63404 Preferential Trade Agreement Policies for Development A HANDBOOK Editors Jean-Pierre Chau our • Jean-Christophe Maur PREFERENTIAL TRADE AGREEMENT POLICIES FOR DEVELOPMENT PREFERENTIAL TRADE AGREEMENT POLICIES FOR DEVELOPMENT A HANDBOOK Jean-Pierre Chauffour and Jean-Christophe Maur, Editors © 2011 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org All rights reserved 1 2 3 4 :: 14 13 12 11 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denomina- tions, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permis- sion may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750- 4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. ISBN: 978-0-8213-8643-9 eISBN: 978-0-8213-8644-6 DOI: 10.1596/978-0-8213-8643-9 Library of Congress Cataloging-in-Publication Data. Preferential trade agreement policies for development : a handbook / Jean-Pierre Chauffour, Jean-Christophe Maur, editors. p. cm. Includes bibliographical references and index. ISBN 978-0-8213-8643-9 — ISBN 978-0-8213-8644-6 (electronic) 1. Developing countries—Commercial policy. 2. Developing countries—Foreign economic relations. 3. Tariff preferences—Developing countries. 4. Free trade—Developing countries. 5. Economic development—Developing coun- tries. I. Chauffour, Jean-Pierre. II. Maur, Jean-Christophe. III. World Bank. HF1413.P69 2011 382'.9091724—dc22 2011014920 Cover illustration: Barrie Maguire, NewsArt.com Cover design: Drew Fasick CONTENTS Foreword xi Acknowledgments xiii About the Editors and Contributors xv Abbreviations xvii Overview 1 Jean-Pierre Chauffour and Jean-Christophe Maur 1 Beyond Market Access 17 Jean-Pierre Chauffour and Jean-Christophe Maur 2 Landscape 37 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard 3 Economics 69 Richard Baldwin 4 North-South Preferential Trade Agreements 95 Bernard Hoekman 5 Customs Unions 111 Soamiely Andriamananjara 6 Preferential Trade Agreements and Multilateral Liberalization 121 Richard Baldwin and Caroline Freund 7 Agriculture 143 Tim Josling 8 Preferential Rules of Origin 161 Paul Brenton 9 Trade Remedy Provisions 179 Thomas J. Prusa 10 Product Standards 197 Jean-Christophe Maur and Ben Shepherd v vi Contents 11 TBT and SPS Measures, in Practice 217 Andrew L. Stoler 12 Services 235 Aaditya Mattoo and Pierre Sauvé 13 Labor Mobility 275 Sherry Stephenson and Gary Hufbauer 14 Investment 307 Sébastien Miroudot 15 Trade Facilitation 327 Jean-Christophe Maur 16 Competition Policy 347 Kamala Dawar and Peter Holmes 17 Government Procurement 367 Kamala Dawar and Simon J. Evenett 18 Intellectual Property Rights 387 Carsten Fink 19 Environment 407 Anuradha R. V. 20 Labor Rights 427 Kimberly Ann Elliott 21 Human Rights 443 Susan Ariel Aaronson 22 Dispute Settlement 467 Amelia Porges Index 503 Boxes 2.1 Typology of Preferential Trade Agreements 38 6.1 Is Bilateralism Bad? 123 7.1 The WTO Agreement on Agriculture 146 9.1 Antidumping Template 186 9.2 Countervailing Duties Template 186 9.3 Global Safeguards Template 186 9.4 Bilateral Safeguards Template 187 10.1 What Are Meta-Standards? 198 10.2 Proliferation and Growing Importance of Product Standards 198 10.3 Do Voluntary Standards Have Cost Effects, Too? 201 10.4 Inventory Methods versus Direct Measures of Restrictiveness 203 10.5 Facilitating Market Access: Harmonization, Equivalence, and Mutual Recognition 204 Contents vii 10.6 Trade Effects of Harmonization: Empirical Evidence 205 10.7 Trade Effects of Mutual Recognition: Empirical Evidence 207 10.8 How Small ASEAN Countries Manage to Access Certification and Accreditation Services 209 10.9 The Codex Alimentarius and Preferential Trade Agreements 213 11.1 WTO Standards and Guidelines on TBT and SPS Measures 218 11.2 Success Story: Orchids to Australia 224 11.3 Impact of SPS Measures in the China-New Zealand PTA 226 11.4 WTO Assessment of TBT Implementation Costs 230 11.5 Using the PTA’s Living Agreement Institutions for Capacity Building: An Example 230 11.6 Dispute Settlement of TBT and SPS Measures in the WTO and within the Andean Community 231 12.1 WTO+ and WTO-Extra Provisions in U.S. and EU PTAs 241 12.2 Harmonization and Mutual Recognition in Services: Promise and Pitfalls 248 12.3 PTAs and Digital Trade 259 12.4 Tourism Liberalization in the EU–CARIFORUM EPA 260 12.5 Cultural Cooperation and Aid for Trade in the EU–CARIFORUM EPA 263 13.1 Labor Mobility in Statistical Terms 277 13.2 Quantitative Estimates of Overall Gains from Greater Labor Mobility 279 14.1 Rules on Investment at the WTO 308 15.1 Definition and Scope of Trade Facilitation in Selected PTAs 328 15.2 Trade Facilitation and the WTO 329 15.3 The Trans-Kalahari Corridor 330 15.4 Mercosur’s Transit and Cross-Border Transport Agreement 339 16.1 Competition Policy and International Cooperation 356 17.1 Persistence of Discrimination: Procurement Practices and the Global Economic Crisis 369 17.2 Three International Government Procurement Instruments 372 17.3 Examples of Flexible Provisions in Government Procurement PTAs 375 18.1 Patent-Registration Linkage and Test Data Protection: The Case of Chile 394 19.1 Considerations for Developing Countries in Negotiating PTAs 412 20.1 Sweatshop Scandal Insurance for Brand-Name Buyers 429 20.2 Labor Rights and the WTO 430 20.3 NAFTA as a Tool for Promoting Rights of Mexican Migrants in the United States 438 20.4 Responding to a Sweatshop Scandal through Capacity Building and Monitoring 439 21.1 Transparency, Due Process, and Democracy Spillovers from the WTO 448 22.1 The Protocol of Olivos 482 Figures 1.1 Most Favored Nation (MFN) Tariff Rates, Weighted Mean, All Products 20 2.1 Total PTA Notifications Received by the World Trade Organization, by Year, 1949–2009 40 2.2 All PTAs Notified to the GATT/WTO, by Year of Entry into Force, 1949–2009 40 2.3 PTAs Notified to the GATT/WTO and in Force, by Year of Entry into Force, 1959–2009 41 2.4 PTAs Notified to the GATT (Pre-1995) and the WTO (Post-1995), by Legal Provision 41 2.5 Evolution of Notified PTAs in Force, by Type of Partner, 1958–2009 43 2.6 Number of PTAs under Negotiation and Signed, by Type of Partner, as of February 2010 43 2.7 Bilateral versus Plurilateral PTAs Notified to the GATT/WTO 44 2.8 Cross-Regional and Intraregional PTAs Notified to the GATT/WTO 44 2.9 PTAs, by Region and by Year of Entry into Force, 2000–09 45 2.10 PTAs in Force and under Negotiation, by Region 46 2.11 PTAs in Force and under Negotiation by Selected Countries and Groupings, as of February 2010 46 2.12 Issues Covered in Regional Trade Agreements, 1989–2009 47 2.13 Participation in Notified PTAs as of February 2010 (Goods) 49 2.14 Participation in Notified EIAs as of February 2010 (Services) 50 2.15 Network of Plurilateral Groupings in Europe and Central Asia 51 2.16 Network of Plurilateral Groupings in the Americas and the Caribbean 52 2.17 Network of Plurilateral Groupings in South Asia, East Asia, and the Pacific 53 2.18 Network of Plurilateral Groupings in Africa and the Middle East 54 viii Contents 2.19 Evolution of the Share of Intra-PTA Imports in Total Imports, 1970–2008 58 2.20 Evolution of the Share of Intraregional Trade in Gross Domestic Product, 1970–2008 60 2.21 Most Favored Nation Applied Tariffs, Trade-Weighted Average of All PTA Members Selected Periods 62 2.22 Proportion of Tariff-Free Imports as a Share of Total Imports, All Goods, Selected PTAs and Periods 62 2.23 Percentage Changes in Trade from Entry into Force of a Preferential Trade Agreement to 2008, Based on Gravity Model Estimates 63 3.1 Trade Pattern for a Simple Preferential Trade Association 70 3.2 Trading Equilibriums in a Preferential Trade Association 71 3.3 Ambiguous Net Welfare Effects 72 3.4 Effects of Preferential Frictional Barrier Liberalization on Prices and Imports 73 3.5 Welfare Effects of Preferential TBT Liberalization: Viner’s Ambiguity Vanishes 73 3.6 Competition (COMP) and Break-Even (BE) Curves 74 3.7 Prices, Output, and Equilibrium Firm Size in a Closed Economy 75 3.8 Prices, Output, and Equilibrium Firm Size with Integration 77 3.9 Welfare Effects of Complete Liberalization 78 3.10 Demand-Linked Circular Causality 79 3.11 Input-Cost-Linked Circular Causality 80 3.12 Locational Equilibrium Diagram 81 3.13 Locational Equilibrium Diagram with Trade Liberalization 82 3.14 Trade Arrangements and Industrialization 82 3.15 Location of Japanese Auto and Electrical Machinery Plants in East Asia, 1975–2004 85 3A.1 Johnson’s Diagram, Small Home and Partner Countries 87 3A.2 The Small PTA Diagram: A Simple Case 88 3B.1 Monopoly Profit Maximization 90 3B.2 Duopolist as Monopolist on Residual Demand: Example of a Nonequilibrium 91 3B.3 Duopoly and Oligopoly: Expectation-Consistent Outputs 91 6.1 Net Welfare Effects, Preferential Trade Agreement to Global Free Trade 125 6.2 Relationship between MFN Tariffs and Home Welfare 125 6.3 Juggernaut Logic 127 6.4 Juggernaut Building Block Logic 128 6.5 Imported MFN Liberalization 131 6.6 An Economic Theory of the GATT 132 8.1 Restrictiveness (R-Index) of Rules of Origin in Free Trade Agreements 171 9.1 Hub-and-Spoke and Cross-Regional Arrangement of PTAs 188 9.2 Intra-PTA Antidumping Filings, Sample of 74 PTAs 194 10.1 Elements of a Standards Infrastructure 200 12.1 Services-Related PTAs as a Share of Total PTA Notifications to the WTO, 2010 236 12.2 Services-Related PTAs Notified to the WTO, by Country Group 236 12.3 Sectoral Coverage of PTAs and of GATS Offers and Schedules, Selected Countries 257 12.4 GATS+ Advances in East Asian PTAs with Services Provisions, by Sector 261 12.5 GATS+ Advances in East Asian PTAs with Services Provisions, by Mode 262 12A.1 GATS and EU–CARIFORUM Commitments Compared: Barbados 266 12A.2 GATS and EU–CARIFORUM Commitments Compared: Dominican Republic 267 12A.3 GATS and EU–CARIFORUM Commitments Compared: Jamaica 268 12A.4 GATS and EU–CARIFORUM Commitments Compared: Trinidad and Tobago 269 13.1 Theoretical Gains from Liberalization of Mode 4 278 13.2 Theoretical Effect on Developed Countries of Liberalization of Mode 4 279 13.3 Theoretical Effect on Developing Countries of Liberalization of Mode 4 279 13A.1 Provisions on Mode 4 in PTAs between the United States and Developing Countries 300 13A.2 Provisions on Mode 4 in PTAs between Canada and Developing Countries 301 13A.3 Provisions on Mode 4 in PTAs between the European Union (EU) and Developing Countries 302 13A.4 Provisions on Mode 4 in PTAs between Japan and Developing Countries 302 13A.5 Provisions on Mode 4 in PTAs between Australia and New Zealand and Developing Countries 303 13A.6 Provisions on Mode 4 in PTAs between Developed and Developing Countries 304 14.1 Total Number of PTAs and Number with Investment Provisions, 1970–2009 308 17.1 PTAs Containing Government Procurement Provisions, 2009 371 Contents ix Tables 1.1 Types and Scope of Regulatory Objectives in Selected Areas Covered by Trade Liberalization Agreements 29 2.1 Deep Commitments in Selected EU and U.S. PTAs, by Type of Provision 48 2.2 Correlation between Intra-PTA and Total Export Growth Rates, 1970–2008 61 2.3 Estimation Results of the Gravity Model of the Average Trade between Two Partners, Selected PTAs 61 2A.1 Membership of Selected Plurilateral Preferential Trade Agreements 64 3.1 ASEAN Tariffs on Engines and Automobiles, Most Favored Nation (MFN) Tariffs and Common Effective Preference Tariffs (CEPTs), 2008 84 3.2 Intraindustry Trade as a Share of Internal and External Trade of PTAs 86 4.1 European Union (EU) Instruments and EU and National Objectives 106 5.1 Selected Customs Unions, in Force and Planned 112 7.1 Summary of Provisions Affecting Agriculture in NAFTA, U.S.–Chile, U.S.–Australia, and CAFTA Agreements 156 8.1 Summary of Methods for Determining Origin 164 9.1 Trade-Contingent Initiations and Measures in PTAs, 1995–2007 181 9.2 Contingent Protection Rules in Selected PTAs 182 9.3 Characteristics of PTAs 184 9.4 Summary of Contingent Protection Rules in PTAs 185 9.5 Cross-Tabulation of Contingent Protection Rules, by Hub 189 9.6 Antidumping Template for Selected PTA Hubs 189 9.7 Countervailing Duties Template for Selected PTA Hubs 190 9.8 Global Safeguards Template for Selected PTA Hubs 191 9.9 Bilateral Safeguards Template for Selected PTA Hubs 192 9.10 Characteristics of PTAs That Have Disallowed Trade Remedies 193 9.11 Antidumping Activity, by PTA Status 194 10.1 Prevalence of Harmonization and Mutual Recognition in Preferential Trade Agreements 207 10.2 Content of Preferential Trade Agreements Relating to Provisions on Standards 208 11.1 Comparison of Main Features Relating to TBT and SPS Measures, 11 PTAs 219 12.1 Key Disciplines in PTAs That Cover Services 244 12.2 Negotiating Approaches in Services Trade 253 12.3 Key Features of PTAs That Cover Services 255 12.4 Average Percentage of Subsectors Subject to Market Access Commitments on Mode 3, Selected Country Groupings 258 12.5 GATS Commitments, GATS DDA Offers, and “Best� PTA Commitments for All Members Reviewed, Selected Sectors 258 12A.1 Preferential Trade Agreements (PTAs) That Include Provisions on Trade in Services 264 12A.2 Classification of Preferential Trade Agreements (PTAs) Featuring Services Provisions by Country Group 265 13.1 Bilateral Labor Agreements with Developing-Country Partners: Government Programs for Temporary Workers 288 13A.1 Quantitative Estimates of Gains from Increased Labor Mobility 292 13A.2 Agreements between the United States and Developing Countries 293 13A.3 Agreements between Canada and Developing Countries 294 13A.4 Agreements between the European Union and Developing Countries 295 13A.5 Agreements between Japan and Developing Countries 296 13A.6 Agreements between Australia and New Zealand and Developing Countries 297 14A.1 Overview of Recent PTAs Covering Investment 322 14A.2 Selected Empirical Studies on the Impact of Bilateral Investment Measures 324 15.1 Functions of Selected Corridors 338 16.1 Cartel Overcharges and Deterrent Effect, Vitamin Industry, 1990s 354 16.2 Models of Regional Competition Regimes 357 18.1 U.S. Free Trade Agreement (FTA) Landscape 390 18.2 Principal TRIPS+ Provisions in U.S. Free Trade Agreements (FTAs) Ratified between 2001 and 2006 392 18.3 European Union (EU) Preferential Trade Agreements (PTAs) and Economic Partnership Agreements (EPAs) 398 20.1 Sanctions Authorized for Labor Violations in U.S. Preferential Trade Agreements 436 21.1 Examples of Human Rights Embedded in PTAs: Demandeurs and Position of Provisions in Agreement 445 x Contents 21.2 The Universal Declaration on Human Rights and Its Two Covenants 446 21.3 Examples of Avenues and Actions at the WTO Related to Human Rights, 2005–10 447 21.4 Human Rights in Preferential Trade Agreements: Comparing EFTA, the EU, the United States, and Canada 449 21.5 Examples of Human Rights Embedded in Preferential Trade Agreements 450 FOREWORD Regional integration is increasingly recognized as a key avenue for promoting economic growth and reducing poverty. Preferential trade agreements (PTAs) have become a central instrument of regional integration in all parts of the world. Beyond market access and the progressive elimination of barriers at the border, PTAs are increasingly being used to address a host of behind-the-border issues, also known as “deep integration� issues, in order to promote cooperation in the areas of investment, trade facilitation, competition policy, and government procurement, as well as wider social issues related to the regulation of the environment and the protection of labor and human rights. While the multilateral route to trade integration remains the first-best option, the stalling of the Doha Round of nego- tiations has led to a temporary impasse. Countries—developed and developing alike—have turned to the regional or bilat- eral route. With close to 300 PTAs notified to the World Trade Organization, regionalism has become a reality on the ground. Many countries are members of multiple PTAs, and the pace of negotiations on new agreements is accelerating. While it has been known for a long time that the traditional preferential market access elements of PTAs are likely to be suboptimal from a welfare perspective, as compared to multilateral or even unilateral liberalization, and that third parties often suffer from these arrangements, policy makers around the world expect these costs to be dwarfed by the deep inte- gration benefits of modern PTAs. The purpose of this handbook on preferential trade agreement policies for development is to explore the various ways in which policy makers and trade negotiators in the developing world can limit the costs and maximize the benefits of their regional integration efforts. Today’s modern PTAs are shaping a broad and comprehensive reform agenda that developing countries can adopt and implement with full ownership and mutual accountability. Preferential market access is no longer the predominant motive. Increasingly important is the use of PTAs to promote competitiveness, upgrade production stan- dards, liberalize services, modernize regulatory regimes, promote labor mobility, protect intellectual property, improve governance, and foster transparency and the rule of law; and, in time, to help build common regional values and norms for a more peaceful and prosperous world. Open regionalism as a complement to a freer and more transparent, rules-based multilateral trading system has been promoted by the World Bank for many years. Regional integration continues to play a positive transformational role in Europe, North America, East Asia, and Latin America, and the same forces are poised to deepen integration in the Middle East and North Africa, South Asia, and Sub-Saharan Africa. In the course of wide-ranging consultations on the World Bank’s upcoming international trade strategy, regional integration and cross-border trade cooperation emerged as one of the four main themes. I hope that this handbook—the collective effort of some of the world’s most renowned trade economists—enhances the understanding of various institutional arrangements and their possible development implications, thereby helping realize the promise of open regionalism and trade for poverty reduction. Mahmoud Mohieldin Managing Director World Bank Group xi ACKNOWLEDGMENTS This handbook is the product of a rich and fruitful collaboration among an outstanding set of distinguished individuals— economists, lawyers, and professional practitioners from around the world on the challenges and opportunities of preferential trade agreements (PTAs) for developing countries. This collaboration was made possible by the support of the Multi- Donor Trust Fund for Trade and Development (MDTF-TD) financed by contributions from the governments of Finland, Norway, Sweden, and the United Kingdom. The editors would like to extend their special thanks to the 26 authors who contributed to this volume for the quality of their research, their professional insights, and their patience in dealing with our multiple requests from the initial concep- tual stages, first drafts, lectures, and final assembly of the material. Without their unique knowledge and expertise, the preparation of this handbook would simply not have been possible. This book is the result of the close cooperation of two World Bank vice presidencies: Otaviano Canuto, vice president of the Poverty Reduction and Economic Management (PREM) Network, and Sanjay Pradhan, vice president of the World Bank Institute (WBI). Strong support from the Development Economics and Africa vice presidencies must also be acknowledged, in particular through substantive contributions to chapters of this book and insightful comments. The World Bank project on preferential trade agreements was originally the brain child of Uri Dadush, director of the World Bank International Trade Department, who in the Global Economic Prospects 2005: Trade, Regionalism, and Develop- ment (Washington, DC: World Bank, 2005) rightfully identified regionalism as an increasingly complex yet potentially pro- ductive avenue for promoting trade, economic integration, and development. When Bernard Hoekman took over the direction of the department, he not only provided the overall intellectual guidance for this project but also magnified its reach and helped us assemble the best possible crew of advisers and reviewers. Among them, we are particularly grateful to Richard Baldwin, Clem Boonekamp, Olivier Cattaneo, Jaime de Melo, Antoni Estevadeordal, Carsten Fink, Caroline Freund, Daria Goldstein, Mona Haddad, Gary Hufbauer, Nuno Limão, Patrick Low, Richard Newfarmer, Marcelo Olarreaga, and Sherry Stephenson for their overall guidance and support in various phases of this project. This endeavor was complemented by a World Bank Institute initiative, under the lead of Roumeen Islam, manager of the Poverty Reduction team, to bring focus and an entirely new program of activities on regional integration. This initiative turned out to be instrumental in the conception of this handbook. Notwithstanding fierce competing work priorities, Mona Haddad, sector manager (PREM), and Raj Nallari, manager (WBI), provided constant support without which the realization of this large project would not have been feasible. The editors would also like to acknowledge the contributions of many reviewers of individual chapters: Susan Aaronson, Rolf Adlung, Julia Almeida Salles, Bruce Blonigen, Olivier Cadot, Steve Charnovitz, Meredith Crowley, Jaime de Melo, Daria Goldstein, Lee-Ann Jackson, Steven Jaffee, Michael Jensen, Muthukumara Mani, Toni Matsudaira, Gerard McLinden, Roberta Piermartini, Daniel Sokol, and Jon Strand. Other colleagues at the World Bank have volunteered to share their knowledge and enthusiasm and provided invaluable advice and recommendations throughout the process, including Jean- François Arvis, Aaditya Mattoo, Maurice Schiff, Ravindra Yatawara, and Gianni Zanini. A special appreciation goes naturally to our experienced team of peer reviewers, who shared their international expert- ise and helped improve both the scope and focus of the entire manuscript: Ndiame Diop, Antoni Estevadeordal, and Richard Newfarmer. A book cannot exist without an effective production team. This book benefited from the impeccable professionalism of the World Bank’s Office of the Publisher. Stephen McGroarty and Mark Ingebretsen managed the publication process in xiii xiv Acknowledgments the smoothest possible way. Nancy Levine did an outstanding job at copyediting the entire volume. She deserves most of the credit for ensuring the readability of the technical parts of the book, and for making it, if not a page turner, an accessi- ble tool for experts. We would also like to thank the dedicated and professional support provided by the administrative team in the International Trade Department and WBI, including Cynthia Abidin-Saurman, Anita Chen, Nene Mane, Rebecca Martin, Anita Nyajur, Vasumathi Rollakanty, and Amelia Yuson. Special thanks also to Charumathi Rama Rao, who provided support on the financial management aspects of the project, and to Stacey Chow, who effectively coordinates the International Trade Department’s publication program. The editors would also like to thank the participants in the conference on the European and Asian approaches to Deep Integration co-hosted by the Centre for the Analysis of Regional Integration at Sussex (CARIS) at Sussex University on September 14–15, 2009; in particular, Michael Gasiorek, Peter Holmes, Jim Rollo, Zhen Khun Wang, and Alan Winters for useful and constructive feedback on early drafts of the chapters. Our thanks also go to the participants to the 2009–10 WBI courses on PTA and Development in Washington, DC, Dakar, and Arusha, and particularly to those who have contributed to their success: Raymond Boumbouya, Caiphas Chekwoti, Göte Hansson, Peter Kiuluku, Tharcisse Ntilivamunda, and Ina Hoxha Zaloshnja. A particular thought goes to the late Dipo Busari and his family. We owe a lot to the excellence of instructors and speakers who contributed to the courses: Richard Baldwin, Paul Brenton, Nora Dihel, Kimberley Elliott, Simon Evenett, Carsten Fink, Caroline Freund, Larry Hinkle, Bernard Hoekman, Peter Holmes, Gary Hufbauer, Oliver Jammes, Tim Josling, Charles Kunaka, Thea Lee, Javier Lopez, Aadittya Mattoo, Bonard Mawpe, Mary Mbithi, Abdoulaye Ndiaye, Ibrahima Bouna Niang, Tom Prusa, Andrew Roberts, Sebastian Sáez, Pierre Sauvé, Ben Shepherd, Yolanda Strachan, David Tarr, and Gianni Zanini. These thanks should not associate in any way our collaborators and partners in the production of this book to any remaining errors and shortcomings, which remain solely those of the editors. ABOUT THE EDITORS AND CONTRIBUTORS The Editors Jean-Pierre Chauffour is lead economist in the World Bank’s International Trade Department, Poverty Reduction and Economic Management (PREM) network, where he works on regionalism, competitiveness, and trade policy issues. Prior to joining the Bank in 2007, he spent 15 years at the International Monetary Fund (IMF) where he held various positions, including mission chief in the African Department and representative to the World Trade Organization (WTO) and United Nations (UN) in Geneva. Mr. Chauffour has extensive economic policy experience and has worked in many areas of the developing world, most extensively in Africa, the Middle East, and Eastern Europe. He holds master’s degrees in economics and in money, banking, and finance from the Panthéon-Sorbonne University in Paris, France. He is the author of The Power of Freedom: Uniting Human Rights and Development (Washington, DC: Cato Institute, 2009). Jean-Christophe Maur is senior economist in the growth and competitiveness practice of the World Bank Institute and a fellow with the Group d’Economie Mondiale at the Institut d’Etudes Politiques de Paris. His current areas of work cover regional integration issues, services liberalization, and contributing to the World Bank’s Development Debates platform. Mr. Maur joined the World Bank in 2008 from the U.K. Department of International Development, where he participated in U.K. trade negotiations in trade facilitation and goods market access. He was also tasked to manage the multilateral trade assistance cooperation with the World Bank and contributed to create the multi-donor trust fund for trade and develop- ment. His research interests cover regional trade integration and public goods, trade facilitation, standards, and intellectual property rights. He holds a doctorate in economics from Institut d’Etudes Politiques de Paris, and is a graduate of the ESSEC Business School. He was also a visiting fellow at Harvard University. The Contributors Susan Ariel Aaronson, associate research professor of international affairs, George Washington University, Washington, DC, and research fellow, World Trade Institute, Bern. Rohini Acharya, chief of regional trade agreements section, World Trade Organization, Geneva. Soamiely Andriamananjara, senior economist, Growth and Competitiveness Practice, World Bank Institute, Washington, DC. Richard Baldwin, professor of international economics, Graduate Institute of International and Development Studies, Geneva. Paul Brenton, lead economist (Trade and Regional Integration), Poverty Reduction and Economic Management, Africa Region, World Bank, London. Jo-Ann Crawford, counselor, World Trade Organization, Geneva. Kamala Dawar, Graduate Institute of International and Development Studies, Geneva. Kimberly Ann Elliott, senior fellow, Center for Global Development, Washington, DC. xv xvi About the Editors and Contributors Simon Evenett, professor of international trade and economic development, University of St. Gallen, Switzerland. Carsten Fink, chief economist, World Intellectual Property Organization, Geneva. Caroline Freund, chief economist, Middle East and North Africa Region, World Bank, Washington, DC. Bernard Hoekman, sector director, International Trade Department, World Bank, Washington, DC. Peter Holmes, Jean Monnet Reader in the Economics of European Integration, University of Sussex, Brighton, U.K. Gary Hufbauer, Reginald Jones Senior Fellow, Peterson Institute for International Economics, Washington, DC. Tim Josling, professor emeritus, Food Research Institute, Stanford University, Stanford, California. Maryla Maliszewska, economist, Development Prospects Group, World Bank, Washington, DC. Aaditya Mattoo, research manager, Trade and International Integration, World Bank, Washington, DC. Sébastien Miroudot, trade policy analyst, Organisation for Economic Co-operation and Development (OECD), Paris. Amelia Porges, principal, Law Offices of Amelia Porges PLLC, Arlington, Virginia. Thomas Prusa, professor of economics, Rutgers University, Brunswick, New Jersey. Anuradha R.V., partner, Clarus Law Associates, New Delhi. Christelle Renard, RTA database administrator, World Trade Organization, Geneva. Pierre Sauvé, deputy managing director and director of studies, World Trade Institute, University of Bern. Ben Shepherd, principal, Developing Trade Consultants Ltd., New York City. Sherry Stephenson, head, institutional relations, Organization of American States, Washington, DC. Andrew Stoler, executive director, Institute for International Trade, University of Adelaide, Australia. ABBREVIATIONS AB Appellate Body (WTO) ACP African, Caribbean, and Pacific (States) ACWL Advisory Centre on WTO Law ADR alternative dispute resolution AFAS ASEAN Framework Agreement on Services AFTA ASEAN Free Trade Area AGOA African Growth and Opportunity Act (U.S.) ANZCERTA Australia–New Zealand Closer Economic Relations Trade Agreement ANZGPA Australia and New Zealand Government Procurement Agreement APEC Asia-Pacific Economic Cooperation APTA Asia-Pacific Trade Agreement ASEAN Association of Southeast Asian Nations ATIGA ASEAN Trade in Goods Agreement ATJ Andean Tribunal of Justice AUSFTA Australia–U.S. Free Trade Agreement BIT bilateral investment treaty BLA bilateral labor agreement BTA bilateral trade agreement CACJ Central American Court of Justice CACM Central American Common Market CAFTA Central America Free Trade Agreement CAFTA–DR Dominican Republic–CAFTA CAN Comunidad Andina (Andean Community) CAP Common Agricultural Policy (EU) CARICOM Caribbean Community CARIFORUM Caribbean Forum of African, Caribbean, and Pacific (ACP) States CEC Council for Economic Cooperation (NAFTA); Commission of the European Communities CEFTA Central European Free Trade Agreement CEMAC Economic and Monetary Community of Central Africa (Communauté Économique et Monétaire de l’Afrique Centrale) CEPA Closer Economic Partnership Agreement (China) CEPAL/ECLAC Comisión Económica para América Latina/Economic Commission for Latin America and the Caribbean CEPT common effective preference tariff CET common external tariff CIF cost, insurance, and freight CIS Commonwealth of Independent States COMESA Common Market for Eastern and Southern Africa CTC change of tariff classification xvii xviii Abbreviations CUSFTA Canada–U.S. Free Trade Agreement CVD countervailing duties DDA Doha Development Agenda DSB Dispute Settlement Body (WTO) DSM Dispute Settlement Mechanism (WTO) DSU Dispute Settlement Understanding (WTO) DTT double-taxation treaty EAC East African Community; environmental affairs council EBA Everything But Arms (EU) ECJ European Court of Justice ECLAC/CEPAL Economic Commission for Latin America and the Caribbean/ Comisión Económica para América Latina ECO Economic Cooperation Organization ECOWAS Economic Community of West African States EDI electronic data interchange EEA European Economic Area EEC European Economic Community (for historical references) EFTA European Free Trade Association EGS environmental goods and services EIA economic integration agreement; environmental impact assessment ENP European Neighborhood Policy EPA economic partnership agreement ESM emergency safeguard mechanism EU European Union FAO Food and Agriculture Organization (of the United Nations) FDI foreign direct investment FOB free on board FTA free trade agreement; free trade area FTAA Free Trade Area of the Americas FTC Free Trade Commission (NAFTA) GAFTA Greater Arab Free Trade Agreement GATS General Agreement on Trade in Services GATT General Agreement on Tariffs and Trade GCC Gulf Cooperation Council GDP gross domestic product GI geographical indication GNP gross national product GPA Government Procurement Agreement (WTO) GSP generalized system of preferences HS Harmonized System ICCPR International Covenant on Civil and Political Rights ICESCR International Covenant on Economic, Social and Cultural Rights ICN International Competition Network ICSID International Centre for the Settlement of Investment Disputes ICT information and communication technology ICTSD International Center for Trade and Sustainable Development IMF International Monetary Fund IPR intellectual property right ISO International Organization for Standardization IT information technology ITO International Trade Organization Abbreviations xix LAIA/ALADI Latin American Integration Association/Asociación Latinoamericana de Integración Mercosur Southern Cone Common Market (Mercado Común del Sur) MFA Multifibre Arrangement MFN most favored nation MLAT mutual legal assistance treaty MNC multinational corporation MNE multinational enterprise MRA mutual recognition agreement MTN multilateral trade negotiation NAAEC North American Agreement on Environmental Cooperation (side agreement to NAFTA) NAALC North American Agreement on Labor Cooperation (side agreement to NAFTA) NACEC North American Commission for Environmental Cooperation NAFTA North American Free Trade Agreement NGO nongovernmental organization NTM nontariff measure OAS Organization of American States OCT Overseas Countries and Territories OECD Organisation for Economic Co-operation and Development OSBP one-stop border post PAFTA Pan-Arab Free Trade Area PICTA Pacific Island Countries Trade Agreement PRTR pollutant release and transfer registry PTA preferential trade agreement REIO regional economic integration organization ROO rule of origin RoW Rest of the World RTA regional trade agreement; reciprocal preferential agreement SAARC South Asian Association for Regional Cooperation SACU Southern African Customs Union SAD single administrative document SADC Southern African Development Community SAFTA South Asian Free Trade Area; Singapore–Australia Free Trade Agreement SAR special administrative region SIA sustainability impact assessment SME small and medium-size enterprise SPARTECA South Pacific Regional Trade and Economic Co-operation Agreement SPS sanitary and phytosanitary TBT technical barriers to trade TDCA Trade, Development, and Cooperation Agreement (EU–South Africa) TEC Treaty establishing the European Community TFAP trade facilitation action plan TIFA trade and investment framework agreement TPA Trade Promotion Act of 2002 (U.S.) TPR Tribunal Permanente de Revisión (Mercosur) TPRM Trade Policy Review Mechanism (WTO) TPSEP Trans-Pacific Strategic Economic Partnership TRIMS Trade-Related Investment Measures TRIPS Trade-Related Aspects of Intellectual Property Rights TRQ tariff-rate quota TWP temporary worker program UDHR Universal Declaration of Human Rights xx Abbreviations UNCITRAL United Nations Commission on International Trade Law UNCTAD United Nations Conference on Trade and Development USITC U.S. International Trade Commission USTR United States Trade Representative VER voluntary export restraint WAEMU/UEMOA West African Economic and Monetary Union/Union Économique et Monétaire Ouest-Africaine WCO World Customs Organization WIPO World Intellectual Property Organization WTO World Trade Organization Overview Jean-Pierre Chauffour and Jean-Christophe Maur Preferential trade agreements (PTAs) have become a cor- improvements in productivity brought about by the trans- nerstone of the international trade system. The surge in fer of improved technology, learning by doing, and their number and scope is fast reshaping the architecture of increased competition. Bilateral or regional integration can the world trading system and the trading environment of be an important engine of trade competitiveness, both for developing countries. The integration of these diverse small, very poor, landlocked countries and for less region- agreements into a multilateral framework that facilitates ally integrated or diversified middle-income countries. the expansion of trade is likely to be one of the main chal- At the same time, the multitude of PTAs is becoming lenges facing the world trading system in the coming years. cumbersome to manage for many developing countries. Hundreds of preferential agreements—free trade agree- As agreements proliferate, countries become members of ments and customs unions that involve reciprocal tariff several different agreements. The average African country, reductions—are currently in force, including close to 300 for instance, belongs to four different agreements, and the that had been notified to the World Trade Organization average Latin America country belongs to seven. This (WTO) as of end-2010. Such a proliferation of regional creates what has been referred to as a “spaghetti bowl� of and bilateral PTAs, together with the sluggishness of Doha overlapping arrangements, often with different tariff Round negotiations toward a new multilateral trade agree- schedules, different exclusions of particular sectors or ment, pose serious challenges to the promotion of a more products, different periods of implementation, different open, transparent, rules-based multilateral trading system. rules of origin, different customs procedures, and so on. Although PTAs may promote development, they necessar- The proliferation of bilateral and regional PTAs may ily discriminate against nonmembers and can therefore undermine progress toward a more open, transparent, and lead to trade diversion in a way that hurts both member rules-based multilateral trading system. countries and excluded countries. At the same time, PTAs This Handbook offers an introduction to the complex create larger and more competitive markets and benefit world of modern PTAs. It follows in the steps of earlier, producers and consumers through economies of scale and seminal World Bank publications on the economics and lower prices, among other effects. Beyond market access practice of PTAs, notably New Dimensions in Regional Inte- motives, PTAs are increasingly used as engines of change in gration (De Melo and Panagariya 1996), Trade Blocs (World many developing countries, to promote, implement, and Bank 2000), and Regional Integration and Development lock in reforms in a wide range of policy areas such as (Schiff and Winters 2003). Supplementing these earlier investment regimes, competition rules, and government publications, this volume aims at taking its audience procurement. beyond the traditional market access paradigm to consider In many developing countries, regional integration has more broadly and systematically the numerous regulatory become a key means of promoting economic growth and policy dimensions that are contained in modern PTAs. In fighting poverty. In fact, no low-income country has man- particular, it offers a framework for understanding a num- aged to grow and sustainably reduce poverty without ber of behind-the-border policies typically covered in global or regional trade integration. In the short term, PTAs, including labor mobility, investment, trade facilita- regional trade contributes to growth by expanding markets tion, competition, and government procurement, as well as for goods and services. In the medium to long term, other societal and more normative policies related to intel- regional integration contributes to growth through lectual property, environment, labor rights, and human 1 2 Jean-Pierre Chauffour and Jean-Christophe Maur rights. These latter are increasingly among the policies tions of these links for development, rather than to assess driven by powerful trading blocs as they strive to influence the individual merits of given PTA initiatives. developing countries and the evolution of the global trad- The remainder of this overview offers a brief survey of ing system. the contents of the Handbook to help readers navigate The Handbook is also inspired by the numerous among the topics and to put the various parts in perspective. requests received by the World Bank from developing countries or groups of developing countries worldwide for A Road Map of the Handbook advice on PTAs, including those currently being negoti- ated, as an aid in understanding the obligations and the This volume consists of 22 mostly original and unpub- possible economic and development implications of vari- lished chapters written by renowned international trade ous provisions. In light of these requests, and in view of the academics and experts. They originate from the course on growing but fragmented knowledge on PTAs, the Hand- preferential trade agreements designed by the World Bank book seeks to provide a first point of entry into the issues, Institute and from the regional integration work of the to inform the policy debate in this area, and to help trade World Bank’s International Trade Department. policy makers, researchers, and practitioners, among Chapter 1 highlights the main theme and raison d’être others, better understand and navigate the world of PTAs. of the Handbook: that modern PTAs are essentially preoc- In particular, the book has been written with a view to cupied by a wide range of behind-the-border and deep informing policy makers and trade negotiators in devel- integration issues that represent considerable opportuni- oping countries about the economic and development ties and challenges for low-income countries, beyond the implications of different approaches to the most impor- benefits of market access. Chapter 2 maps the landscape of tant sectoral provisions in PTAs. The ultimate objective of regionalism, chapter 3 presents the economic theory of the Handbook is to help developing countries use PTAs PTAs, and chapter 4 describes the characteristics of poten- proactively, as instruments to bolster their trade competi- tially development-friendly PTAs. Chapter 5 examines the tiveness and leverage the global economy to promote specificities of customs unions, one of the most advanced growth and poverty reduction. forms of PTA. Chapter 6 discusses the tension between The content of the Handbook is largely nontechnical regionalism and multilateralism. Next, chapters 7 through and has been written to be accessible for a large audience 17 survey current practice in the main policy areas typically of policy makers, general academics, and other stake- covered in PTAs: agriculture policy, rules of origin, product holders who are not necessarily economists or legal spe- standards, technical barriers, trade remedies, services, labor cialists. It is designed to be used for both graduate and mobility, and deep integration areas related to investment, undergraduate teaching in economics, international rela- trade facilitation, competition policy, and government tions, political science, and law, but it may also be of spe- procurement. Chapters 18 through 21 review a number of cial interest to informed readers who may not have more normative issues that are increasingly incorporated detailed knowledge of all aspects of PTAs. One of its key into PTAs—intellectual property, environment, labor objectives is to offer an entry point to specialized areas rights, and human rights. Chapter 22 closes the volume covered in PTAs. For this reason, each chapter can be with a discussion of the cross-cutting issue of dispute read independently, as a guide to the most salient issues settlement in PTAs. arising in PTAs today. Given the rapidly evolving nature From a methodological standpoint, each policy area of many issues discussed in the Handbook, the volume is has been researched with a set of generic questions in to be regarded as a window into the issues covered by mind. How do provisions in a given policy area compare PTAs, not as a definitive appraisal. across selected relevant PTAs and (when applicable) with Although the global evolution of preferentialism WTO rules? What are the legal and economic implications betrays the influence of particular countries and integra- of different formulations? Are there particular economic tion initiatives, the Handbook purposely chooses not to development benefits or costs associated with different discuss as stand-alone issues important bilateral or pluri- approaches? To what extent are third parties being dis- lateral initiatives led by large economies such as the criminated against, and are PTA provisions conducive to European Union (EU) and the United States. Rather, it open regionalism? How binding are the provisions in the refers throughout to specific relevant examples. The focus short term and the longer term? What dispute settlement is chiefly to promote understanding, from an analytical mechanisms would be used in case of disagreements, and perspective, of the links between sectoral issues in deep how? Are there particular approaches (e.g., hard law ver- integration and preferential liberalization and the implica- sus soft law) that can be advocated or that should be Overview 3 avoided? What are the implications of the provisions for characterized by low external barriers to trade, liberalized building up institutions and for technical assistance services markets, nonrestrictive rules of origins for services needs? To what extent do the provisions help deliver as well as for goods, a focus on reducing transaction costs regional public goods or tackle market failures that are at borders, and transparency and the availability of due regional in nature? process. Another answer to complexity, and one not sufficiently considered, according to the authors, by developing coun- Beyond Market Access tries, is selectivity. Liberalization is a complex matter, not Beyond market access, PTAs have become potential only from a capacity standpoint, but also politically. Over- instruments of choice for many developing countries for loading the negotiating agenda (which will later become promoting wide-ranging reforms and promoting owner- the implementing agenda) creates a distraction from what ship, mutual accountability, and results. In chapter 1, Jean- may be achievable and where gains may be the most Pierre Chauffour and Jean-Christophe Maur discuss important. Agreements bloated by too many issues may why, for many low-income countries, PTAs are increas- lose significance and fail to achieve much. ingly the core of a credible development strategy for Finally, the dynamics of North-South, South-South, and accelerating economic growth and reducing poverty. North-North PTAs differ considerably. Asymmetric agree- Choosing meaningful issues, with the right partner, with ments make cooperation less easy and may provide less adequate technical assistance, and employing a cooperative scope for transnational public goods and mutual recogni- approach, may bring about substantial progress toward lib- tion but may open greater prospects for lock-in and for eralization and can serve as a positive signal or trigger for access to imported regulatory regimes, when needed. Mar- more challenging areas. To be sure, the maintenance of ket access considerations will dominate for the small part- high border barriers toward third parties entails discrimi- ner, whereas the larger partner will seek, beyond that, to nation and costs, and traditional PTAs clearly represent an diffuse its regulatory norms, including values norms, and inferior solution compared with multilateral liberalization. to trigger competitive liberalization effects in partner But market access is no longer the only or even the main countries. item on the agenda of negotiators, especially those of developing countries, since deep integration really involves Landscape the contemplation of a domestic reform strategy. In this respect, prioritization of core objectives should be a central Today’s multilateral trading system is characterized by a consideration of negotiators. criss-crossing web of ever more complex PTAs. In chapter 2, From a theoretical standpoint, the economic paradigm Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, of traditional or shallow PTAs does not necessarily apply to and Christelle Renard survey the landscape of PTAs and deep and comprehensive PTAs. Concepts such as mercan- note that recent developments and trends can be summed tilist reciprocal liberalization, trade creation and diversion, up in a number of stylized facts: or a textual approach toward the design of PTAs may still underpin the reasoning of many policy makers, but they • PTAs have become ubiquitous, and participation in are often not valid, or only partially explanatory, for deep PTAs is becoming more diverse, spreading to most geo- integration liberalization. Discrimination is still an issue in graphic regions, especially East Asia and the Pacific. deep PTAs, but it may not be the pervasive problem it is in North-South preferential partnerships are on the rise, as a goods-only PTAs, especially if the principle of open access number of developing countries elect to forgo unilateral to any regulatory treatment is respected. programs such as the generalized system of preferences More worrying than discrimination, perhaps, in view of (GSP) in favor of reciprocal agreements. Cross-regional the desirable objective of multilateral liberalization, is the agreements are also expanding, in an attempt to keep on inherent complexity created by overlapping and conflicting a level playing field with other countries that are also regulatory regimes in the myriad of PTAs. This concern is gaining preferences. already clearly identified, and the call for multilateralizing • There is some consolidation of PTA networks as bilateral regionalism already voiced, in the WTO. Complexity relationships are replaced by plurilateral PTAs among the requires that some core principles should be followed. The same partners. This is the case, for instance, in Latin key to making PTAs complementary to a nondiscrimina- America. Agreements between regional blocs—for exam- tory multilateral system and supportive of development is ple, between the European Free Trade Association (EFTA) to strive for open regionalism in the shape of agreements and the Southern African Customs Union (SACU)—are 4 Jean-Pierre Chauffour and Jean-Christophe Maur also on the increase. In Asia, countries long resistant within the PTA to ensure welfare gains for all. The poten- to preferential trade liberalization are catching up, tial negative effect of PTAs on third countries could be with the emergence of plurilateral PTAs that coexist addressed by lowering PTA external tariffs to leave the alongside bilateral PTAs involving the same sets of third countries’ trade levels unchanged. Even though this is partners. rarely how PTAs operate in practice, they are, in theory, not • The structural configuration of PTAs is changing, necessarily bad for world welfare, from a static perspective. with bilateral PTAs becoming increasingly the norm. There is more to the economics of PTAs than the mere Such PTAs can be concluded more rapidly, and they mechanistic static effects. PTAs can generate complex confirm a shift away from using PTAs as a means of dynamic gains that operate by changing the rate at which forging traditional regional partnerships among geo- new factors of production, mainly capital, are accumu- graphically proximate countries and toward using lated. Signers of regional trade deals have long emphasized them to negotiate strategic, bilateral market access, the importance of the dynamic, or growth, effect of PTAs, often among countries in different regions. Indeed, which may sensibly alter the economic benefits of prefer- cross-regional PTAs account for two-thirds of those ential liberalization. One of the mainstay justifications for currently under negotiation. PTAs is the belief that uniting small economies will make • The regulatory scope of PTAs is rapidly becoming regional firms more efficient and more competitive by broader and deeper. On issues that fall under the cur- allowing them access to a bigger market. Indeed, in the rent mandate of the WTO, some countries have elected presence of imperfect competition and economies of scale, to take on bilateral commitments that exceed those they the size of the market matters. Preferential liberalization have accepted at the multilateral level—that are and defragmentation of national markets lead immediately “WTO+.� In addition, some countries are undertaking to more competition because more firms are present in the “WTO-extra� commitments in PTAs on issues that lie market. This, in turn, results in industrial restructuring outside the current WTO mandate. and upgrading because firms need to grow to cover their • For a number of plurilateral PTAs, intra-PTA imports as costs in the more competitive environment, and the least a share of total imports have increased, and both intra- efficient firms exit the market. In the end, the region is left PTA exports and total exports have grown. PTA partners with a more efficient industrial structure, with fewer, big- in selected plurilateral PTAs seem to trade more inter- ger, more efficient firms competing more effectively with nally than would be expected in the absence of a PTA, each other. and the impact on extra-PTA exports and imports is A further dynamic effect from preferential liberaliza- largely positive. tion may stem from the location decisions of firms in the PTA—a source of potential distributional impact within the preference zone. Firms may benefit from the existence Economics of positive externalities (or agglomeration forces) associ- The theoretical effect of PTAs depends on the efficiency ated with being located close to where markets and other with which economic resources are allocated within and producers are. However, there are also negative externali- among countries as a result of partial trade liberalization, ties (or dispersion forces) linked with geographic concen- as opposed to full and multilateral liberalization. Chapter 3, tration, such as higher cost for land or labor. The balance by Richard Baldwin, provides a comprehensive review of of agglomeration and dispersion forces is altered with the theoretical economic foundation of PTAs. PTAs gener- preferential liberalization, but in a complex way. In theory, ate a number of static effects that lead to a one-time reallo- lowering trade costs reduces dispersion forces but also cation of resources. The first effect is that exporters from a diminishes agglomeration forces. The alteration of the country that is enjoying lower (preferential) tariffs will balance has implications that are potentially important— benefit from the improved market access. A second, oppo- but are indeterminate in theory and case-specific—for site, effect is that countries left outside the preferential the location decisions of firms. Members of a PTA are trade area will lose out as their trade with PTA members is thus likely to benefit unequally from these complex displaced by trade between members. A third, and largely dynamic gains. uncertain, static effect is that PTAs may divert trade and enable producers from within the PTA to displace other, North-South PTAs more efficient producers that had been able to serve the market when all faced the same tariff regime. Diversion Issues related to the distributional effects of PTAs among concerns can be addressed through lump-sum transfers their members are of particular concern in the context of Overview 5 agreements between developed and developing countries. that help implement PTA disciplines and enhance their In chapter 4, Bernard Hoekman discusses the key develop- credibility. But deep and comprehensive PTAs between ment policy challenges associated with North-South PTAs developed and developing countries also bring new risks. and offers a number of rules of thumb and approaches for The proposed norms will not necessarily benefit the devel- making these PTAs development friendly. North-South oping countries if the provisions limit policy freedom in PTAs should strive to remove barriers to trade in the devel- inappropriate ways or lead to the allocation of resources to oped country for the products that the developing country activities that yield few immediate benefits or are complex produces; to lower trade barriers in the developing country and costly to implement. partner or partners that raise the prices of goods and serv- ices consumed by firms and households; to promote more Customs Unions general liberalization based on the most favored nation (MFN) principle, as this best serves global development A customs union (CU) is a trade agreement whereby coun- prospects; to support the adoption of complementary tries preferentially grant tariff-free market access to imports measures and actions that allow the potential benefits of from each other and agree to apply a common external tar- trade opportunities to be realized; and to create mecha- iff (CET) to imports from the rest of the world. It is one of nisms through which the private sector can be regularly the most advanced forms of PTA and generally requires informed of progress with implementation of the PTA and considerable coordination among members. In chapter 5, can provide feedback to authorities. Soamiely Andriamananjara observes that CUs have recently Achieving these objectives requires that changes be become less popular than simple free trade agreements made in the way North-South PTAs are usually designed. (FTAs). This trend reflects the nature of the current wave of If the goal is development, a first overarching objective regionalism, which is characterized by smaller cross- should be far-reaching liberalization. This can be achieved regional deals, flexibility, selectivity, and, most important, through a redesign of the approach to PTAs: high-income speed. Recent FTAs tend to be more pragmatic than their countries should liberalize in all sectors (not just “substan- precursors and to focus more on strategic commercial mar- tially all�), on a preferential basis, with liberal and simple ket access and less on geographic considerations or political rules of origin. Developing-country signatories should ambitions. By contrast, CUs are usually set up by a relatively reduce their tariffs and apply negotiated trade policy com- large number of geographically contiguous countries and mitments on an MFN (nondiscriminatory) basis. involve a certain loss of policy-making autonomy. A second necessary change is to build in significant pol- When a country joins a CU, it agrees to relinquish some icy flexibility, as well as stronger accountability mecha- of its national sovereignty over the formulation and imple- nisms. The goal should be to use PTAs as a tool for helping mentation of trade policy. The implication is that a CU developing-country governments pursue priority national member considers the loss of some autonomy to be more regulatory policy objectives, rather than have these objec- than offset by the economic benefits of securing access to a tives dictated by the trade partners. Economic cooperation larger and more harmonized regional market and of based on institutions that help identify good practices enhancing the depth and effectiveness of the ongoing would be a desirable and risk-free way forward. An regional integration process. Some regional groupings con- important corollary is that governments should be held sider the establishment of a CU a prerequisite or a neces- accountable for performance and outcomes. This requires sary step toward the future establishment of some deeper mechanisms that generate the necessary information—an form of economic integration, such as a common market. area in which the private sector has an important role Other groups regard a CU as a useful way of pooling coun- to play. tries’ market power, coordinating their trade policies, and A third plank of a development-focused approach is combining their negotiating powers to deal advantageously expanded development assistance. Such aid should not be with the rest of the world. limited to the issues that are covered by an agreement, and The establishment of a CU involves reaching consensus it should be aimed at improving trade-related regulation on three other important issues not relevant to PTAs. First, and its implementation, as well, including services inputs, members need to agree on a CET. The economic impact of both public (government services such as customs and the CU will be closely related to the degree of discrimina- trade facilitation) and private (transport, distribution, tion it entails, which depends on the selected tariff level. A finance, etc.). The extension of the PTA agenda to regulatory higher CET entails more trade diversion and greater net issues can be beneficial to developing countries, especially if welfare loss (although different economic agents will be accompanied by financial transfers and technical assistance affected differently). Whether CUs lead to higher external 6 Jean-Pierre Chauffour and Jean-Christophe Maur tariffs remains an open question. A number of arguments trading partners’ incentives for further liberalization. A key seem to suggest that CUs engender more protectionist question about the proliferation of PTAs, beyond their pressures than PTAs, but this is an empirical question to direct welfare impact for the countries signing them, there- which the existing literature has not been able to provide fore relates to their systemic effects on the world trading an unequivocal answer. system. Are PTAs—leaving aside the objective of mutually Second, members need to decide where and how to col- beneficial multilateral reduction of trade barriers—a force lect CET duties: at the initial port of entry into the CU, or for good or for bad? This question raises the more general at the final import destination. Collecting import duties at question of the interaction between PTAs and the multilat- the first port of entry could facilitate freer movement of eral trade system. Do PTAs influence multilateral liberal- goods within the CU and minimize intra-CU border con- ization, or is it the other way round? Or do they influence trols, but it requires the appropriate institutional capacity each other? to administer the revenues and, most important, a high A traditional view among economists is that PTAs form level of trust among members. Most existing CUs allocate a stumbling block in the path to freer multilateral trade. revenues according to the final destination principle, First, market access preferences granted in PTAs create which requires sophisticated administrative measures for incentives to resist multilateral liberalization, which would identifying the end destination of each shipment entering erode these preferences by reducing external MFN tariffs. the union. Second, market access preferences are used as bargaining Third, members have to decide which mechanism to use chips by preference-granting nations against nontrade for managing CET revenues. In some cases, the CU allo- concessions. Third, countries tend to choose to pursue lib- cates (a fraction of) these revenues to a joint fund to eralization with trade partners in areas that are less likely to finance regional development initiatives. Pooling customs affect politically sensitive sectors, while still offering liber- revenues presupposes a high level of coordinating capacity alization gains. Moving toward further liberalization may and a certain degree of trust among members and is more mean undertaking much more politically difficult liberal- likely to be sustainable when tariff revenues do not consti- ization that will affect the sensitive sectors. tute an important part of government revenue for individ- At the other end of the spectrum of economic analysis ual members. In other cases, customs revenues are treated are theories supporting the idea that PTAs could be as the property of individual members and are allocated building blocks for overall freer trade. Liberalization either according to the final destination or in line with an would beget liberalization in reciprocal negotiations, agreed sharing formula. generating a virtuous circle, as exporters are transformed The collection and allocation of customs revenues in a from bystanders in the tariff debate to antiprotectionists CU setting is clearly an area in which harmonization of motivated by the prospect of market access. The initial border management (e.g., customs procedures), coopera- reciprocal tariff cuts in PTAs may start a liberalization tion, and modernization, along with capacity building, juggernaut rolling. could be critical. So far, there is little systematic empirical evidence that regionalism is overwhelmingly bad for the multilateral trade system, as some had feared. Analyses tend to show PTAs and Multilateral Liberalization complementarity between PTA and multilateral tariff lib- In chapter 6, Richard Baldwin and Caroline Freund discuss eralization. Most empirical studies find trade creation the relative merits of preferential liberalization and the cir- effects in PTAs, with trade diversion as the exception. cumstances under which PTAs are more likely to serve as According to the natural trading partner hypothesis, coun- building blocks for greater multilateral trade liberalization tries tend to engage in PTAs with only those partners that than to pose stumbling blocks. On the positive side, PTAs can offer positive welfare gains. Among the selection crite- have offered a popular way for countries to liberalize their ria for such natural partners are geography and lower trade trade policies and gain market access. Countries are often costs, large economic size, and economic complementari- unwilling to liberalize on their own because they count on ties (e.g., factor endowments and difference in economics tariffs to raise tax revenues and, at times, protect domestic size). The complementarity effect is stronger in sectors in sectors. Reciprocal liberalization in the context of PTAs which trade bloc partners are more important suppliers, makes liberalization easier because the increased market which is precisely where trade discrimination would be access may counter the political-economy forces opposing more disrupting. liberalization. PTAs do, however, generate trade diversion There is, however, also evidence of PTAs’ acting as stum- and trade preference rents and these distortions alter the bling blocks. The United States and the European Union Overview 7 liberalized less during the Uruguay Round of multilateral public goods. In the case of regional PTAs, the treatment of trade negotiations in sectors in which preferences were these issues will determine the extent to which the expan- granted. sion of intrabloc markets can lead to scale economies and the rationalization of production and investment. In the case of bilateral PTAs among countries that are not in Agriculture the same region, there is less potential for economies of Turning to the key chapters of PTAs, a first consideration is scale and productivity gains, but PTAs could still offer a often to decide whether to include agriculture as part of the way to secure market access for agricultural exports and negotiations and PTA commitments. In chapter 7, Tim to benefit from (at least temporary) advantages over Josling discusses the challenges and opportunities associ- excluded competitors. ated with the incorporation of agriculture into PTAs. The challenges have to do with the additional competition Rules of Origin faced by domestic agriculture from regional partners that may be lower-cost producers or that, in the case of bilateral At the heart of the preferential regime of PTAs are the rules agreements with developed countries, may have signifi- of origin (ROOs) that determine the eligibility of products cantly better marketing and trade infrastructure. The pol- to receive preferential access. In chapter 8, Paul Brenton icy dilemma is whether to subject domestic agriculture to reviews the rationale and practice of rules of origin in further competition in the hope that the farm sector will PTAs. The justification for preferential rules of origin is respond by undertaking structural and technological to prevent trade deflection or simple transshipment, change and becoming competitive regionally and, eventu- whereby products from nonparticipating countries are ally, globally. Given adequate safeguards—in particular, to redirected through a free-trade partner to avoid the pay- avoid import surges—the better strategy would be to ment of customs duties. These rules, however, can be attempt to develop a competitive agricultural sector, but manipulated to achieve other objectives, such as protecting the political economy does not always allow it. domestic producers of intermediate goods. Restrictive As with other PTA chapters, the opportunities from rules of order that go beyond what is necessary to prevent incorporating agriculture in PTAs go far beyond the gains trade deflection could raise the economic costs of supplying from trade creation. They include collective action on the markets of preferential partners and the administrative research, market development, and health and safety regu- costs of proving conformity with the rules. These costs will lations, as well as cooperation on trade policy and on constrain market access relative to what is promised on approaches to development agencies and donors. Where paper in the trade agreement. The rules of origin are there- these arrangements have worked well, regional markets fore a key element determining the magnitude and distri- have been established, health and safety regulations have bution of the economic benefits that accrue from PTAs. been harmonized or made more compatible, and trade has There is no simple and standard set of rules of order developed through investment, as firms have realized the that can be identified as performing the task of preventing potential of closer market integration. By contrast, when trade deflection. Three main criteria are used to establish PTAs opt to exclude some or all agricultural sectors from whether a product imported from a partner that contains the pressures and opportunities that come with freer inputs from other countries has undergone sufficient pro- regional trade, agricultural markets often remain confined cessing or a substantial transformation: (a) a change of tar- within national borders or focused on traditional trade iff classification; (b) a minimum amount of domestic value exchanges with overseas partners. added; or (c) use of a specific manufacturing process. No Overall, the dearth of empirical studies on economic one method is dominant; each has its advantages and dis- integration in agricultural markets makes for a lack of clear advantages, and different rules of origin can lead to differ- guidance for policy makers. To help close that gap, the ent determinations of origin. Other features of the rules of chapter provides a brief review of experience with agricul- origin that can influence whether origin is conferred on a tural provisions in selected PTAs. The issues common to all product include cumulation, which allows producers to PTAs, as they contemplate the agricultural component of import materials from a specific country or regional group the agreement, include the way in which tariffs are cut, the of countries without undermining the origin of the prod- use of tariff-rate quotas to open up markets on a progres- uct; tolerance, or de minimis, rules, which allow a certain sive basis, the employment of safeguards to help domestic percentage of nonoriginating materials to be used without sectors cope with import surges, the treatment of subsidies affecting the origin of the final product; and the absorption (both domestic and on exports), and the provision of principle, which provides that parts or materials that have 8 Jean-Pierre Chauffour and Jean-Christophe Maur acquired originating status by satisfying the relevant rules Furthermore, provisions differ significantly for the same of origin for the particular product can be treated as country across different PTAs, and countries do not com- being of domestic origin in any further processing and monly incorporate the same trade remedy provisions in all transformation. their PTAs. In some cases, trade remedy provisions in PTAs Restrictive rules of origin constrain international spe- make protection easier, but in most cases, the additional cialization and discriminate against small, low-income rules in PTAs tend to make protection harder to impose. countries where the possibilities for local sourcing are lim- Any country considering entering into a PTA should there- ited. Simple, consistent, and predictable rules of origin are fore consider carefully what provisions potential partners more likely to foster the growth of trade and development. have included in prior agreements. Rules of origin that vary across products and agreements Another finding of the chapter is that PTAs are less add considerably to the complexity and costs of participat- likely to alter existing countervailing duty provisions than ing in and administering trade agreements. The burden of they are either antidumping or global safeguard rules. This such costs is particularly heavy for small and medium-size is partly because few PTAs have created common policies firms and for firms in low-income countries. Complex on subsidies and state aid. Without such rules, and given systems of rules of origin add to the burdens of customs the global nature of subsidy distortions, there appears to services and may compromise progress on trade facilita- be little motivation for PTAs to limit the application of tion. Specification of generally applicable rules of origin, countervailing duties against members. PTA provisions with a limited number of clearly defined and justified regarding antidumping, countervailing duties, and global exceptions, is appropriate if the objective is to stimulate safeguards generally make protection more difficult to integration and to minimize the burdens on firms and cus- impose than do the existing WTO rules; that is, these pro- toms services in complying with and administering the visions are WTO+. They work against protection through a rules. Producers should be given flexibility to meet origin combination of additional specific rules that relax the rules by, for example, specifying in the rules that the prod- threshold and duration of application of measures and ucts may satisfy either a change of tariff requirement or a through the creation of institutions that help to defuse dis- value added criterion. putes. By contrast, PTA provisions on bilateral safeguard Preferences granted by member countries of the Organ- actions (e.g., transition safeguards and special industry isation for Economic Co-operation and Development safeguards) offer new avenues for PTA members to restrict (OECD) would be more effective in stimulating exports intra-PTA trade. from developing countries if they were governed by less- Overall, the evidence shows that PTAs reduce the inci- restrictive rules of origin; ideally, specific rules of origin dence of intra-PTA antidumping and safeguard disputes. should be designed, and producers in developing countries At the same time, there appears to be an increase in actions should be able to gain preferential access to all developed- against non-PTA members. So, although the liberalization country markets if their product satisfies a single origin effects of PTA tend to be reinforced by global safeguard test. Restrictive rules of origin should not be used as tools and antidumping rules in PTAs, the trade diversion effects for achieving economic development objectives—they are can also potentially be reinforced. In addition to the dis- likely to be counterproductive. The potential benefits of crimination introduced by preferential tariffs, PTAs can trade agreements among developing countries can be sub- lead to increased discrimination against nonmembers stantially undermined if those agreements contain restric- through more frequent trade remedy actions against them. tive rules of origin. Overall, the chapter calls for vigilance about the impact of trade remedy provisions in PTAs. The sheer number of PTAs with trade remedy rules is heightening the promi- Trade Remedies nence of the issue. The diversity of PTA types, ranging from goods-only agreements to customs unions, means that the political and Standards economic demands for trade remedy provisions vary greatly across PTAs. In chapter 9, Thomas Prusa shows that Provisions relating to the management of standards for the proliferation and diversity of PTAs have produced a protection of human, animal, or plant life or health are complicated pattern in the use and inclusion of trade rem- now a common feature of most PTAs. In chapter 10, Jean- edy provisions across PTAs that defies simple characteriza- Christophe Maur and Ben Shepherd discuss the unique tion. Whereas some PTAs contain long discussions of trade role that PTAs can play in reducing standards barriers to remedy rules, others do not even mention trade remedies. trade through the recognition of equivalence of rules and Overview 9 procedures. They observe that standards provisions in and consultations on legal and administrative matters PTAs are likely to have welfare-enhancing effects on partic- through regular dialogue. ipating members. Although standards are classified as one of many nontariff barriers—the relative importance of Standards, in Practice which is growing with the decline in tariff duties—an important economic distinction is that standards serve a Andrew Stoler, in chapter 11, discusses specific best-practice different policy objective than simple discrimination against provisions in PTAs concerning the treatment of technical foreign goods. Standards represent a quasi-regulatory barriers to trade (TBTs) and sanitary and phytosanitary means of pursuing important public policy objectives such (SPS) measures. He recommends that the parties to a PTA as environmental protection, consumer safety, food quality, aim at using international standards whenever possible and compatibility between different types of apparatus. because doing so guarantees a high level of protection in The policy objective should therefore not be to eliminate the integrated market and makes it easier for third parties standards but to make them more efficient and cost to trade into that market. If the parties to the PTA decide to effective. pursue harmonization of their standards and conformity Harmonization to international standards is an effective assessment procedures, they should accept that it might be way of reducing the duplication of costs of complying with necessary to limit harmonization to essential health and different sets of standards. Standards harmonization may safety standards and to rely on mutual recognition and also facilitate market access and enable countries with no equivalence techniques for other areas. Where one partner or inefficient standard policies to put better practices in is less developed than the other, the PTA will have to place. But it does mean compromising governments’ ability include technical assistance and capacity-building measures to set national standards that may better fit the needs to assist the institutions and exporters of the developing- of local industries and consumers. country partner in adjusting to harmonization. There are, broadly, two models for dealing with stan- If technical regulations and conformity assessment dards measures in PTAs. Where the European Union is procedures cannot be harmonized, it is important for the involved, the agreement often expects the EU partner purposes of the PTA that the parties work to eliminate country to harmonize its national standards and conform- requirements for duplicate or multiple measures or tests ity assessment procedures with those of the EU. PTAs in for the same product. This is particularly important for the Asia-Pacific region and those in which the United small and medium-size enterprises that cannot afford the States is a partner typically address problems resulting high cost of meeting differing regulations and testing from different national standards and conformity proce- regimes. Mutual recognition agreements (MRAs) are dures through a preference for international standards important tools in this respect. or the use of mutual recognition mechanisms. Both Transparency is important for business and consumers approaches can be successful in reducing the negative in this area of international trade. PTA partners should impact of different standards and conformity assessment consider incorporating WTO+ notification obligations procedures, but there is a risk that they can introduce into and a commitment not to implement any technical regu- global markets de facto discrimination, particularly against lation or SPS measure until it has been published and developing countries, because achieving conformity in comments by the PTA partner have been taken into account. technical standards requires capacity and resources. It is recommended that the PTA be drafted as a “living Standards indeed have the potential to discriminate agreement� with a commitment to a work plan or priori- against nonmember countries when mutual recognition tization of problem resolution through harmonization, agreements are not open to third countries or when har- mutual recognition, equivalence measures, and other monization is not carried out on the basis of interna- policy tools that enable elimination or mitigation of trade- tionally agreed standards. As in the case of contingent related problems over time. PTA provisions on TBTs and protection measures and restrictive rules of origin, stan- SPS measures should be legally binding, through a judi- dards may end up raising external barriers against coun- cious combination of “soft� and “hard� law. The provisions tries outside the PTA, thus eroding some of the benefits of should provide a pathway that permits integration to multilateral liberalization. Good practices regarding stan- evolve and deepen over time by allowing the gradual reso- dards in PTAs usually promote institutional arrangements lution of TBT and SPS issues in the bilateral relationship. to supervise the effective implementation of standards Such a pathway should be considered an integral part of provisions. Important objectives of such arrangements are, any PTA that aims to deal effectively with standards, certifi- most frequently, transparency, the diffusion of expertise, cation, and conformity assessment problems. Eventual 10 Jean-Pierre Chauffour and Jean-Christophe Maur recourse to the PTA dispute settlement provisions should listing, most notably in the area of transparency. Studies be an option, along with recourse to the WTO Dispute devoted to the practice of preferential market opening Settlement Understanding (DSU). suggest that North-South PTAs based on a negative-list Finally, PTA parties should agree to an overall commit- approach tend to achieve the deepest, WTO+ liberalization. ment always to apply technical regulations and conformity The chapter also shows that gains from PTAs are likely to assessment procedures on a national treatment basis and to be significant in services areas where there is scope for reap- allow third parties whose technical regulations and con- ing large economies of scale. In principle, these gains can formity assessment procedures can be demonstrated as also be realized through MFN liberalization, but in practice being equivalent to the level agreed to by the PTA partners the full integration of markets may require a deeper conver- to benefit from the arrangements between the partners. A gence of regulatory regimes. Regulatory cooperation may be commitment to open regionalism would help ensure that more desirable, and probably more feasible, within a subset the PTAs support the multilateral system. of countries than if pursued on a global scale. Regional or international harmonization or standardization can be an important and cost-effective way of improving national stan- Services dards. Yet the best partners for regulatory cooperation are In recent years, increasing numbers of PTAs have sought to likely to be those with the soundest regulatory frameworks, include provisions on liberalizing cross-border trade in serv- and they may not always be found within regional compacts. ices, investments in goods- and services-producing activities, There are gains from regulatory cooperation, but also and the temporary movement of business people. In chapter costs. The former will dominate where national regulation 12, Aaditya Mattoo and Pierre Sauvé show that liberalization can be improved and where regulatory convergence or har- of trade in preferential services differs fundamentally from monization can be carried out, taking into account local liberalization of trade in goods in that the effect of many circumstances. The costs are likely to be smallest when for- restrictive measures in services trade (given their regulatory eign regulatory preferences are similar and regulatory insti- nature) is to increase the costs of operation faced by foreign tutions are broadly compatible. providers without necessarily generating equivalent domes- Unless there are liberal rules of origin for investment that tic rents. There is therefore little or no cost to granting pref- confer the full benefits of an integration scheme on third- erential access for services trade because there is little or no country investors, the establishment of preferences may revenue to lose. In such circumstances, preferential liberal- result in entry by inferior suppliers. Because the most effi- ization will necessarily be welfare enhancing, but countries cient suppliers may also generate the greatest positive exter- outside the preferential arrangement may lose. nalities, the downside risks of preferential liberalization may PTAs covering services tend to follow two broad be greater, especially in crucial infrastructural services. This approaches with respect to liberalization of services trade is particularly the case in services sectors with high location- and investment. Some PTAs tend to replicate the use, found specific sunk costs that need to be close to consumers. Pref- in the General Agreement on Trade in Services (GATS), of a erential liberalization may then exert more durable effects positive-list or hybrid approach to market opening whereby on the nature of competition than in the case of trade in agreements list sectors, subsectos, and modes of supply in goods. For instance, concluding an agreement that allows which governments are willing to make binding liberaliza- second-best providers to obtain a first-mover advantage may tion commitments; others pursue a negative-list approach, imply that a country could be stuck with such providers, in which exceptions to liberalization are listed. More than even if it subsequently liberalizes on an MFN basis. half of all the PTAs concluded to date that feature services Finally, PTAs have generally made little progress in tack- provisions use a negative-list approach. Such agreements ling the interface between domestic regulation and trade in are more prevalent in the Western Hemisphere, reflecting services (e.g., subsidies). Despite the greater initial similar- the influence of the North American Free Trade Agreement ities in approaches to regulation and greater cross-border (NAFTA), and in agreements conducted along North-South contact between regulators that geographic proximity can lines (with the exception of the EU and EFTA agreements). afford, progress in the area of domestic regulation has been Although both approaches can in theory generate broadly slow and generally disappointing, even at the PTA level. equivalent outcomes with respect to liberalization, as a prac- tical matter, a negative-list approach can be more effective Labor Mobility in locking in the regulatory status quo. In addition, the process of “getting there� tends to differ; a number of good The temporary movement of natural persons is a key mode governance–enhancing features are associated with negative of service delivery across borders. In chapter 13, Sherry Overview 11 Stephenson and Gary Hufbauer show that, notwithstand- Investment ing the strong sensitivity that surrounds the movement As trade and investment become more and more inter- of people and the confusion about the fine line between twined in the context of international supply chains and temporary movement of workers and permanent migra- firms’ strategies of vertical specialization, countries are tion, PTAs are increasingly being used as vehicles for dealing increasingly incorporating investment provisions into with labor mobility issues. Several PTAs between devel- PTAs instead of negotiating bilateral investment treaties oped and developing countries have included new cate- (BITs). In chapter 14, Sébastien Miroudot presents the gories of labor, such as technicians, nurses, and health care salient features of investment provisions in PTAs. By com- workers, in their trade agreements or have offered an bining trade liberalization provisions with investment lib- expanded number of categories for service suppliers, com- eralization and protection, PTAs emphasize the comple- bined with expanded quotas. Other PTAs have eliminated mentary relationship between trade and foreign direct the quotas altogether for certain categories of services sup- investment (FDI). In contrast to BITs, investment provi- pliers (usually, professionals). To date, nearly all PTAs that sions in PTAs ensure that disciplines are not limited to cover services focus either exclusively or primarily on investment in the postestablishment phase but also deal professional services suppliers. with the preestablishment stage (i.e., market access, Members of South-South regional integration group- national treatment, and most favored treatment), thus pro- ings of geographically contiguous countries in Africa, Asia, viding economic incentives for investors, in the form of and Latin America and the Caribbean often contemplate market access, in addition to the legal incentives (the pro- completely liberalized labor markets as part of their agree- tection of investment, once established). Given liberal rules ments. Progress is slowly being made toward the imple- of origin, investment provisions in PTAs do not seem to mentation of these commitments for all categories of introduce severe distortions among investors; and develop- workers, both for temporary movement and for perma- ing countries have usually been able to maintain some of nent settlement. The chapter confirms the intuition that the restrictions needed for legitimate policy considerations. the countries that have gone furthest in opening their mar- Investment disciplines hold an important place in PTAs kets to temporary labor movement under PTAs are the that they have not found in the multilateral negotiations. ones that are experiencing a relative shortage of labor sup- This is one of the four “Singapore issues� that were sup- ply at home (e.g., Canada, Japan, and New Zealand). Con- posed to become the new additions to the WTO in the versely, the countries that have been most reluctant to open Doha negotiations. (The other three—trade facilitation, their market to temporary labor movement under PTAs government procurement, and competition policy—are are those that are experiencing a strong wave of migration discussed in the following chapters). There are two main pressure—notably, from Mexico, in the case of the United models of investment provisions in PTAs, one inspired by States, and from new EU members in Eastern Europe, for NAFTA, which places investment in goods and services the EU. industries in the same chapter, and one that follows the Looking forward, the chapter observes that developed GATS model for part of the provisions on investment in countries should proactively search for labor market niches services. Despite important differences, these two types of in which additional temporary workers can become valued agreement offer the same degree of protection for invest- members of the workforce and the community. In that ment and are equally liberalizing of investment, even respect, there has been some progress within a new genera- if, empirically, NAFTA-inspired agreements are more tion of bilateral labor agreements (BLAs) and temporary ambitious in their scope and sectoral coverage. For devel- worker programs (TWPs). In the case of workers with oping countries, North-South PTAs with substantive lower skill levels, the best vehicle for promoting greater investment provisions are found to positively affect FDI labor mobility may not be a formal PTA but, rather, the flows. They can also offer an opportunity to address con- more flexible instrument of a TWP that specifies the length cerns about dispute settlement for countries that are not of stay for a certain number of workers in a specific eco- ready to face commercial arbitration. nomic activity. TWPs have the advantage that both govern- ments are better able to enforce the clauses, and they can be more easily adapted to the cycle of economic fluctuations. Trade Facilitation Canvassing the extent of TWPs in the world economy today is challenging, as no one organization is responsible Trade facilitation is a relatively recent domain of interna- for maintaining an inventory of these agreements, but their tional trade cooperation, and countries are still very much numbers appear to be increasing. in the process of learning how to design the best approach. 12 Jean-Pierre Chauffour and Jean-Christophe Maur In chapter 15, Jean-Christophe Maur notes that the cur- countries deal with competition problems, including those rent approaches to trade facilitation in PTAs are mainly emanating from international trade, such as import distri- extensions of customs cooperation provisions, with some bution monopolies and cartels, overseas export cartels, and variations depending on the signatory countries. Trade the abuses of dominant firms from abroad. They observe facilitation provisions are also closely related to the pro- that international trade presents complex competition pol- posals in the Doha Round of trade negotiations concerning icy issues because an abuse of market power can occur trade facilitation involving developed countries. Good unevenly across several markets and jurisdictions, with practice in this area seems to be to strive for living agree- consequences for the jurisdictional limits of a national ments, whereby institutional arrangements are set up for authority. promoting exchange and joint initiatives with a view to Effectively enforced regional competition provisions possibly amending agreements through new provisions on may serve to lock in reforms that are politically difficult to trade facilitation. In addition, PTA signatories should sustain because of the influence of strong domestic lobby- extend their trade facilitation commitments beyond the ing groups that do not immediately benefit from competi- sole remit of customs administration to include other bor- tion law. PTAs can also pioneer or test-run provisions and der agencies (largely ignored in existing agreements) and so facilitate their negotiation at a multilateral level at a later services and standards, which are often treated separately. date. Finally, regional competition regimes offer a demon- Indeed, recent PTAs that incorporate more advanced trade stration effect of the positive gains to be had from effective facilitation provisions demonstrate the increased interest national cooperation to underpin and improve the market in using PTAs to reduce transaction costs. liberalization process. The trade facilitation agenda in PTAs remains largely In the case of North-South PTAs, greater development driven by the most developed partners, which tend to set benefits can be reaped if the more developed party offers the agenda and to propose terms of agreement (e.g., appropriate technical assistance and capacity building. ambitious reforms such as risk management) that may For developing countries with little experience in imple- not necessarily reflect the needs and capacity of their less- menting competition rules, the PTA provisions should developed partners. Reflecting this relative lack of consid- initially be limited to the exchange of information, tech- eration, there are few examples of special and differential nical assistance, and capacity-building support, as it may treatment by trading partners, or of clear commitments on be more beneficial at first to focus on establishing a cul- provision of technical assistance. ture that values competition at the national or subre- More generally, PTAs seem to fail to capitalize on gional level. Subsequent negotiations could expand the opportunities to provide value added: that is, domains of agreement. The use of soft law could be beneficial if gov- potential regional cooperation are not always given prior- ernments are uncertain of the underlying technical issues ity. For instance, PTAs could have a much more active role and the consequences of the provisions. Other nonjudi- in implementation through mutual recognition and assis- cial mechanisms, such as voluntary peer review and con- tance, the eventual sharing of resources, and joint efforts to sultations, may also be appropriate. improve the trade supply chain. Trade facilitation provi- The economic and human resources necessary to imple- sions that focus on reform on the ground, implementation, ment even a minimal regional decentralized competition and monitoring of measurable objectives, as in Asia-Pacific regime arrangement are significant for both developed and Economic Cooperation (APEC), seem to deliver more developing countries. Nevertheless, the emerging evidence effective and successful reforms. Such agreements tend to shows that the economic and welfare costs associated with be pragmatic, flexible, and country specific and are gener- cross-border anticompetitive practices are higher. Short- ally well suited to the type of reforms required in the con- term political costs should be weighed against the under- text of regional trade facilitation. standing that the long-term and sustainable benefits of a strongly enforced regional competition regime will almost always outweigh its costs. Competition Policy Decentralized agreements that only require the exis- The lack of a multilateral competition framework has coin- tence of a local competition law and authority to apply the cided with a surge in PTAs that contain competition provi- law, such as NAFTA, are not as economically demanding as sions intended to address market failures that national a regime that establishes a fully centralized law with a sup- competition laws cannot remedy. In chapter 16, Kamala porting regional authority, such as the EU or the Common Dawar and Peter Holmes discuss how well-designed Market for Eastern and Southern Africa (COMESA). Com- regional competition arrangements can help developing petition provisions in customs union agreements are, in Overview 13 general, more specific and demand higher commitments. A government that has improved its national procure- The economic burdens of implementation can be offset if ment procedures in compliance with PTA provisions may the parties are able to exchange information effectively and not find it worthwhile to maintain separate procurement avoid duplications and conflicting decisions. For instance, processes for bidders from PTA signatories and nonsigna- the competition regime in the Caribbean Community tories. Government procurement provisions therefore (CARICOM) points to potential economies of scale from offer the prospect of being implemented across-the- regional cooperation; the provisions allow for resource board and more formally, on an MFN basis. Third-party pooling among neighboring countries when national MFN rules are an important way of furthering nondis- capacity is not adequate to implement and enforce the criminatory liberalization over time and of diffusing regional framework. good practices. Although improvements to procurement regimes required by PTAs, such as increased transparency, may Government Procurement be costly, the expense could be mitigated through techni- In the same vein, according to Kamala Dawar and Simon cal assistance and capacity-building programs negoti- Evenett, in chapter 17, government procurement provi- ated as part of the PTA. Flexibilities are available to sions may have received greater attention in PTAs because negotiators, including exclusion of certain entities from of the prospects they offer for reforming national procure- the agreement, the threshold set in order for procure- ment systems, The authors note that all the PTAs that ment to be covered by the agreement, the exclusion of include government procurement provisions promote a sectors such as defense and financial services, and the use liberal agenda, acknowledging the benefits of transparent of offsets or set-asides to accommodate domestic poli- nondiscriminatory and competitive procurement markets. cies such as the promotion of indigenous communities Agreements differ in the scope and strength of their com- or small businesses. mitment to progressive liberalization of procurement mar- kets, but the provisions typically cover a dozen aspects of Intellectual Property procurement policies, including nondiscrimination princi- ples, procedural rules, dispute settlement mechanisms, As Carsten Fink shows in chapter 18, rules for the protec- commitments to cooperation and further negotiations, tion of intellectual property rights (IPRs) have become a state entities and sectors covered, and special exceptions common, albeit controversial, feature of PTAs, especially know as offsets. those involving developed partners. Over the past decade For many countries, government procurement outlays or so, the United States has negotiated numerous PTAs are a sensitive matter, making for reluctance to take on that contain ambitious IPR chapters. These chapters strong and legally binding international obligations. The introduce standards of protection for all types of IPRs that desire for value for money from public purchasing has exceed those of the WTO Agreement on Trade-Related often been tempered by other objectives, such as support Aspects of Intellectual Property Rights (TRIPS)—that is, for specific domestic industries and interests. Accordingly, they are TRIPS+. The most prominent, and sometimes in many PTAs, particular attention has been paid to elimi- most controversial, standards include patent term exten- nating the more transparent forms of discrimination, such sion, patenting of life forms, patent-registration linkage as price preferences. This may have had the unintended for pharmaceutical products, exclusive rights to test consequence of driving discrimination into nontransparent data, prolonged copyright protection, rules on technolog- forms such as contract thresholds, limitation on number of ical protection measures and the liability of Internet serv- bidders, exclusion of sensitive sectors, and the like. Yet the ice providers, and more stringent requirements for the rules governing procurement provisions in PTAs appear less enforcement of IPRs. These rules seek to deepen and discriminatory than the provisions governing market update preexisting multilateral IPR rules, as embedded in access. One reason is that government is only one among the TRIPS Agreement. In May 2007 the United States numerous national buyers, which limits the impact on mar- adopted a new trade framework, the Bipartisan Agree- ket outcome of its decisions to award specific preferences. ment, to roll back some of the TRIPS+ provisions as they Another focus of attention in PTAs has been improvement relate to pharmaceutical products. Although the Biparti- of the transparency of those public institutions in signatory san Agreement covers only three PTAs, it marks an impor- countries that are responsible for state purchasing. tant shift in U.S. trade policy toward greater sensitivity Transparency provisions in PTAs tend to generate bene- with respect to the implications of global IPR rules for fits for nonmembers of the PTA, as well as members. public health concerns. 14 Jean-Pierre Chauffour and Jean-Christophe Maur Until recently, the agreements concluded by the EU dispute in this area so far, the rapid spread of PTAs and were less far-reaching than those of the United States. In BITs may well lead to the initiation of arbitration claims the past, the main TRIPS+ element of the EU’s PTAs took in the future. the form of separate agreements on wines and spirits that included lists of geographic names to which signatories Environment had to apply rigorous geographical indication protection. In 2006 the EU embarked on a new set of negotiations, Practically all countries recognize the critical importance and it is demanding the inclusion of more comprehensive of environment, but many do not believe that dealing with IPR chapters in these new agreements. The EU’s eco- environment issues under trade agreements is necessarily nomic partnership agreement (EPA) with the Caribbean the best approach. In chapter 19, Anuradha R. V. does not Forum of African, Caribbean, and Pacific (ACP) States seek to build a rationale for the trade-environment linkage or (CARIFORUM), gives a sense of the new EU approach, to discuss whether environmental issues should be addressed including as it does TRIPS+ provisions not only on geo- in trade agreements or in stand-alone environmental agree- graphical indications but also in other areas, notably IPR ments. Instead, she focuses on the main contours of the trade enforcement. Future EU agreements with more developed and environment debate at the multilateral level and on how trading partners may well be more ambitious than the environmental provisions are increasingly being incorpo- agreement with CARIFORUM. rated into PTAs, especially by major developed economies. TRIPS+ standards of protection are found in several The United States, the EU, Canada, and New Zealand PTAs that do not involve the United States or the EU. are the principal proponents of environmental provisions Notably, several trade agreements negotiated by EFTA pro- in PTAs. On the basis of a review of existing agreements, it vide for patent term extension and exclusivity of pharma- would appear that the key issues in the negotiation of envi- ceutical test data. Although agreements among developing ronmental provisions in a PTA include an understanding countries usually do not go beyond TRIPS, several of the of (a) the nature of legal obligations emerging from provi- PTAs signed by Chile and Mexico include lists of geograph- sions relating to the environment under a PTA; (b) the ical indications that benefit from protection in the signa- potential economic costs of specific environmental require- tory countries. ments; (c) areas where technical assistance and capacity The adoption of TRIPS+ standards is often an impor- building would be necessary to ensure compliance with tant element in the overall package of quid pro quos neces- environmental obligations; (d) the nature and extent of the sary to conclude a preferential trade deal, reflecting the financial assistance required; and (e) the nature of dispute importance of IPRs as a market access concern for devel- settlement and enforcement mechanisms. oped countries. Yet the logic of negotiating new IPR stan- Even before negotiations, pre-PTA impact assessments dards differs from the traditional logic that economists could provide important benchmarks for assessing the have applied to the reciprocal dismantling of tariff barriers, scope of the trade-environment linkage. For instance, the in several ways: EU’s use of sustainable impact assessments (SIAs) to appraise the environmental implications of the PTA for • Optimizing the degree of IPR protection entails striking both parties offers a basis for addressing the linkages a proper balance between the interests of IPR owners between environmental provisions and sustainable devel- and the public at large. There is no guarantee that ever- opment in the trading partners. Technical and financial higher standards of IPR protection will necessarily assistance and capacity-building support by the party that improve economic welfare. has more experience with such assessments should be a • Because of the MFN obligation under TRIPS, PTA par- necessary aspect of the PTA process. ties usually cannot implement TRIPS+ standards in a preferential way, which affects the bargaining incen- Labor Rights tives of countries that are engaged in multiple PTA negotiations. Provisions on labor rights in PTAs are becoming increas- • IPRs have been included in the definition of investment ingly common, particularly in PTAs between large, power- in many PTA investment chapters and in BITs. Such a ful, developed countries and smaller, poorer developing broad definition of investment raises the possibility that countries. In chapter 20, Kimberly Ann Elliott observes private rights holders may directly challenge government that demands to include labor rights in PTAs began with measures affecting IPRs under the terms of an invest- the United States and have since been taken up, albeit in ment accord. Although there has been no investment somewhat different forms, by Canada, Chile, and, more Overview 15 recently, the EU. The definition of labor standards in PTA Although growing numbers of countries include or negotiations usually refers to the International Labour accept human rights provisions, the trend has limits. Many Organization (ILO) Declaration on Fundamental Princi- policy makers in middle-income countries and in the ples and Rights at Work. That document lists four core developing world are reluctant to use trade policies to labor standards as deserving of universal application: free- change the behavior of other countries. As countries grow dom of association, and “effective recognition� of the right richer and more influential, however, these policy makers to collective bargaining; abolition of forced labor; effective may become more willing to accept or to demand human abolition of child labor; and elimination of discrimination rights provisions. Nevertheless, some industrial countries, in employment. Notwithstanding this common definition, such as Australia, that themselves have strong human the negotiating stances of the main demandeurs differ rights records have refused to include human rights widely, especially on questions of implementation and requirements in trade agreements. Other nations, such as enforcement. Overall, it is difficult to assess the implica- China, have accepted such provisions. tions of these agreements for developing countries because The human rights embedded in these PTAs exhibit a great sustained attention to implementation is rare. PTA negoti- variety; they include civil and political rights to privacy, polit- ations with the United States have led to changes in existing ical participation, due process, and access to information, but labor laws in several cases in which those laws were deemed also economic, social, and cultural rights, rights of indige- inadequate, but monitoring and enforcement after agree- nous peoples, and access to affordable medicines. Although ments are signed are sporadic, at best. countries with interests in promoting human rights may Four key findings emerge from the review of labor pro- prod their trade partners to change their laws or devote visions in PTAs: (a) enforceable provisions for labor stan- greater resources to human rights, little is known about the dards are a condition for negotiating a PTA with the United actual effect of provisions in PTAs on human rights condi- States and are likely to remain so; (b) labor provisions in tions. These provisions, especially the more positive eco- PTAs not involving the United States or Canada are nomic, social, and cultural rights, are probably expensive unusual and are almost always hortatory, rather than for developing countries to implement—just like IPRs. legally binding or enforceable; (c) the EU appears to Developing countries must devote scarce resources to human be joining the trend, if the model used in its EPA with rights, perhaps before they have the national income or will CARIFORUM is replicated in other similar agreements, to do so. If human rights provisions are carefully designed, but the language on labor, although nominally binding, they can focus on improving governance (the supply side), includes no sanctions for noncompliance; and (d) even in as well as on empowering people to demand their rights. U.S. PTAs—which have the strongest language, on paper— If these rights provisions are to be workable and lasting, enforcement is rare, sanctions have never been applied, and policy makers will need to understand their effects on trade financial, technical, and capacity-building assistance to and governance. Policy makers, scholars, and activists improve implementation of labor standards is uncommon. should use human rights impact assessments, as well as widely accepted datasets, to gain greater understanding of how to make the match between trade and human rights Human Rights effective and enduring. The growing number and scope of PTAs containing human rights provisions reflects a new reality: policy mak- Dispute Settlement ers perceive that economic integration will not be success- ful without a stronger focus on improving governance In theory, the parties to a PTA are the masters of their own among trade partners. In chapter 21, Susan Ariel Aaronson treaty and could design an original dispute settlement notes that the United States, Canada, the EU, and the mem- mechanism from the ground up, or have no dispute settle- bers of EFTA are the main demandeurs of human rights ment mechanism at all. In chapter 22, Amelia Porges language in PTAs. The EU and EFTA focus on human observes that, in practice, almost all PTAs employ one of rights under the Universal Declaration on Human Rights three general types of dispute settlement mechanism: (UDHR), but they rely on aspirational language and on diplomatic settlement, use of standing tribunals, or a dialogue. Canada and the United States focus on specific WTO-type system in which ad hoc panels determine human rights, embed these provisions in the body of the whether disputed measures are in conformity with the trade agreement, and often make them binding. These PTA’s obligations. approaches have become more similar over time, but they Dispute settlement is useful in several ways. A PTA must remain distinct. provide an orderly way for its members to settle disputes 16 Jean-Pierre Chauffour and Jean-Christophe Maur and move on, or the disputes will poison bilateral relations, will be helpful, but it will require an up-front investment. reduce the benefits from the PTA, and perhaps even lead to The details of panel procedures can be left to be agreed later, the demise of the agreement. Dispute settlement is also although negotiators can set key parameters in advance. essential to ensure that the promises set forth in a PTA are Various approaches exist for dividing the expenses of dis- kept. By participating in a PTA with strong dispute settle- putes, handling translation and documentation issues, and ment provisions, a government signals its level of commit- regulating presentation of evidence and arguments. ment to private and public interests at home and abroad. PTAs with dispute settlement experience have sought Even if no disputes are anticipated, enforcement provisions (like the WTO negotiators) some way to ensure sound and in a PTA reinforce the governments’ precommitments, consistent panel decisions: almost all ad hoc panel systems make their promises more credible, and signal that the PTA allow the parties to comment on panel reports in draft is a sound platform for investment that will create jobs and form, and two PTAs have committed to appellate mecha- economic growth. Solid dispute settlement is even more nisms. Then there is the question of participation in the important in North-South (or South-South) PTAs with process. Who are the decision makers? Every PTA has a asymmetrical power relations. Recently concluded PTAs in process for selecting members of panels or standing tri- Latin America, Europe, and Asia demonstrate to a striking bunals and ensuring their impartiality. Who (other than extent that as PTA obligations deepen, become more the governments) may provide input into this process? complex, and provide more value, PTA partners seek Every government that engages in dispute settlement relies more certainty than purely diplomatic dispute settlement on its private sector to identify problems, to provide rele- can provide. vant factual information, and to identify commercial pri- The WTO-type ad hoc panel model is an often-used orities. Will the PTA permit panels to consider input from option, but some PTAs employ a standing tribunal instead. civil society in general? Must submissions and hearings be A tribunal may involve more fixed investment in infrastruc- open to the public? ture than ad hoc panels, but its permanence may make it Finally, it is important to have some mechanism for available to take on more functions for the integration determining whether government measures breach PTA process. Thinking ahead toward coexistence and even mul- obligations—but formal dispute settlement is not always tilateralization of PTA networks, however, it may be easier the quickest way to resolve a measure that is causing a com- to mesh ad hoc panel systems than court-based systems. mercial problem. The committees and other institutions In doing so, one should be aware that even an ad hoc created by a PTA provide a practical setting for addressing system involves many procedural choices: and resolving such issues. Some PTAs go further and build in a role for mediation, conciliation, or other rule-agnostic • Should the PTA limit panels to determining whether practical settlement methods. The benefits of such media- a PTA government has violated PTA law? Some PTAs tion are sure to exceed its costs. follow the General Agreement on Tariffs and Trade The final, and unavoidable, problem is compliance. (GATT) by making provision for disputes about gov- Unless a PTA’s dispute settlement instruments can (like the ernment measures that do not violate the rules but still European Court of Justice) impose fines on members for frustrate reasonably expected PTA benefits. noncompliance, PTAs’ leverage to obtain compliance con- • Should the PTA provide that all of its obligations are sists of authorizing withdrawal of PTA benefits or exercis- subject to dispute settlement? Some do, whereas others ing moral suasion. To the extent that a PTA’s preference exclude particular areas (for instance, soft law or com- margin is low relative to MFN benefits in the WTO, a PTA petition law). member may have much more leverage litigating in the • How should the PTA deal with overlap between PTA WTO, if it can. Indeed, the record of experience with PTA obligations and those of the WTO or other PTAs? In disputes shows that except in the Southern Cone Common practice, most give the complaining party the right to Market (Mercosur), where MFN tariffs are high, many PTA ask for consultations in either or both forums, but they parties have chosen to turn to the WTO. prohibit a second panel process after a panel has been requested in one forum. References The parties to a PTA will also need to consider how they De Melo, Jaime, and Arvind Panagariya, eds. 1996. New Dimensions in Regional Integration. Cambridge, U.K.: Cambridge University Press. want dispute settlement to operate. Institutions are a thresh- Schiff, Maurice, and L. Alan Winters. 2003. Regional Integration and old question: if negotiators want greater consistency and Development. Washington, DC: World Bank. predictability in panel decisions, the support of a secretariat World Bank. 2000. Trade Blocs. Washington, DC: World Bank. 1 Beyond Market Access Jean-Pierre Chauffour and Jean-Christophe Maur Economists have repeatedly warned against them, non- This vision is expressed in two ways in recent PTAs. The governmental organizations (NGOs) have fought them, first is the pursuit of what can be termed a “WTO+� agenda, and some governments have signed them begrudgingly (at focusing on disciplines already espoused by the World least in appearance). Yet in the past 20 years, preferential Trade Organization (WTO) but often expanding their trade agreements (PTAs) have multiplied unremittingly. depth and breadth and seeking enforceability. The second What is even more striking is that their scope has broad- is through rules and disciplines that are not covered by the ened at the same time as their numbers have grown. Deep WTO, or are covered very imperfectly (WTO extra). In integration provisions in PTAs have become ubiquitous. practice, PTAs often pursue both objectives, to varying This first chapter looks at the background of the drive degrees. North American PTAs, for instance, focus more on toward deep integration PTAs and at how they differ, in WTO+ disciplines, while adding a few WTO extras to the content and implications, from traditional market access mix. By contrast, European PTAs include numerous WTO- agreements. It then discusses the theoretical and practical extra aspects. Horn, Mavroidis, and Sapir (2010) identify motivations behind today’s deep PTAs. Finally, it highlights no fewer than 38 areas in U.S. and European Union (EU) key areas for policy makers to consider as they contemplate PTAs that aim to go beyond WTO disciplines. their future PTA strategies. The proliferation and deepening of PTAs may offer developing countries vast opportunities to modernize and upgrade their rules and disciplines with a view to greater A Preference for Deep Integration economic efficiency. At the same time, these trends pose a Gaining market access or preserving a level playing field serious challenge for policy makers, especially in low- has remained an important motivation for entering into income countries, because of the added burden of covering PTAs. But with the liberalization of trade around the world an increasingly large and complex set of issues with limited and the related diminishing size of preferential rents, the administrative resources for negotiation and implementa- growing success of PTAs cannot be explained by traditional tion, and frequently with no preexisting experience.1 market access motives alone (even factoring for the possi- Indeed, PTAs are increasingly addressing policy areas that ble substitution of tariffs for less transparent forms of pro- are entirely new to developing countries. These broader tection). Countries are also interested in a host of other agreements may deeply affect countries’ development objectives—importing higher policy standards, strength- processes. To take an often-cited example, it is possible that ening regional policy coordination, locking in domestic the inclusion in PTAs of the most advanced forms of intel- reforms, and even addressing foreign policy issues (see lectual property rights (IPR) protection may require an alter- Schiff and Winters 2004; Hoekman, ch. 4 in this volume). native economic development model whereby knowledge All this translates into a beyond-market-access vision and know-how are no longer acquired through imitation for PTAs that includes a broad set of rules and disciplines and reverse engineering (as happened with the generic phar- governing areas such as investment regimes, technical and maceutical industries in middle-income countries), but sanitary standards, trade facilitation, competition policy, through a less optimal, more demanding, and yet-unproved government procurement, intellectual property, environ- process of accumulation of capital and knowledge. ment protection, migration, labor rights, human rights, and The deep integration commitments in new PTAs, with other “behind the border� issues. their concomitant challenges, stand in sharp contrast to 17 18 Jean-Pierre Chauffour and Jean-Christophe Maur older trade agreements, which chiefly had to do with dis- mechanisms for redress, whereas positive integration mantling barriers to trade and making trade policy sim- requires taking active steps toward integration by defining pler to administer. Although multilateral trade agreements common policies and setting up the legal and administra- under the WTO have pursued a similar path toward tive framework to implement them. The difference is, how- greater complexity—for instance, with the 1994 agree- ever, not as clear-cut as it appears at first (see, e.g., Ortino ment on Trade-Related Aspects of Intellectual Property 2004; Torrent 2007). In both cases, a certain degree of legal Rights (TRIPS) agreement—nowhere is the policy ambi- alignment is required, as is the establishment of minimal tion as sweeping as in PTAs, under which regulatory disci- common institutions. For instance, agreeing on new rules plines are spreading to nontrade areas. that limit the way governments can intervene in markets Before looking at the issues in detail, it is useful to try to could be seen as an instance of either positive or negative capture the essence of the difference in the nature of the integration.3 liberalization challenge posed by the new disciplines in Nevertheless the distinction remains useful for thinking PTAs. The trade literature usually characterizes this process broadly about important characteristics of deep integra- in terms of positive versus negative integration and of tion, because new dimensions of PTAs clearly imply behind-the-border versus at-the-border integration. greater retooling of legal frameworks at the domestic level. Positive integration can be conducted in various ways, depending on how it is legally instrumentalized. Torrent Positive versus Negative Integration (2007) notes, for instance, the substantive differences The first Nobel laureate in economics, Jan Tinbergen, artic- between U.S. and EU agreements regarding procurement ulated the notions of positive and negative integration in provisions. The U.S. approach is more normative in that it characterizing the process of international economic inte- inserts the rules in the agreement, whereas the EU adopts a gration (Tinbergen 1954). Negative integration refers to more progressive approach by defining the rules through the removal of trade barriers and the principle of nondis- specialized organizations such as expert committees. Rela- crimination. This is the traditional remit of trade negotia- tive to negative integration, positive integration entails tions.2 Tinbergen defines positive integration as substantial differences in the drafting of language in agree- ments (the instruments of implementation being more [the] creation of new institutions and their instruments complex) and therefore in negotiations and, probably, in or the modification of existing instruments. . . . More the predictability of implementation. For instance, when generally, positive integration should consist of the cre- tackling trade facilitation issues, it is not sufficient to agree ation of all institutions required by the welfare optimum on items that should be prohibited (e.g., the use of con- which have to be handled in a centralized way. (Tinbergen sular fees) or on simple positive obligations such as trans- 1954, 79) parency; countries must also agree on standards for proce- Analysts have often retained the first part of the dures, such as use of risk management screening at definition—that integration is not just about removal of borders, and must monitor agency conduct. These obliga- barriers to trade flows but about “rule making� to facilitate tions are not easily incorporated into normative commit- these flows. Interestingly, though, Tinbergen offers in the ments in trade agreements—Messerlin and Zarrouk second part of the definition a vision that suggests that (2000), for example, take the view that they should not be. the creation of intergovernmental public goods could also Beyond adopting new policies designed to open markets, be welfare enhancing. This is an important aspect, to positive integration also seeks coordination of policies with which we will return. trading partners, which may imply some form of institu- Various interpretations of the Tinbergen characteriza- tional arrangements. tion have survived in the literature (e.g., Pelkmans 1984; Hoekman and Kostecki 2009; Ortino 2004, 18–34; Torrent Behind-the-Border versus At-the-Border Policies 2007). We take from Tinbergen’s definition the basic intu- ition that positive integration calls for public intervention Another important dimension is characterized in the liter- to tackle market failures that would otherwise prevent eco- ature as behind-the-border versus at-the-border measures. nomically optimal levels of integration. National treatment and uniformization of obligations Positive and negative integration have substantively dif- indeed differ in substance from most favored nation (MFN) ferent implications for the process of integration. Negative obligations in that they require countries to change policies integration would mainly seek the prohibition of a narrow that affect internal transactions that are not necessarily set of policies, as well as joint surveillance and, eventually, related to trade. Beyond Market Access 19 The question of the impact of domestic regulations on bilateral or regional setting rather than at the multilateral trade is not new and is well recognized in the WTO. GATT or WTO level. Implied by his analysis was a sense Domestic policies have the potential to be designed so as to that the removal of traditional trade barriers was not nec- discriminate against foreign producers. Article III of the essarily solving the issue of market access. Schiff and General Agreement on Tariffs and Trade (GATT) accord- Winters (2004) subsequently reviewed alternative ratio- ingly requires that internal regulations comply with the nales for PTAs. These rationales, either nonstandard or not national treatment principle, which states that other well represented by traditional theoretical models, include nationals should be treated the same as one’s own. Beyond domestic policy anchoring, importation of good regula- addressing discrimination per se through the national tory practices, supranational coordination to achieve treatment principle, there is also a desire on the part of pol- regional policy goals, export of regulatory standards by icy makers to reduce the costs of having to comply with hegemons, and foreign policy considerations. These eco- multiple and heterogeneous requirements. As the world nomic, societal, and political-economy motives for con- economy becomes more integrated and supply chains cluding a PTA are discussed in detail next. incorporate sourcing from many countries, the calls for some uniformization are growing. This is an area in which Economic Motives PTAs play an increasing role. Behind-the-border policies directly affect domestic Market access mercantilism is the traditional force behind transactions and thus have obvious direct welfare implica- the push for trade liberalization. Led by the false logic that tions that differ from the indirect effect through prices import barriers should be lowered only if reciprocal access and volumes of trade goods. Their effect also implies a dif- for exports is granted, countries mutually agree to liberal- ferent political-economy equilibrium. Moreover, as we ize their markets, and in most cases, the result is welfare- will see later, the notion of behind-the-border measures enhancing liberalization. In a globalized world, countries could be expanded to measures that are included in trade seek to gain competitive advantage over their neighbors by agreements not because of their direct or indirect effects negotiating special (preferential) market access with key on trade, but merely because trade agreements provide a destination markets. convenient vehicle for international negotiation or Several facts challenge this traditional explanation. Pref- enforcement. erences, to start with, may not be as important as in the In sum, deep integration measures may impinge on past. Tariffs have been falling worldwide (figure 1.1), and in domestic policies that are not necessarily directly trade a very general sense, even the most protected markets now related. They require more advanced reform of the legal tend to exhibit tariff levels that are moderate compared environment and, generally, a more complex set of instru- with those of 15 years ago. There are obviously many ments for implementation. They also may involve active exceptions at the product level. Developed economies and supranational coordination. It is not hard to imagine how middle-income countries exhibit, on average, lower levels demanding and complex liberalization of these measures of protection than low-income countries. might be. Moreover, as PTAs grow in number, so does the number of recipients of preferences, leading to the erosion of the preference margins held over competitors. Carrère, de Motivations for Deep Integration Melo, and Tumurchudur (2010) construct an adjusted The reality of the new PTA landscape raises questions about market access measure of what countries receiving EU the motives for entering into regional agreements. Why preferences actually enjoy when the preferences given to would policy makers around the world invest time, political other partners are taken into account. When this measure capital, and resources in negotiating trade arrangements is compared with the unadjusted measure of preference that discriminate among trading partners and offer uncer- over the MFN tariff, it turns out that real market access is tain welfare benefits, when a multilateral approach of often much lower—for example, less than half for Cambo- nondiscriminatory market access provides a superior solu- dia, a recipient of the EU’s Everything But Arms (EBA) tion? The answer can only be that policy makers are look- preferences. In some instances, as in the case of a generalized ing for benefits that extend beyond market access for goods system of preferences (GSP) recipient, such as Indonesia, and services. there is no effective market access preference at all. As Krugman (1993) assumed that one reason for the suc- noted by Levy (2009), the reciprocal incentive apparently cess of PTAs was the convenience of dealing with the vari- fails to explain the rationale behind asymmetric North- ety, complexity, and opacity of modern trade barriers in a South types of agreement. Many developing countries 20 Jean-Pierre Chauffour and Jean-Christophe Maur Figure 1.1. Most Favored Nation (MFN) Tariff Rates, Weighted Mean, All Products 26 21 percent 16 11 6 1 88 90 92 94 96 98 00 02 04 06 08 19 19 19 19 19 19 20 20 20 20 20 high income: non-OECD high income: OECD lower middle income upper middle income low income Source: World Bank, World Development Indicators database, http://data.worldbank.org/data-catalog/world-development-indicators. Note: OECD = Organisation for Economic Co-operation and Development. already benefit from very good market access in their decreasing and sources of comparative advantage can be northern partner countries. found in small cost differences (“thin� margins of compar- To sum up: there is a tendency toward diminishing ative advantage). Even preferences that are small on paper MFN tariffs; preference margins are actually smaller than may become attractive for prospective partners. they appear; and some developing countries already enjoy Market access conditions are not determined only by virtually tariff-free access to major markets under the tariffs. First, customs procedures and other domestic poli- GSP, EBA, and other preferential regimes. Under those cies, such as standards, may affect foreign exporters’ costs circumstances, can market access incentives alone explain of access to the market. As noted by Bagwell and Staiger reciprocal liberalization in the PTA context? (2001), when governments choose these policies unilater- Market access may persist as a motive in North-North ally, there is a possibility that market access might be set at and South-South agreements. In the global South, in par- a lower and less optimal level than under reciprocal liberal- ticular, tariffs remain fairly substantial. Other incentives ization negotiations. may also be in play. Countries at the periphery of a net- Second, the market access question is not limited to work of agreements (for instance, the partners of the EU goods. Foreign investment is another way of gaining access and the United States) may suffer because industries shift to foreign markets, and the inclusion in agreements of dis- toward the hub of the network and away from peripheral ciplines relating to investment can be an additional motive countries (the spokes) and because of erosion of the outly- for reciprocal liberalization commitments.4 Many PTAs ing countries’ preferential access, since location in the hub now include investment disciplines that go beyond those provides preferential access to many more markets. This of the WTO. WTO rules are limited to the supply of services reality is what has led countries such as Chile, Mexico, and following an investment (commercial presence), as speci- Singapore to pursue “spoke-spoke� strategies by mirroring fied in the General Agreement on Trade in Services (GATS), their large trading partners’ PTA policies and pursuing and to the trade-related investment measures (i.e., the agreements with the same partners, even though their Trade Related Investment Measures [TRIMS] agreement). trade with such distant partners might be small. The strat- Moreover, GATS relies on an “enterprise-based� definition egy of the European Free Trade Association (EFTA) in par- of investment, whereas bilateral rules generally refer to a allel to the EU, and the accession of new countries to the broader “asset-based� definition that covers portfolio EU, might be seen as being driven by a similar motive (see investment and different forms of tangible and intangible Baldwin 1994). Bhagwati (2008) also argues that even property (Miroudot, ch. 14 in this volume). modest margins of preference have a sizable impact in a Third, because traditional PTA analysis focuses on globalized world in which overall transaction costs are trade in goods, trade in services is often omitted from the Beyond Market Access 21 discussion. Yet the services sector represents the largest, pricing below cost for the poorest customers in such and a growing, share of gross domestic product (GDP) in services sectors as water, electricity, finance, and trans- many developed and developing countries; many services port. Hoekman, Mattoo, and Sapir (2007) observe that (e.g., electricity, telecommunications, transport, and pro- not only do political-economy calculations become more fessional services) are key inputs into the production of complex in this environment, but also the usual reciproc- goods and other services; and the information technology ity mechanism of trade liberalization may not work any (IT) revolution has increased the tradability of services. In more because of the difficulty of clearly separating meas- these circumstances, services liberalization may offer con- ures that promote market access from measures that pur- siderable gains, both from increased trade flows and from sue legitimate regulatory objectives. reduced input costs for firms.5 For some country group- Aside from market access considerations, PTAs have a ings, such as South-South agreements, preferential integra- role to play in transnational regulation. This involve- tion in goods may bring little benefit; small countries ment reflects the standard economic-efficiency motive for with similar production structures and with small and regulation—addressing market failures. Three often-cited inefficient manufacturing sectors might not have much to market failures are in the areas of monopoly power, exter- gain from engaging in goods-only PTAs. A promising next nalities and the provision of public goods, and informa- step might be to explore other integration dimensions in tion asymmetries. which complementaries might be beneficial, such as 1. Monopoly power and supranational competition. services (Mattoo and Sauvé, ch. 12 in this volume), invest- Economies of scale and, more generally, market failures ment (Miroudot, ch. 14 in this volume), and labor mobility give rise to the possibility of monopoly power and abusive (Stephenson and Hufbauer, ch. 13 in this volume). Hoekman conduct by private firms. Trade liberalization may go some and Sekkat (2010) examine this option in the case of the way toward creating competition by making markets con- Pan-Arab Free Trade Area (PAFTA). testable, but this will not always be sufficient. Domestic Yet the reality is that even if some limited sectoral enforcement of competition rules is linked to market advances (on movement of professionals, for instance) access. If, in a national jurisdiction, competition is weak have been recorded in recent agreements, PTAs have made because of lack of enforcement, market concentration only modest inroads where access to services markets is and collusion in the domestic market may deter entry by concerned. Regulatory policies tend to pursue noneco- foreign suppliers. In such instances, competition policy nomic objectives along with economic concerns (such as should complement trade liberalization to secure the lowering the costs of barriers and compliance), and this, gains from the opening of markets. Hoekman, Mattoo, and Sapir (2007) remark, makes for a The threats of market power and abusive conduct may particularly complex political-economy calculus. As with not justify the inclusion of competition rules and disci- tariffs, the transaction costs imposed by deep integration pline in a PTA on economic grounds alone. After all, policies will lead incumbent services industries sheltered countries can individually opt to implement competition by regulatory protection to resist liberalization. policies unilaterally. But such policies may not be effective The bias against liberalization can be reinforced in the in dealing with the risk of cross-border externalities and case of services by reluctance on the part of consumers and the abusive behavior of exporters abroad. Competition government. Consumers may fear that regulatory liberal- rules may be particularly relevant in PTAs where the risk ization will affect their well-being—for instance, through of abuse of market power or collusive practice involves slacker standards and lower quality of products and serv- more than one national jurisdiction and where interna- ices.6 Government and regulatory agencies may also view tional legislation and cooperation could effectively curb liberalization reluctantly, for several reasons: (a) regulation anticompetitive behavior. For instance, a firm may use its may be a source of indirect taxation, in that governments market power in one market to extract monopoly rents in benefit from rents generated by regulatory protection (as is another; a dominant position may span several countries common in the area of standards); (b) governments may (as with Microsoft), potentially leading to anticompetitive fear that their latitude to pursue regulatory objectives will market conduct; or firms may have agreed in one jurisdic- be curtailed because cross-border supply could undermine tion to collude in another, making it necessary for author- local suppliers while being subject to different (lower) ities to cooperate in order to collect evidence. regulatory requirements; and (c) governments also pur- Because such competition issues are related to trade and sue redistribution objectives by, for instance, imposing investment, there are complementarities in dealing with requirements for universal provision (e.g., in water sup- them in the same forum as trade arrangements. PTAs offer ply, telecommunications, and postal services) and for a scope for creating disciplines that the WTO does not. 22 Jean-Pierre Chauffour and Jean-Christophe Maur Arguably, the degree of cooperation in international com- • Joint decision making to ensure that national policies are petition arrangements will depend on the size of individual coordinated at the regional level (e.g., management of economies, the level of trade, and the enforcement capacity food stocks). of the actors. • Transfer of resources to solve externality problems when 2. Externalities and provision of public goods. An exter- contributions by individual member states are required. A nality (or transaction spillover) is a cost or benefit, not common instance is when institutions are weak and transmitted through prices, which is incurred by a party capacity building is needed to bring a partner country that did not participate in the action causing the cost or to a higher standard for the regional common good benefit. As the examples of climate change and the deple- (e.g., customs enforcement). tion of fish stocks show, externalities are not necessarily confined within the borders of a given country. In some Regional externalities should arguably be dealt with in cases, externalities may be best tackled by a small group of those jurisdictions in which they occur, and they therefore countries; for instance, river management and some trans- require transnational mechanisms of cooperation. It is not port issues should involve neighboring countries. External- entirely clear whether regional externalities (positive or ities are closely related to the need to provide public negative) should necessarily be addressed in the specific goods—that is, goods that are nonrivalrous and nonex- context of a PTA (Schiff and Winters 2002). There is always cludable.7 In the presence of externalities, markets may not the possibility (as for all commitments agreed in PTAs) of spontaneously provide goods, such as clean air, that are addressing these issues through dedicated agreements and socially desirable. transnational institutions such as bilateral customs or Addressing regional externalities should logically be a water management agreements. Historically, many such priority of regional PTAs, given the need for some form of problems have been addressed in this way. supranational coordination to help internalize the exter- 3. Information asymmetries. Sometimes goods charac- nalities or share them fairly. Coordination can take several teristics may not be discernible to buyers before consump- forms: tion. Credence goods do not—for example, chemicals may be harmful to health, unbeknownst to the people exposed • Alignment (for instance, through mutual recognition to them; thus regulation is needed to inform consumers agreements) or harmonization of policies. These meas- before purchase. Information asymmetries may affect pro- ures eliminate segmentation of markets and duplication ducers themselves in situations where consumers’ charac- of the costs generated by barriers at the border. teristics are hidden (e.g., in the insurance market). In most • Alignment and harmonization of policies to avoid leakage. instances, the market itself deals with these information Leakage is a concern when, for instance, one jurisdiction asymmetries through information dissemination and in the PTA has lower regulatory standards that might brand signaling. When, however, the asymmetries are not undermine the regulatory efforts of its trading partners. addressed, the market outcome is suboptimal. The problem An example is a country’s deficient control of animal of information is particularly acute for services because of epizooties or pests that spill over to neighbors. (Animal their intangibility, which makes it harder for the buyer to border crossings cannot be totally controlled.) learn about quality prior to consumption (Hoekman, • Alignment and harmonization of policies to create net- Mattoo, and Sapir 2007). Regulation may then be called works and to facilitate information exchange. This for—perhaps through licensing or the imposition of com- method essentially refers to the adoption of common pulsory standards. International cooperation may help standards and regulatory language in order to facilitate reduce the overall complexity of the regulatory framework flows within the region (for instance, ensuring interop- for international traders by aligning and harmonizing erability of national networks at a regional level). Such regulations. Failure to tackle issues of this kind in a coordi- alignment is of particular relevance for services sectors nated fashion may generate negative externalities. such as finance and insurance, IT, professional services, In the specific PTA context, the challenge will be to transport, and electricity. assess whether information asymmetry problems are best • Pooling of efforts to create infrastructure or pooling finan- tackled at the bilateral or regional level, rather than in cial and human resources to provide a regional public other international forums. In most instances, this will be a good. For example, combined financing might be question of judging the trade-offs between the transaction needed for a large infrastructure serving a region, such costs of cooperating with a limited number of countries, as as a hydroelectric dam, or for a large port. against the international community, and between the Beyond Market Access 23 benefits of coordination at the PTA versus the global level. procedures) in international flows arising from the trade There is a clear risk, for example, that regional standards in agreement would spill over into other domestic areas.8 PTAs may exclude third, nonadhering, countries. Never- Clearly, in the case of trade and development, there are com- theless, it may be much easier to agree on a common plementarities between openness and poverty-alleviating approach with a small number of countries and with coun- growth, and PTAs help target specific countries. In other tries with similar preferences. Finally, in some cases (e.g., instances, and more prosaically, there might merely be a regional epizooties), a neighboring-country approach will quid pro quo between market access in the North and be appropriate. concessions on other fronts in the South. Another motivation relates to the search for the best available forum for promoting the international sharing Societal Motives of societal norms, focusing on issues that are not already Beyond the economic motives, a PTA can be driven by present in other agreements such as the WTO (for instance, societal motives or, as Bhagwati (2008) calls them, value- labor rights or environment protection) or pushing for related demands. Each society has moral and social norms higher standards than currently exist in the international and preferences that may be undermined by market forces community. (See also “Institutions for reform,� below.) left to operate on their own. For instance, trade in danger- This is a clear objective of the new U.S. trade policy of pur- ous weapons or in morally or religiously reprehensible suing PTAs that was initiated under the George W. Bush material may need to be restricted. What is considered administration. In a 2001 speech, Robert Zoellick, then dangerous or morally reprehensible will vary significantly U.S. trade representative, noted that “we need to align the according to country and culture. global trading system with our values. . . . We can encour- Social norms and values may be undermined by trade age respect for core labor standards, environmental protec- liberalization; after all, it is easier to control borders than to tion, and good health . . . And we must always seek to control a whole territory, and foreign producers may not strengthen freedom, democracy and the rule of law� hold themselves to a particular country’s standards. This (quoted in Evenett and Meier 2008). Related to this objec- issue has long been recognized in multilateral trade agree- tive is the desire to use every trade forum to reaffirm these ments, leading to the inclusion of safeguard provisions and choices, with a view toward mutual complementarity and general exceptions (on moral grounds, for instance). Safe- reinforcement between the different instruments. guard mechanisms and the language of general exceptions may, however, prove insufficient, and countries may want Political-Economy Motives to negotiate sector-specific conditions in PTAs, such as reservations concerning universal provision of services. Beyond the need for coordinating policy making with trad- Conversely, trade agreements may help further societal ing partners, PTAs also serve as forums for policy objec- objectives. Development policy concerns, for example, are tives that are strictly related neither to exchanges nor to the increasingly present in agreements such as the EU-sponsored preferential nature of PTAs. PTAs can be seen as efficient economic partnership agreements (EPAs). Northern part- forums for achieving broader geopolitical, institutional, ners also push for provisions related to good governance, and policy-anchoring objectives. democracy, labor rights, and human rights (Elliott, ch. 20 in Geopolitical objectives. Geostrategic considerations have this volume; Aaronson, ch. 21 in this volume). historically commanded the formation of PTAs. There are What is the specific value added of PTAs in helping to numerous examples of trade agreements that have been achieve these objectives? One motivation might be that used to promote peace. Chief among them is the EU, threats to societal preferences are localized in a limited which was born from the desire to prevent war from hap- number of partners, and thus it makes sense to deal with pening again in Europe. Winston Churchill called in 1946 those countries directly. Some PTAs, for example, have for a “United States of Europe,� but it was with economic been specifically linked to measures for fighting narcotics integration and the 1951 European Coal and Steel Com- production and trafficking; an example is the Central munity that European integration began (Winters 1997; America Free Trade Agreement and the U.S. stipulations Baldwin 2008).9 Other examples of agreements used for concerning narcotics in that PTA (Hornbeck 2003). stability purposes, as noted by Bergsten (1996) include the PTAs might be seen as a locus of positive spillovers Southern Cone Common Market (Mercosur, Mercado between trade and societal policy issues. For instance, Común del Sur) and Asia-Pacific Economic Cooperation provisions on governance (e.g., open and transparent (APEC). More recently, the push by the United States to 24 Jean-Pierre Chauffour and Jean-Christophe Maur conclude PTAs has had foreign policy motives (Bhagwati Sapir (2009), who show that binding and nonbinding pro- 2008; Evenett and Meier 2008), as has Europe’s neighbor- visions coexist in agreements on nearly all the issues cov- hood policy (European Commission 2007).10 ered. Such flexibility, it should be noted, might appear as a Thus, PTAs can contribute to delivering peace and sta- virtue to policy makers but may not necessarily contribute bility as a regional public good (Schiff and Winters 2004; to factual reform. World Bank 2005, box 2.6). Two mechanisms may come In contrast to multilateral forums, PTAs often feature into play. First, trade exchanges increase economic interde- innovative institutions. One innovation is the involvement pendence and, thus, act as a disincentive for conflict. They of the private sector, from participation in stakeholder may also help increase familiarity and trust and defuse forums to the possibility of lodging complaints in, for trade-related disputes.11 Second, and more specific to example, the European Court of Justice, the General Secre- PTAs, institutions themselves serve as a conduit for diplo- tariat of the Andean Community, or under the investment macy, allowing for frequent and repeated interaction provisions of the North American Free Trade Agreement among officials and for better exchange of information (NAFTA). Some PTAs offer more substantial transfers of (Haftel 2007). Deep PTAs seem more attractive in this sovereignty. Governments can also opt to devolve some of respect because they have more sophisticated institutions. their authority to institutions created by PTAs, such as Empirically, Mansfield and Pevehouse (2000) have regional competition authorities, as described by Dawar found that membership in a PTA significantly decreases and Holmes (ch. 16 in this volume). the likelihood of armed conflict. More recently, Lee and The transaction costs of agreement are lower in PTAs Pyun (2009) provide statistically significant evidence that with a small number of participating countries. In addi- PTA institutions decrease the probability of conflict tion, small PTAs do not lend themselves to free riding, between members, whereas WTO membership seems only which is a practice that poses a key obstacle to successful marginally significant. Martin, Mayer, and Thoenig (2010) global liberalization (Krugman 1993). Lower transaction test the interaction between conflict and PTAs over the costs allow for more binding constraints on each partner period 1950–2000 and find that the hypothesis of geopolit- (noncooperation is more difficult) and for legal flexibility. ical motivations behind the agreements is supported by Since the number required to reach consensus is lower, the evidence. Yet for PTAs to help ease the probability of con- agreement could be amended and revisited more often flict, there must be sufficiently large trade gains between than is the case with multilateral agreements. A smaller the partners. Economics and political motives thus com- number of participant countries enables more frequent plement each other. and probably less formal interactions, which can con- Institutions for reform. By offering a different set of insti- tribute to problem solving and deeper relations. That tutions and related services from those of other forms of seems an important feature for the regulatory aspects, international agreements, PTAs provide an infrastructure which require agreement on complex issues (such as for institutional dialogue and cooperation. As noted by the mutual recognition arrangements) and the setting up of World Bank (2005), many issues covered by PTAs, such as expert bodies. This is the road followed by the EU under the externality problems described earlier, could well be the Florence Forum.12 handled without a trade agreement. If PTAs are used, this Resource transfers are more likely to occur in the frame- must be because they are perceived as offering a good work of PTAs than in other international agreement set- framework for achieving progress. tings. Many PTAs—North-South ones, in particular—do PTAs are relatively flexible instruments insofar as they incorporate such transfers. Agreements signed by the EU allow for various levels of legal commitment and offer are the most striking examples; other cases include U.S. nearly infinite ways of creating policy space. For example, free trade agreements (FTAs) with Latin American partners the options with dispute settlement are numerous: there and South-South agreements such as the Common Market may be no mechanism at all, or one or several dispute set- for Eastern and Southern Africa (COMESA). Resource tlement mechanisms (Porges, ch. 22 in this volume). Each transfers matter, in particular, for deep and asymmetric PTA can come with its own ad hoc instruments, which may PTAs. Arguably, deep integration places heavier demands be sector specific or may refer to external mechanisms such on capacity. Less developed trade partners may have diffi- as international arbitration or WTO dispute settlement. culties in, for example, meeting the regulatory standards of Various ways of reaching settlement before recourse to for- their partners and thus obtaining effective market access. mal dispute settlement are available, such as good offices, They may lack capacity to compensate for some of the third-party mediation, and conciliation. Indirect evidence adjustment costs of reform; to contribute effectively to the of legal flexibility is also provided by Horn, Mavroidis, and production of regional public goods; and, even more Beyond Market Access 25 broadly, to help achieve geopolitical and societal objectives Rica and Peru used their FTA negotiations with the United (development, conflict prevention, and so on). States to push domestic reforms. Levy (2009) does not find Policy anchoring. A traditional political-economy expla- much in the way of market access motives for Peru, except nation for a country’s entering into binding international a desire for greater certainty about future access (trading trade commitments is the pursuit of a domestic reform temporary preferences for more permanent ones), but the agenda and the use of external commitments to lock in agreement did help the country cement its economic pol- progress and prevent future reversals. The opportunity to icy reforms. Some of these reforms, notably in the areas of lock in is also a motive for including behind-the-border services and investment, were part of the FTA implementa- aspects in agreements. PTAs may be perceived as more tion program, but the agreement also helped lock in prior effective lock-in mechanisms than other international policy reforms, such as tariff reductions. Another motiva- agreements, and they may complement other external tion for signing the FTA with the United States, Levy notes, instruments in the process of reinforcing and consolidat- may have been the hope that it would generate broader ing domestic reforms. positive spillover effects on Peru’s governance and the rule By extending their reach to regulatory issues, PTAs offer of law. By imposing good disciplines to protect foreign a way of improving policy credibility (Hoekman, ch. 4 in investors and market access, the FTA would signal a com- this volume). What are the differences, then, between the mitment to a better legal environment, in general. sort of anchor offered by PTAs and that provided by the In sum, there are strong rationales for policy makers to WTO? Aside from the obvious point that PTAs may offer embark on deep and comprehensive PTAs, but the relative commitments in WTO+ and WTO-extra areas, they may merits of regional integration are also issue specific and have specific advantages. The possibility of picking a part- country specific. The choice of whether to include regula- ner may help reinforce credibility, as the partner of choice tory aspects in PTAs is essentially dictated by a dual concern: may be perceived as a strong proponent of reform. The EU, securing market access and addressing market failures, the United States, and other developed countries do, in whether national (through the lock-in effect of policy fact, promote various agendas through their respective reform and policy upgrading) or regional. Market failures PTAs. Picking a partner or a group of partners may also will be of different natures and will involve different sets of signal a preference for a certain regulatory approach. In countries, depending on the sector and the issue at hand— addition, lock-in through PTAs can be complemented by hence the need, as with any regulation, for a case-by-case transfers of finance and knowledge. approach. Other noneconomic considerations, such as fos- An argument put forward by Schiff and Winters (1998) tering societal choices, may apply to some issues of a regu- is that PTAs may actually be well suited for locking in poli- latory nature. Finally, specific institutional characteristics cies because of the credibility of enforcement in these and advantages may motivate the choice of PTAs as ade- agreements. Incentives to enforce commitments are greater quate forums for reform. in a PTA because there is less possibility of free riding and fewer of the coordination problems that may arise in Specificities of Deep Integration multilateral forums. In addition, there is more scope for retaliation because concessions in a PTA may go beyond The increased scope and depth of PTAs create opportuni- just tariffs. Schiff and Winters note, however, that the disci- ties but also pose extra challenges to policy makers as they plining effect is limited to the partner countries in the PTA, negotiate and implement the complex market-access and not third-country members. (They cite the peso crisis of regulatory web of these agreements. Policy makers may 1994–95, when Mexico raised its tariffs on 500 items for have to reevaluate their approach when negotiating and non-NAFTA suppliers.) These dynamics are echoed by the implementing deep integration PTAs. In particular, to what conclusion reached by Prusa (ch. 9 in this volume), that extent are the multiple goals of PTAs consistent and PTAs tend to discipline the use of contingent protection congruent? Do the new disciplines incorporated in PTAs measures among partners while, at the same time, the use of create a different category of obligations? Does the deepen- protection against third countries seems to be increasing. ing and popularity of PTAs create new challenges for the There are several recent examples of countries that have multilateral trading system? In this section, we suggest four used PTAs to pursue an ambitious domestic agenda; Schott major areas of emphasis for policy makers, especially in (2003) cites Mexico and Chile. Similarly, the accession of developing countries, as they refine their regional trade Eastern European countries to the EU was strongly moti- strategies: reexamining the question of discrimination and vated by the desire to break irrevocably with socialism and preferential access, adopting a holistic approach, building consolidate market economy reforms. More recently, Costa in flexibility, and focusing on implementation. 26 Jean-Pierre Chauffour and Jean-Christophe Maur Preferences government procurement rules do not affect the market as a whole, and therefore, exclude suppliers only from serving Are the traditional concerns of discriminatory liberaliza- the public share of the domestic demand (Dawar and tion valid for the new areas of deep commitments in Evenett, ch. 17 in this volume). Depending on the size of PTAs? Do deep integration measures generate trade diver- public markets, this may not be enough to exclude foreign sion? Can they harm the liberalizing country? Do they act suppliers from the market altogether. as stumbling blocks to further liberalization? These Impact on third parties. The standard effect of discrimi- issues—in particular, those related to the impact of deep nation will still be harmful for third parties (the excluded integration on multilateral architecture—have generated countries). This is, for instance, the case when countries considerable interest as of late (OECD 2003; Baldwin, adopt European standards instead of international ones Evenett, and Low 2009; Estevadeordal, Suominen, and Teh (Maur and Shepherd, ch. 10 in this volume; Stoler, ch. 11 in 2009), thanks to mounting evidence provided by new this volume). It is important to ask, however, whether pref- PTAs. The new regulatory commitments found in today’s erential measures are always discriminatory. PTAs are discussed here in light of the three classical eco- An important characteristic of regulatory measures is nomic concepts for analyzing market access discrimina- that de jure preferential treatment might be difficult to tion in PTAs: trade diversion, third-party effects, and apply, making de facto MFN liberalization a preferable systemic effects. option. That is, devising a new regulatory regime applica- Trade diversion. Discrimination in deep integration ble to each PTA may be impracticable, or the concept of agreements can secure the benefits of market access with- rules of origin that applies to product characteristics sim- out generating the potential cost of trade diversion. In this ply cannot be as easily applied to regulations or intangible sense, regulatory discrimination does not raise the same transactions.16 For instance, provisions on protection of concerns as tariff discrimination would. The certitude that intellectual property rights apply equally to all origins, better market access will be beneficial and that no diversion including domestic ones. Carving out specific regimes for costs will occur leads to an important consideration for some countries (as in the case of the WTO Article 6 excep- policy makers: all things being equal, PTA partners will tion for least-developed countries) requires complex legal unambiguously gain in preferential deep integration and practical arrangements. Similarly, for customs proce- efforts. This may explain why deep integration issues are dures, although trade rules may differ depending on the winning popularity in PTAs. (See Baldwin, ch. 3, and origin of the product, it makes sense to maintain, as much Baldwin and Freund, ch. 6, in this volume.) as possible, similar procedures regardless of the origin of Protection afforded by lack of regulatory openness is not the good because most objectives of border controls apply necessarily protectionist in intent. Regulatory requirements to all imports. Those examples show that the concern often impose a transaction cost on the exporter without about negative impact on excluded parties can largely dis- generating rents for the home country.13 A case in point is appear in the case of deep commitment provisions and superfluous or antiquated border controls, which create that preferential liberalization could generate positive additional costs without any corresponding benefits.14 externalities for third countries.17 In such instances, liberalization of services, harmo- This is, however, not a universal rule. A characteristic of nization of standards, trade facilitation, investment liber- deep integration liberalization is that there are instances in alization, and openness of government procurement can which discrimination is inevitable and even necessary. The generate benefits even if carried out preferentially. It is, main illustration of this conclusion is provided by mutual however, important to stress that this positive effect only recognition agreements (MRAs). MRAs can be negotiated occurs if there is no sizeable rent transfer from domestic to in any regulatory area and are basically a way of lowering foreign producers.15 In trade in services, for instance, the barriers to entry into the domestic market for foreign impact of preferential liberalization will be determined by producers without outright harmonization of rules, thus the nature of the regulatory barriers present. If lack of preserving regulatory diversity and allowing countries to competition is an issue, regulatory liberalization may well maintain national objectives and preferences. Under this replace a domestic monopolist practice with a foreign one principle, parties agree, in essence, to maintain their own (Mattoo and Sauvé, ch. 12 in this volume). Similarly, where regulatory procedures provided that they meet minimum access to services markets is subject to some form of licens- common objectives. Recognition can be agreed both for ing, rents may arise, and with them, the cost of trade diver- regulatory standards and for their testing and can be sion. In the case of government procurement, there is an applied in several areas: services (e.g., professional stan- additional aspect at work: restrictive and discriminatory dards and transport), trade facilitation (e.g., declarations Beyond Market Access 27 made with foreign customs), and technical trade barriers component of an overall dynamic of liberalization (Bald- and phytosanitary measures. Another instance in which win and Freund, ch. 6 in this volume). discrimination is needed relates to customs controls. Mod- In the context of deep integration, similar concerns pre- ern and efficient, risk-based border management calls for vail. How do complex and largely ad hoc PTAs touching on the selective control of imports, focusing on categories that services and behind-the-border measures interplay with present the highest risk of noncompliance. Risk criteria the multilateral order? Part of the answer was provided in discriminate, for instance, by product category, country of the previous section, where the point was made that origin, and identity of shipper, allowing simplified controls liberalization is often MFN in nature, thereby removing for authorized economic operators and express shippers. concerns about stumbling-block effects in these instances. Finally, the fact that liberalization in preferential set- There is also more to this story, as discussed by Baldwin, tings could de facto lead to MFN liberalization has pro- Evenett, and Low (2009) and in OECD (2003). Several found implications for overall liberalization negotiating mechanisms that support further liberalization are actually strategy. Concessions given to one partner cannot be found in PTA provisions: offered again to another when they are nondiscriminatory and are implicitly offered to the rest of the world. One • PTAs may enforce or encourage adherence to interna- implication might be that such liberalization is more diffi- tional standards in, for example, sanitary and phytosan- cult to achieve because it is more likely to be resisted by itary measures and technical barriers to trade (Lesser domestic firms that would lose not only to the preferential 2007). Numerous PTAs also refer directly to WTO rules. partners, as would be usually the case in a trade-diverting • Third-party, nonparty MFN clauses are often found in PTA, but also to the world as a whole (Krishna 1998). The services provisions (Fink and Molinuevo 2007) and in reciprocity rationale for signing North-South agreements government procurement (Baldwin, Evenett, and Low would also be undermined because offsetting market 2009). According to third-party MFN rules, future and access preferences for goods (the objective of the South) more advantageous commitments with other partners against deep regulatory commitments (the objective of the should be granted to PTA partners as well, thus trigger- North) seems to make little sense for developing coun- ing automatic liberalization. A benefit of such rules is tries. Preferences are bound to be eroded over time, but that small countries avail themselves of the bargaining regulatory commitments are both permanent and MFN. power of more powerful countries with common trade Alternatively, as argued by Limão (2007), this asymmetry partners and so gain increased market access (Baldwin, could provide an incentive for PTAs to maintain high bar- Evenett, and Low 2009). riers against third countries (through high preferences), in • When regimes operate under liberal rules of origin order to provide greater incentives for cooperation in non- (ROOs) or liberal “denial of benefits� provisions, the trade areas and postpone to a distant future the threat of provisions are applied not only to preferential trade in preference erosion. goods but also in any instance requiring the establish- Another related consideration is that parties that want ment of the origin of the partner subject to preferential to export a certain regulatory model—one more advanta- rules. There are various instances in PTAs; they include geous to their own firms—could gain from being the first access by third parties to MRAs, ROOs that apply to for- to negotiate with a given country. This may be one aspect eign firms that establish local presence in the partner of the competitive liberalization framework described by country (Mattoo and Sauvé, ch. 12 in this volume),18 Bergsten (1996). and government procurement (Dawar and Evenett, Systemic effects: Deep integration as a building block. Are ch. 17 in this volume). ROOs applying to regulations deep commitments in PTAs building blocks or stumbling often happen to be liberal either because they otherwise blocks with respect to multilateral liberalization? This is a become complex to administer or because, as we saw legitimate question, given that the slew of new commit- earlier, it does not make sense to operate parallel regula- ments in PTAs makes these agreements much more inva- tory systems instead of an MFN system. sive and, by adding new dimensions, may create even more • The diffusion of identical and liberalizing rules in PTAs hurdles for the welfare-superior objective of multilateral has been particularly noted in “contiguous� PTAs hav- liberalization. Even when the more traditional aspect of ing one partner in common. This occurrence can be tariff preferential liberalization is considered, the answer to seen in investment provisions in agreements in North this question is not entirely clear, with some analysts argu- and South America, in procurement provisions, and in ing that PTAs fundamentally undermine the multilateral contingent protection (Baldwin, Evenett, and Low system (Bhagwati’s “termites�) and others seeing in PTAs a 2009). There could, however, be a downside, as large 28 Jean-Pierre Chauffour and Jean-Christophe Maur trading powers export their own—and not necessarily rules concerning movement of persons; and standards compatible—vision of a liberalization agenda. Prusa policies. (ch. 9 in this volume) describes a phenomenon of rules The other aspect of the broader agenda is, of course, the diffusion in which the EU and the United States act as regulatory one and the inclusion of domestic and other spokes in their respective networks. More broadly, tem- policies that have the objective not of protection but of plate approaches to liberalization are often used in remedying some sort of market failure. Such policies are PTAs. Rules relating to investment, services liberaliza- included in PTAs at least in part because trade liberaliza- tion, or standards tend to replicate one of two or three tion interacts with their objectives in ways that may often existing models. seem to make these objectives more difficult to achieve. Trade policies can no longer be designed on the assump- In sum, discrimination in the implementation of deep tion of their separability from other policies.19 commitments in PTAs should not be underestimated, but Deep integration is as much about trade as it is about there are ways of dealing with it. It should not be underes- other dimensions of economic management and public timated because deep integration creates stealthier and policy. Starting with liberalization of services, all deep inte- more complex ways to discriminate. Trade partners can gration policies meet specific objectives, and the liberaliza- push for specific regime designs with the aim of carving tion question cannot be divorced from the consideration of out more favorable market access conditions. One example these goals. Policy makers should carefully think about why is the insistence by the United States on including customs and how trade agreements should serve these objectives in rules in its FTAs that allow for preferential treatment for the specific context of PTAs. Table 1.1 offers a snapshot of express carriers—an industry in which the United States is the variety of such objectives. well represented. However, the parallel with the tariff In Termites in the Trading System (2008), Bhagwati analysis of preferential liberalization does not necessarily pointedly mocks the ever-expanding notion of trade- hold because there is less risk of trade diversion, and the related policies: “If I sneeze and use imported cough syrup, welfare implications of preferential liberalization are then that immediately affects imports; if I use domestic cough necessarily positive. In such cases, PTAs would contribute to syrup, that potentially reduces exports of the syrup I have overall welfare gain. Moreover the rather complex nature of used up.� It is true that by pushing the logic ad absurdum, regulation tends to work to the advantage of MFN liberal- every issue becomes trade related and has a trade effect. ization because managing multiple regulatory regimes to Although this does not mean that the impact of non-trade- create specific preferences is often too complicated. related policies on trade (and vice versa) should be ignored, it is important to be clear about the primary objectives of policies and how to achieve them. The ques- Policy Complementarities: Taking a Holistic Approach tion for regulatory issues, which are, in essence, behind the The expansion of PTAs into new disciplines implies that pol- border and not unique or specific to traded goods, is of icy makers are confronted with multiple policy choices with three orders: different objectives and complex interactions. In essence, new PTAs capture a broader paradigm than traditional ones. 1. What are the issues of true international dimension that Evans et al. (2006) characterize one aspect of this expanded can only be addressed through international agreements? paradigm by pointing out that, unlike traditional trade 2. How should behind-the-border rules in PTAs be liberalization that focuses chiefly on goods trade, deep inte- designed to minimize trade-distorting effects? gration aims at broad factor mobility, including liberalization 3. How should policy makers prepare themselves to nego- of investment (capital movement), trade in services, and tiate or resist such rules? migration and labor standards. Perhaps nowhere are all the liberalization dimensions explored as deeply and compre- The first two questions roughly ask, what is the actual hensively as in PTAs. The complementarities created might link with trade issues? On the first question, as was seen explain the attraction of PTAs (Mattoo and Sauvé, ch. 12 in earlier, market failures and externalities of a supranational this volume). A good example is the trade facilitation nature can be addressed using PTAs; for instance, interna- agenda, which embraces such goals as the streamlining of tional transit is a trade facilitation concern that clearly has numerous border measures (all of which have specific reg- a regional dimension. Arguably, these issues can also be ulatory objectives in sectors such as health, immigration, addressed in separate, dedicated agreements such as bilat- and security controls); the inclusion of services sectors that eral cooperation treaties for competition law or stand- facilitate trade (transport, logistics, insurance, and so on); alone transit agreements (see Maur 2008; Dawar and Beyond Market Access 29 Table 1.1. Types and Scope of Regulatory Objectives in Selected Areas Covered by Trade Liberalization Agreements Area Regulatory objectives Services Universal provision (access, prices) Standards (professional, safety, interconnection of networks) Prudential regulations (banking) Cultural exceptions (media) Standards for goods Human, plant, and animal health Safety Network economies Intellectual property rights (IPRs) Innovation and creativity Trade facilitation Fiscal revenue Border security Prohibitions Immigration control Enforcement of domestic laws with respect to foreign goods Transit Government procurement Preference for national goods Protection of sensitive sectors (defense) Consumer law Consumer information and protection Labor and human rights Minimum standards Environment Public goods Minimum standards Movement of persons Immigration management Source: Authors’ compilation. Holmes, ch. 16 in this volume). International trade may be standards in such a way as to exclude foreign products), or an important source of market failure—for instance, with by raising standards abroad (e.g., by exporting new regula- respect to environmental protection (Anuradha, ch. 19 in tory requirements that increase the cost of production this volume).20 Because of the binding nature of interna- abroad and shift comparative advantage patterns). tional agreements and the international trade dimension of On the third question, that of negotiating or resisting externalities and market failures, there is a space for regula- such rules, a first step is to clarify the policy-making tory frameworks in the context of PTAs. Issues that were process involved in making future commitments—to previously dealt with under dedicated bilateral instruments, begin with, by involving the key ministries and administra- such as bilateral investment treaties, customs cooperation tions that oversee the nontrade objectives and by ensuring agreements, and cooperation on competition policy, are mutual understanding and coherence of objectives. Histor- now increasingly incorporated into PTAs. Although the ically, PTA negotiations have typically been led by finance, jury is still out as to the most effective instrument for foreign affairs, or trade ministries. These ministries would implementation, PTAs may be superior because of the seldom coordinate with other ministries or specialized possibility of issue links and institutional-savings costs, bodies of government, and sometimes their understanding given that one body serves several purposes (Devlin and of the issues at stake is limited. Another step is to minimize Estevadeordal 2006). the costs of meeting the regulatory objectives, with reason- On the second question, that of reducing distortionary able statistical confidence. This step is often not done; effects, the approach should be to minimize the conflicts instead, solutions that meet the objective of regulation irre- between regulatory and trade liberalization objectives. For spective of the costs caused by trade distortions are chosen. instance, harmonization to a low standards level would The notion of risk is often not embedded in the regulatory maximize trade liberalization objectives but clearly would design because agencies have no direct interest in consider- not meet regulatory objectives. (Low standards may not ing the costs borne by other parts of the economy in meet- meet a country’s preferred level of enforcement.) An obvi- ing their objectives and naturally opt for regulatory solu- ous approach is to ensure that the clarity of regulatory tions that minimize risk rather than costs. An example is objectives is such that protectionist intents cannot hide border controls, where 100 percent checking of consign- themselves behind the disguise of rules. Hidden protec- ments is not rare, to the exclusion of economically efficient tionism operates by raising the costs of (foreign) rivals, by methods of targeting only risky shipments. The marginal imposing discriminatory rules at home (e.g., by designing costs of meeting regulatory objectives (in particular, the 30 Jean-Pierre Chauffour and Jean-Christophe Maur costs for trade) should be balanced against the marginal integration requires flexibility and customization in the expected benefits. (Maur and Shepherd, ch. 10 in this vol- way provisions are drafted: ume, discuss this issue in the context of standards.) An important aspect that may justify resisting the • Provisions in the agreement compose only one of the incorporation of regulatory objectives into PTAs is the building blocks of broader cooperation, which may importance of national endowments and preferences, include institutional arrangements, whether hard (e.g., which will differ among countries. Bhagwati (2008) sug- a common institution) or soft (e.g., expert consultations), gests that the rationales for different labor standards apply as well as technical assistance and capacity building. to countries that are at different stages of development and • Gradual implementation is advisable, as reform may have different economic contexts. Although harmoniza- not be carried out overnight and will present unique tion eliminates the costs associated with duplication and challenges in a given country context. complexity, it can undermine national objectives by • Flexibility is necessary for joint projects such as harmo- departing too much from these aims, and in the case of nization work and creation of common regional tools. upward harmonization to stricter regulatory levels, it can • Areas of cooperation will need to be reexamined as reg- raise the costs faced by some countries—often, the poorest ulatory needs change over time. ones. The exportation of (higher) regulatory standards • Deep integration areas will have to be revisited as under- and practices has been flagged in the recent literature as standing of how best to address regulatory dimensions integral to the strategy of the two biggest proponents of evolves. PTAs, the United States and the EU (Maur 2005; Bhagwati • Implementation of policy has to be monitored. 2008; Horn, Mavroidis, and Sapir 2009). A closer examina- • Recourse to venues other than dispute settlement is tion of PTA disciplines suggests that template approaches needed. to PTA liberalization are indeed promoted by both hubs. This is worth highlighting because the EU and U.S. Reliance on a rigid interpretation of an international approaches generally differ quite substantially. In this agreement and on enforcement through dispute settlement volume, the influence of the EU and U.S. hubs is noted, for is insufficient for the deep integration dimensions of liber- instance, for contingent protection rules (Prusa, ch. 9), the alization (Hoekman, ch. 4 in this volume). Deep integra- use of mutual recognition versus equivalence for standards tion requires a combination of hard and soft law and (Maur and Shepherd, ch. 10; Stoler, ch. 11), differing areas enhanced capacity. The reasons for this are (a) practicali- of focus regarding IPR protection (Fink, ch. 18), and the ties, (b) uncertainty, and (c) political economy. diffusion of procurement and investment rules (Miroudot, The first problem is practical and procedural: the ch. 14; Dawar and Evenett, ch. 17). PTAs—and bilateral appropriate implementation of behind-the-border policies agreements in general—offer a means of dealing with requires a set of actions, ranging from enactment of legal heterogeneity of preferences through the principle of provisions to establishment of adequate structures, includ- equivalence, which often takes the form of mutual recog- ing appropriate governance and rules, material and per- nition, as was discussed earlier. sonnel for operationalizing the policies, and reporting mechanisms. All this is highly complex and is difficult to specify in full in an international agreement, and it might Flexibility and Customized Problem Solving as well be left to domestic authorities to work out. The flexibility offered by PTAs in terms of tailoring the The second problem arises from the fact that there is level of ambition of given disciplines to particular trading uncertainty about the most appropriate design for regula- partners is one finding from the early work on the new tory policies and their implementation, since these aspects wave of PTAs (Heydon 2003). Although special and differ- probably depends on country circumstances. A good ential treatment and policy space are important features of example is the variety of competition provision rules and modern trade negotiations involving developing countries, setups in PTAs (Dawar and Holmes, ch. 16 in this volume). here we take the logic farther, arguing that it is not only the Configurations of common competition regimes and pro- nature of partner countries and their capacity that dictate visions are greatly influenced by national regimes and by the need for such flexibility but also the regulatory issues partners’ size and level of development. Another source of themselves. uncertainty is time, because technological changes, for Many of the new policies captured in the latest genera- instance, may fundamentally affect the nature of the goods tion of PTAs do not lend themselves to reduction to stand- and services exchanged and the way markets, market alone legal language in a trade agreement. Rather, deep operators, and government bodies conduct their work (in, Beyond Market Access 31 for example, telecommunication services, standards poli- economic size—has implications for the choice of degree cies, and border controls). of flexibility and informality. In this sense, a flexible Finally, the political economy of PTAs that rely on soft- approach can be seen as a building block for more formal law mechanisms might be more supportive of actual liber- arrangements down the line. Porges (ch. 22 in this volume) alization than is the case with a top-down approach in notes that dispute settlement tends to become more legalis- which rules are rigidly imposed by a powerful trade part- tic when the relationship is symmetric, whereas when it is ner. Ownership can be enhanced through cooperation and asymmetric, political and diplomatic approaches are pre- stakeholder involvement, which would be part of a process ferred. Although the softer structure may be seen as a way of identifying appropriate regulatory solutions for liberal- of affording flexibility to smaller partners and as reflecting ization (Hoekman, ch. 4 in this volume). This is related to the unequal balance of power, it leaves the solving of dis- respect for country preferences while reducing differences putes to less transparent conduits in which power may be with a view toward mutual recognition of regulations. more easily wielded. Thus, flexibility seems to be an important dimension to be considered in deep integration. This recognition gives Implementation rise to a recommendation for “living agreements� that incorporate a work program and for associated institutions The inadequacy of a solely legalistic approach to commit- that establish a pathway allowing deeper integration over ments implies that negotiations will not settle every issue time and the resolution of standards issues. Hoekman and that, in addition to the ex ante work of negotiators, an (ch. 4 in this volume) similarly argues in favor of a con- important ex post agenda awaits countries signing PTAs. It structive rather than adversarial process in North-South can be argued that the implementation agenda is on paper PTAs. A problem with purely adversarial procedures is that more important in the case of PTAs than in the WTO. they tend to leave unaddressed public good issues, whereas There are essentially two reasons for this. The first is that supranational institutions have an incentive to pursue such PTAs commit parties to effective liberalization, whereas the matters. PTAs then become instruments of cooperation, in WTO often merely commits parties to bind only maximum addition to integration, and can provide a problem-solving levels of protection and provides numerous exemptions forum for countries that are undertaking reform and and exceptions for developing countries.21 The second rea- upgrading their regulatory capacities. son is that deep PTAs cover newer and more ambitious Another aspect of the flexible approach toward deep ground than does the WTO. integration is the implication for approaches to dispute In some areas, the track record of PTAs in implementa- settlement. On the one hand, negotiating and implement- tion has been relatively poor. Reviews of services (Mattoo ing deep integration dimensions is costly, creating a motive and Sauvé, ch. 12 in this volume), and of competition pro- for ensuring a return on this investment by the establish- visions (Dawar and Holmes, ch. 16 in this volume) suggest ment of strong dispute settlement mechanisms (Porges, unimpressive results, while in areas that have seen more ch. 22 in this volume). This may explain the observed trend pressure toward implementation, such as IPRs, the evi- toward more legalistic forms of dispute resolution, replac- dence shows much more substantial changes (Biadgleng ing the more diplomatic approaches of older agreements. and Maur forthcoming). The different treatment for IPRs In this regard, WTO-like ad hoc panels (which, among is the direct result of the greater prominence of implemen- other things, permit recourse to the expertise of specialists) tation and enforcement in recent PTAs involving the are often preferred. On the other hand, dispute settlement United States and Europe. is only one of the several mechanisms in PTAs contribut- Dealing with deep integration issues requires prepared- ing to enforcement. Panel-type disputes only occur in ness that goes well beyond the negotiation stage, and most exceptional cases, and smaller disagreements are resolved likely it entails the dedication of some permanent resources through other channels established in the PTAs. The latter to managing the agreement. The resource and policy impli- approach is what can be described as soft law. In this cations of deep integration agreements are likely to be, in respect, common institutions play an important role, part, unforeseen, as Hoekman (ch. 4 in this volume) sug- allowing technical and ad hoc approaches to solving what gests, and the problems may be compounded by countries’ are often complex issues and facilitating the involvement of lack of preparedness. Examples of possible unintended third parties, such as the private sector. consequences of commitments include incompatibilities As has already been indicated, the extent to which a PTA between PTA commitments and the existing (domestic is symmetrical or not—that is, the degree to which the and international) legal environment;22 political-economy partners are equal or similar in level of development or constraints, where commitments are not accepted by 32 Jean-Pierre Chauffour and Jean-Christophe Maur domestic constituencies, including legislators who may and which of the hard-law–soft-law approaches, or combi- have to vote on new laws; limitations on regulatory free- nations thereof, contribute most to liberalization. The dom; inefficiency in implementing the new regulatory inference from the above discussion is that although many environment; and economic implications that are less examples and many distinctive approaches to integration beneficial than initially thought. now exist, there is still relatively little basis for recommen- Beyond the principle of liberalization agreed in the PTA, dations on how to appropriately implement deep integra- a work program of implementation must be devised in tion provisions in PTAs, beyond a core set of principles. order to make liberalization a reality. The implementation For developing countries, one attraction of PTAs with of the provisions of an agreement may require different lev- more developed partners, at least in theory, is the prospect els of intervention. First, institutional changes may be nec- of access to capacity building and transfer of resources. Yet essary because the implementation of new areas of policy whether and how development assistance contributes to may call for the establishment of new regulatory agencies or implementation is generally difficult to assess, and is even the reorganization of existing institutions. Second, new laws more so in the context of PTAs, given the naturally non- for regulatory reform will be required to reflect PTA com- transparent nature of institutions. mitments. The need for such legislation will vary, depend- An important issue that appears prominently in the ing on the legal standing of international commitments in discussion on competition policy, government procure- domestic law: some commitments require translation into ment, and standards is to what extent the process of imple- domestic law, while others have direct effect, and some law mentation should be run from the center. For instance, systems rely more on a case law approach. The third dimen- competition regimes in PTA contexts range from regional sion of implementation consists of the administrative, pro- institutions to national institutions that cooperate on cedural, and operational changes required to comply with international issues. Several considerations affect whether the new regulatory framework. This can include the man- the implementation process will be left to national gov- agement of the agreement itself, including transparency and ernment or devolved to a transnational body (Dawar and monitoring requirements. Finally, enforcement of the newly Evenett, ch. 17 in this volume). adopted regulations needs to be considered; the requisite The first, obvious point is that a prerequisite for com- staff and resources will have to be allocated. Attention mon institutions or rules is the willingness of trade part- should be given to the quality of enforcement and to meas- ners to abandon some of their sovereignty. When this does ures for assessing the effectiveness of the application of the not happen (often, in the context of North-South agree- laws (Biadgleng and Maur forthcoming). ments), only “lighter� options remain, and only core prin- The text of the PTA is only one initial element of the ciples guaranteeing good policy and governance can be process of integration. Implementation issues must also be agreed. This is the solution chosen for procurement provi- carefully examined to determine whether liberalization is sions. A related concern is the choice between maintaining effective. Monitoring and accountability matter. This is a national preferences and adopting international stan- more complex process than verifying that trade barriers are dards. In this area, the answer to the question of harmo- effectively dismantled, and the necessary information is nization versus recognition depends on whether the often not readily available. In general, implementation in benefits of harmonization outweigh the costs of loss of PTAs is not a very transparent process, and sustained atten- preferences. National ability to issue regulations is also a tion to implementation is rare. This is an additional reason parameter to be taken into account. In particular, one for the constructive cooperative approach recommended motivation for preserving diversity in regulation and by Hoekman (ch. 4 in this volume), given the complemen- implementation is to derive the benefits from competition tarity between information generation and exchange and between different regulatory solutions and procedures. the process for discovery of the best trade-facilitating regu- Messerlin and Zarrouk (2000) advocate less centralization latory solution. in the context of conformity assessment in order to pro- In spite of some evidence that monitoring is taking mote competition between different conformity bodies place, information about implementation remains scarce, and their services (e.g., testing, surveillance, inspection, and most of the analysis of PTAs rests on the evidence pro- certification, etc.). vided by the agreements themselves and on some measures A second point is the degree of coordination that is of outcomes such as trade flows.23 Although such analysis required by the integration policy. For some transnational provides useful insights, the policy recommendations that public goods, common institutions and a top-down can be drawn from it are limited. Little is known about approach may be preferable in order to solve the coordina- which liberalization strategies work best as agents of change tion problems that lead to inadequate supply of the goods. Beyond Market Access 33 Beyond public goods per se, common institutions appear already been clearly articulated, and a call for multilateral- particularly necessary when frequent interactions, decision izing regionalism has been voiced in the WTO (Baldwin making, adjudication, and exchange of information are and Low 2009). The concern that PTAs “compete� with needed, as in customs unions. (See Andriamananjara, ch. 5 multilateral negotiations for the attention of negotiators in this volume.) A related dimension is the desire to and represent an untidy way of proceeding to liberalization achieve certain scale or efficiency effects. Competition pol- remains legitimate. icy illustrates the possibility of opting for a common, cen- Yet, an important emerging lesson is that the economic tralized competition regime—as in the cases of the EU, the paradigm of shallow PTAs does not necessarily apply to Caribbean Community (CARICOM), COMESA, and the deep and comprehensive PTAs. “Old� concepts such as Andean Pact—when parties to a PTA do not differ much mercantilist reciprocal liberalization, trade creation and in their preferences regarding the type of competition diversion, or a textual approach to negotiating PTAs may enforcement. still underpin the reasoning of many policy makers but are For developing countries in particular, a top-down often archaic or incomplete for deep integration liberaliza- approach may prove attractive. A first incentive might be to tion. Failure to understand the new paradigm of preferen- improve governance. By decentralizing decision making tial integration may, in turn, explain why most PTAs either within a PTA, member countries may be better able to have not fully exploited the liberalization opportunities of shield policies from the risk of reversal and guarantee their behind-the-border measures or have not prioritized the independence. This strategy may be helpful, for example, opportunities closest to the parties’ interests—although it for competition policy, where it might be difficult in some might be naïve to attribute lack of progress solely to a lack countries to escape the influence of particularly large firms of understanding. and to discipline them. More generally, as discussed earlier, 2. Deep integration PTAs are potentially powerful tools for countries with weak judicial systems might find it advanta- pushing wide-ranging government-owned reforms. Beyond geous to rely on supranational judicial institutions as a way market access, deep integration PTAs create opportunities of anchoring or locking in policies. Poor administrative to complement trade liberalization with other behind-the- resources at the country level could also motivate the pool- border reforms. In addition, they offer unique instruments ing of resources among a group of countries—something for promoting bilateral or plurilateral cooperation and small island countries have done. resource transfers, transparency mechanisms, mutual Finally, a third consideration is efficiency motives, equivalence, informal mechanisms for dispute resolution, which may lead some countries to seek to replace their in-depth and expert dialogue, and deeper liberalization existing regulations with superior systems “imported� among the willing parties. These are not approaches that from partner countries and, through the institutional can be easily, if at all, replicated in the large and formal set- mechanisms of the PTA, to obtain access to the superior ting of multilateral institutions. expertise and systems of the partner country. Yet PTAs are worthwhile only if governments are them- selves committed to reform and liberalization. PTAs offer a variety of mechanisms by which the process of reform will Conclusions become more effectively and irremediably set in motion, Modern PTAs are evolving rapidly. They are increasingly but a prerequisite is that meaningful commitments be deep, and they affect all countries and regions of the world, agreed to in the first place. Deep integration PTAs should including the most remote quarters. As academics and pol- therefore strive to provide open access to regulatory rules icy makers try to deal with this new generation of PTAs, and disciplines in order to ensure equality of treatment of four tentative conclusions can be suggested regarding this all members and nonmembers and so minimize the occur- changing landscape. rence of “regulatory preferences.� This means—beyond 1. Deep integration introduces a change of paradigm. To national treatment—liberal rules of origin, transparency, be sure, PTAs still have to do with preferences, discrimina- and the availability of due process. Good regulatory practice tion, and exclusion. They may lead to suboptimal out- should lead de jure preferential liberalization to become, in comes and could complicate and even undermine progress effect, MFN liberalization. The question of discrimination toward a more open, rules-based, and nondiscriminatory in deep PTAs is likely to remain convoluted. Discriminatory multilateral trading system. More worrying than discrimi- regulations could take many forms, whether codified in nation, perhaps, is the additional inherent complexity that rules or not. De facto preferences can arise from rules that is created by the overlapping and conflicting regulatory look nonpreferential on paper, or from preferential enforce- regimes promoted by myriads of PTAs. This concern has ment. Furthermore, discrimination may, paradoxically, 34 Jean-Pierre Chauffour and Jean-Christophe Maur be the only form of acceptable liberalization, as parties norms, and to trigger competitive liberalization effects in mutually agree to accept each other’s rules. partner countries. A related logic can be observed in the EU, 3. Deep integration should be pursued strategically and which, as Maur (2005) notes, is attempting to leverage its selectively. Another answer to complexity that is, in our PTAs to shape South-South agreements in its recent wave of view, not sufficiently considered by developing countries, is agreements with the Mediterranean countries, the Balkans, selectivity. (This is also suggested by Hoekman and Sekkat and the African, Caribbean, and Pacific (ACP) countries. 2010.) Liberalization is a complex matter, not only from a capacity standpoint but also from a political one. The Notes political economy of deep integration involves many (often The authors thank Jaime de Melo, Ndiame Diop, Bernard Hoekman, and opposing) interests and a large set of potential stakehold- Richard Newfarmer for helpful comments and suggestions. ers. Overloading the negotiating agenda (which will later 1. In a WTO working paper, Fiorentino, Verdeja, and Toqueboeuf become the implementing agenda) diverts the focus from (2006) note that country negotiating resources are being shifted away from multilateral negotiations toward negotiations on PTAs. what may be achievable and from the areas where gains 2. The General Agreement on Tariffs and Trade (GATT) architecture may be the most important. Agreements bloated by too was historically built, essentially, around the notion of negative integra- many issues may lose significance and fail to achieve much. tion and the prohibition of the most detrimental policies through elimi- Picking meaningful issues with the right partner, along nation of border trade barriers and the espousal of nondiscrimination principles (Hoekman and Kostecki 2009). Recently, however, new forms with having adequate technical assistance and a coopera- of economic integration have been included in the multilateral trade tive approach, may bring about substantial progress on lib- framework, starting with the Kennedy Round and the General Agreement eralization and serve as a positive signal or trigger for more on Trade in Services (GATS) in 1979 and continuing with the WTO (and the incorporation of TRIPS, in particular). The WTO incorporates much challenging areas. Market access should not be the only more significant elements of positive integration than previously. PTAs item on the agenda of negotiators, especially those of follow the same trend and go even further in many instances. developing countries, since deep integration really entails 3. Torrent (2007) provides the following example: “The European Community directives on the liberalization of movements of capital seem thinking about a domestic reform strategy. Prioritization to be a clear example of ‘negative’ integration, but they were enacted of core objectives and sequencing should be central con- according to what, in Tinbergen’s terms, would be a clear example of ‘pos- cerns of negotiators. Sound regulatory practice should itive integration’ (and they would be defended in this way by many in the underpin liberalization to minimize the prevalence of reg- European Commission). In political terms, NAFTA’s [North American Free Trade Agreement’s] Chapter XI on investments would be looked at ulatory preferences and ensure the overall consistency of by many around the world as a typical example of ‘negative integration’ liberalization and regulatory objectives. that sharply reduces the capacity of Governments to intervene in the 4. There is no one-size-fits-all model of deep integration. economy. It is also an example of ‘positive integration’ that creates com- mon rules that go beyond the liberalization of access (for example on As policy makers more and more integrate these new protection of investments).� dimensions, we can expect that they will use PTAs more 4. According to Ethier (1998) and, subsequently, Levy (2009), attract- intensively to further liberalization objectives—it may be ing foreign direct investment (FDI) is one of the main incentives for entering into PTAs. hoped, in a way complementary to multilateral efforts. 5. Technology allows services to be traded under several modal Liberalization in any sector is not a simple matter. It forms; for instance, medical diagnostics can now be provided at a dis- escapes easy characterization, as well as one-size-fits-all tance, thanks to electronic imagery. types of answers. This complexity means that there are few 6. Note that well-being can be understood in broad terms as includ- ing not only economic welfare but also value-related preferences. universal rules to follow; rather, carefully designed and spe- 7. In economics, nonrivalry means that consumption of the good by cific solutions are needed. Deep integration is essentially a one individual does not reduce the availability of the good for consump- sui generis process, as is illustrated by Winters (2010) in the tion by others. Nonexcludability means that no one can be effectively excluded from using the good. case of the EU. Yet complexity means that some core prin- 8. According to Levy (2009), several mechanisms could be in play: ciples should be followed in order to promote, to the extent (a) once a bureaucracy commits to good governance in a trade agreement, possible, market-based solutions. it may make little sense to maintain a different attitude for the domestic market; (b) good governance with respect to international flows could The dynamics of North-South, South-South, and North- serve as a signal, spurring reform on the domestic front; and (c) if the rule North PTAs differ considerably. Asymmetric agreements of law is not followed within the country, there might be costly and make cooperation less easy and may provide less scope for adverse reputational spillovers that could affect the decisions of foreign transnational public goods and mutual recognition but investors and traders. 9. The decision to focus on coal and steel came about not only offer more prospects for lock-in and greater access to because of the economic importance of the two sectors but also because imported regulatory regimes, when needed. Market access these materials were considered the main inputs for making weapons. considerations will be paramount for the small partner, “The pooling of coal and steel production should immediately provide for the setting up of common foundations for economic development as whereas the larger partner will seek, beyond market a first step in the federation of Europe, and will change the destinies of access, to diffuse its regulatory norms, including values those regions which have long been devoted to the manufacture of Beyond Market Access 35 munitions of war, of which they have been the most constant victims� Labor and Environmental Issues.� Journal of Economic Perspectives (Robert Schuman, Declaration of 9 May 1950). 15 (3): 69–88. 10. “The EU can make an important contribution by working around Baldwin, Richard E. 1994. “Towards an Integrated Europe.� Centre the conflict issues, promoting similar reforms on both sides of the bound- for Economic Policy Research, London. http://heiwww.unige.ch/ ary lines, to foster convergence between political, economic and legal sys- ~baldwin/. tems, enabling greater social inclusion and contributing to confidence ———. 2008. “Sequencing and Depth of Regional Economic Integration: building� (European Commission 2007). Lessons for the Americas from Europe.� World Economy 31 (1): 5–30. 11. World Bank (2005), however, notes that some wars began partly Baldwin Richard, Simon Evenett, and Patrick Low. 2009. “Beyond Tariffs: as trade disputes; examples are the U.S. Civil War (1861–65) and the Multilateralizing Non-Tariff RTA Commitments.� In Multilateralizing Soccer War of 1969 between El Salvador and Honduras. Regionalism: Challenges for the Global Trading System, ed. Richard 12. The Electricity Regulatory Forum, or Florence Forum, was set up Baldwin and Patrick Low, ch. 3. Cambridge, U.K.: Cambridge to discuss the creation of a true internal electricity market in the EU. Since University Press. 1998, the forum has met once or twice a year, formerly in Florence and Baldwin, Richard, and Patrick Low. 2009. Multilateralizing Regionalism: now in Rome. Further information on the Florence Forum is available on Challenges for the Global Trading System. Cambridge, U.K.: Cambridge the Web page of the European Commission’s Directorate General for University Press. Energy; see http://ec.europa.eu/energy/gas_electricity/forum_electricity_ Bergsten, Fred. 1996. “Globalizing Free Trade.� Foreign Affairs 75 (3): florence_en.htm. A similar body, the Madrid Forum, has been established 105–20. for natural gas markets. Bhagwati, Jagdish. 2008. Termites in the Trading System: How Preferential 13. Thus, regulation does not necessarily generate positive terms-of- Agreements Undermine Free Trade. Oxford, U.K.: Oxford University trade effects for the home country (e.g., customs duties revenues, in the Press. case of tariff protection). Biadgleng, Ermias Tekeste, and Jean-Christophe Maur. Forthcoming. 14. In most cases, the situation is not that clear-cut. Regulatory “The Influence of Preferential Trade Agreements on Implementation burdens often create additional jobs for administrations and provide of Intellectual Property Rights in Developing countries: A First Look.� opportunities for graft. Issue Paper (forthcoming), ICTSD Programme on Innovation, Tech- 15. In normal circumstances, there should be no prior rent capture by nology, and Intellectual Property, International Centre for Trade and domestic interests, since the objective of regulatory controls is not to raise Sustainable Development, Geneva, Switzerland. revenue or afford protection. Carrère, Céline, Jaime de Melo, and Bormolaa Tumurchudur. 2010. “Dis- 16. The distinction is often blurred, and when identification methods entangling Market Access Effects of Preferential Trading Arrange- are imperfect, origin is often used as a very imperfect proxy for other ments with an Application for ASEAN Members under an ASEAN-EU characteristics (as is country of citizenship for migrants). FTA.� World Economy 33 (1): 42–59. 17. This claim was tested empirically by Czubala, Shepherd, and Czubala, Witold, Ben Shepherd, and John S. Wilson. 2009. “Help or Hin- Wilson (2009) in the context of adoption of international standards. drance? The Impact of Harmonized Standards on African Exports.� 18. Baldwin, Evenett, and Low (2009) cite the example of ROOs in Journal of African Economies 18 (5): 711–44. East Asian PTAs as particularly liberal, since they only require incorpora- Devlin, Robert, and Antoni Estevadeordal. 2006. “Trade and Coopera- tion under the laws of the trade partner and do not impose any other tion: A Regional Public Goods Approach.� In Multilateral and nationality requirements on the entity, regarding, for example, the nation- Regional Frameworks for Globalization: WTO and Free Trade ality of the people controlling the firm. Agreements, ed. Wonhyuk Lim and Ramon Torrent. Seoul: Korea 19. Although separability never really existed in practice, the tradi- Development Institute. tional approach was to consider trade policy in relative isolation from Estevadeordal, Antoni, Kati Suominen, and Robert Teh. 2009. Regional other policies, including other economic policies. Rules in the Global Trading System. Cambridge, U.K.: Cambridge 20. Environmental externalities could also stem from purely domes- University Press. tic economic activities. Ethier, Wilfred J. 1998. “The New Regionalism.� Economic Journal 108 21. In terms of liberalization, PTAs will at least lock in the status quo if (449): 1149–61. disciplines in an area are included. This is furthered by the use of negative- European Commission. 2007. A Strong Neighbourhood Policy. Communi- list approaches, as in some services provisions. Thus, the value of commit- cation COM(2007)774 final. Brussels: European Commission. ments in PTAs is higher than in the WTO. In this sense PTAs are more Evans, David, Peter Holmes, Leonardo Iacovone, and Sherman Robinson. rigid. This may be a reason developed countries have intensively used PTAs 2006. “Deep Integration and New Regionalism.� In Assessing Regional with smaller developing countries as a way to lock in liberalization in a way Trade Agreements with Developing Countries: Shallow and Deep Inte- that was not necessarily happening in the WTO. But with stricter commit- gration, Trade, Productivity, and Economic Performance, ch. 2. Report ments looming, resistance to liberalization may also be stronger. for DFID Project Number 04 5881, University of Sussex, U.K. 22. Biadgleng and Maur (2010) cite the example of commitments Evenett, Simon, and Michael Meier. 2008. “An Interim Assessment of made by the Arab Republic of Egypt with the EU to ratify, in contradic- the US Trade Policy of ‘Competitive Liberalization.’� World Economy tion to its own law, the Convention for the Protection of New Varieties of 31 (1): 31–66. Plants established by the International Union for the Protection of New Fink, Carsten, and Martin Molinuevo. 2007. “East Asian Free Trade Agree- Varieties of Plants (UPoV). ments in Services: Roaring Tigers or Timid Pandas?� World Bank, 23. Examples of different approaches to monitoring implementation Washington, D.C. are seen in the United States, which “certifies� implementation of FTAs Fiorentino, Roberto V., Luis Verdeja, and Christelle Toqueboeuf. 2006. before congressional approval; the EU, which reports regularly in the case “The Changing Landscape of Regional Trade Agreements: 2006 of accession and in more ad hoc fashion with other partners on implemen- Update.� WTO Discussion Paper 12, World Trade Organization, tation; APEC implementation action plans; and South-South agreements Geneva. such as COMESA, for which the secretariat monitors country progress. Haftel, Yoram Z. 2007. “Designing for Peace: Regional Integration Arrangements, Institutional Variation, and Militarized Interstate Disputes.� International Organization 61 (1): 217–37. References Harrison, Glenn W., Thomas Rutherford, and David Tarr. 2002. “Trade Bagwell, Kyle, and Robert Staiger. 2001. “The WTO as a Mechanism for Policy Options for Chile: The Importance of Market Access.� World Securing Market Access Property Rights: Implications for Global Bank Economic Review 16 (1): 49–79. 36 Jean-Pierre Chauffour and Jean-Christophe Maur Heydon, Ken. 2003. “Key Findings. Regionalism: A Complement, Not a Martin, Philippe, Thierry Mayer, and Mathias Thoenig. 2010. “The Geog- Substitute.� In Regionalism and the Multilateral Trading System. Paris: raphy of Conflicts and Free Trade Agreements.� CEPR Discussion OECD Publishing. Paper 7740, Centre for Economic and Policy Research, London. Hoekman, Bernard, Aaditya Mattoo, and André Sapir. 2007. “The Political Maur, Jean-Christophe. 2005. “Exporting Europe’s Trade Policy.� World Economy of Services Trade Liberalization: A Case for International Economy 28 (11): 1565–90. Regulatory Cooperation?� CEPR Discussion Paper 6457, Centre for ———. 2008. “Regionalism and Trade Facilitation: A Primer.� Journal of Economic and Policy Research, London. World Trade 42 (6): 979–1012. Hoekman, Bernard, and Michel Kostecki. 2009. The Political Economy of the Messerlin, Patrick, and Jamel Zarrouk. 2000. “Trade Facilitation: Technical World Trading System, 3rd ed. Oxford, U.K.: Oxford University Press. Regulations and Customs Procedures.� World Economy 23 (4): 577–93. Hoekman, Bernard, and Khalid Sekkat. 2010. “Arab Economic Integra- OECD (Organisation for Economic Co-operation and Development). tion: The Missing Links.� CEPR Discussion Paper 7807, Centre for 2003. Regionalism and the Multilateral Trading System. Paris: OECD Economic and Policy Research, London. Publishing. Horn, Henrik, Petros C. Mavroidis, and André Sapir. 2010. “Beyond the Ortino, Federico. 2004. Basic Legal Instruments for the Liberalisation of WTO? An Anatomy of EU and US Preferential Trade Agreements.� Trade: A Comparative Analysis of EC and WTO Law. Oxford, U.K.: The World Economy 33(11), 1565–88. Hart. Hornbeck, J. F. 2003. The U.S.–Central America Free Trade Agreement Pelkmans, Jacques. 1984. Market Integration in the European Community. (CAFTA): Challenges for Sub-Regional Integration. Report for the U.S. Studies in Industrial Organization, vol. 5. The Hague: Martinus Nijhoff. Congress, RL31870. Washington, DC: Government Printing Office. Schiff, Maurice, and L. Alan Winters. 1998. “Regional Integration as Krishna, Pravin. 1998. “Regionalism and Multilateralism: A Political Diplomacy.� World Bank Economic Review 12: 271–95. Economy Approach.� Quarterly Journal of Economics 113 (1): 227–50. ———. 2002. “Regional Cooperation, and the Role of International Krugman, Paul. 1993. “Regionalism versus Multilateralism: Analytic Organizations and Regional Integration.� Policy Research Working Notes.� In New Dimensions in Regional Integration, ed. Jaime de Melo Paper 2872, World Bank, Washington. DC. and Arvind Panagariya. Cambridge, U.K.: Cambridge University Press. ———. 2004. Regional Integration and Development. Oxford, U.K.: Lee, Jong-Wha, and Ju Hyun Pyun. 2009. “Does Trade Integration Con- Oxford University Press. tribute to Peace?� Asian Development Bank Working Paper on Schott, Jeffrey J. 2003. “Assessing US FTA Policy.� In Free Trade Agreements: Regional Integration 24, Asian Development Bank, Manila. US Strategies and Priorities, ed. Jeffrey Schott, ch. 13. Washington, DC: Lesser, Caroline. 2007. “Do Bilateral and Regional Approaches for Reduc- Institute for International Economics. ing Technical Barriers to Trade Converge towards the Multilateral Tinbergen, Jan. 1954. International Economic Integration. Amsterdam: Trading System?� OECD Trade Policy Working Paper 58, Organisation Elsevier. for Economic Co-operation and Development, Paris. Torrent, Ramon. 2007. “The Legal Toolbox for Regional Integration: A Levy, Philip I. 2009. “The United States–Peru Free Trade Agreement: What Legal Analysis from an Interdisciplinary Perspective.� Prepared for the Did You Expect?� American Enterprise Institute Working Paper Series Euro-Latin Study Network on Integration and Trade (ELSNIT) Con- on Development Policy 1, American Enterprise Institute, Washington, ference, Barcelona, October 26–27. DC. Winters, L. Alan. 1997. “What Can European Experience Teach Develop- Limão, Nuno. 2007. “Are Preferential Trade Agreements with Non-trade ing Countries About Integration?� World Economy 20 (7): 889–911. Objectives a Stumbling Block for Multilateral Liberalization?� Review ———. 2010. “Europe Is Sui Generis—But It Provides Lessons for of Economic Studies 74 (3): 821–55. Developing Countries.� Prepared for the Annual Conference of the Mansfield, Edward D., and Jon C. Pevehouse. 2000. “Trade Blocs, Trade Global Development Network, Prague, January 16–18. Flows, and International Conflict.� International Organization 54 (4): World Bank. 2005. Global Economic Prospects: Trade, Regionalism, and 775–808. Development. Washington, DC: World Bank. 2 LANDSCAPE Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Preferential trade agreements (PTAs) are an enduring developments. Unless otherwise stated, the data presented feature of the contemporary multilateral trading system. take account of all bilateral, regional, and plurilateral trade Sixty years after the founding of the General Agreement on agreements of a preferential reciprocal nature that have Tariffs and Trade (GATT), the global trading landscape has been notified to the GATT/WTO (see box 2.1 for defini- changed beyond recognition. Membership in the World tions). The focus is on free trade agreements (FTAs), cus- Trade Organization (WTO) continues to grow steadily, but toms unions (CUs), partial-scope agreements in the area of meanwhile, participation in PTAs is expanding at an trade in goods, and economic integration agreements unprecedented rate. Slow progress in the Doha Round (EIAs) in the area of trade in services.1 trade negotiations has no doubt contributed to this growth trend. As of February 2010, 266 PTAs were in force, and Trends among PTAs this figure did not include a significant number of agree- ments (mostly among developing countries) that had not Recent developments and trends are shaping a PTA land- yet been notified to the WTO, or the many PTAs still in the scape that presents a number of significant features. pipeline. As a consequence, a growing proportion of world The first is ubiquity. PTA participation is becoming trade is, or has the potential to be, conducted under prefer- more diverse, spreading to most geographic regions, but ential terms rather than under the nondiscriminatory especially to East Asia and the Pacific. North-South prefer- regime of the WTO. In addition, trade conducted under ential partnerships are on the rise, with a number of devel- preferential rules is increasingly subject to a plethora of oping countries electing to forgo unilateral programs— crisscrossing regulatory regimes that modify and compli- such as the generalized system of preferences (GSP) or cate the WTO’s multilateral regulatory regime. trading arrangements previously conducted under a WTO All but one of the WTO’s 153 members is a party to at waiver—in favor of reciprocal agreements. These develop- least one PTA (Mongolia is the exception), and most ments are testing the negotiating capacity of developing countries are parties to several. Today’s PTAs are charac- countries, particularly where they have to deal with issues terized by diverse geographic and physical configurations for which no multilateral rules currently exist or where the and differing regulatory content. Although PTAs offer the negotiating framework differs from that of the WTO. Such potential for increased trade and investment among their is the case, for instance, in PTAs covering trade in services. members through enhanced market access, they do so at The second is consolidation. Bilateral relationships are the cost of introducing multiple layers of complexity into being replaced by plurilateral PTAs among the same part- the global trading landscape, rendering trade relations less ners, and agreements between regional blocs are on the transparent and more unpredictable. increase. In Asia, countries long resistant to preferential The aims of this chapter are (a) to provide a snapshot trade liberalization are catching up, and plurilateral PTAs (as of February 2010) of recent developments and trends coexist alongside bilateral PTAs among the same sets of with respect to the number and scope of PTAs, (b) to ana- partners. Notwithstanding some consolidation, the grow- lyze the types of PTA initiatives that are currently under ing number of overlapping plurilateral PTAs—particularly negotiation in each geographic region, and (c) to assess the in Africa and Central Asia but also, increasingly, in the possible impact of a selection of plurilateral PTAs on trade Americas and Asia—points to a further fragmentation of 37 38 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Box 2.1. Typology of Preferential Trade Agreements Care should be taken when categorizing preferential trade agreements (PTAs), given the differences in terminology used by institutions and researchers. In this study, we use the generic term PTA to refer to all reciprocal preferential agreements. The World Trade Organization (WTO), however, uses the term regional trade agreements (RTA) for all reciprocal preferential agreements and reserves PTA for nonreciprocal preferential agreements such as the generalized system of preferences (GSP) and the African Growth and Opportunity Act (AGOA). The terminology employed in this chapter is explained below. Free trade agreement (FTA). An agreement between two or more parties in which tariffs and other trade barriers are eliminated on most or all trade. Each party maintains its own tariff structure relative to third parties. Examples are the North American Free Trade Agreement (NAFTA) and the Japan–Singapore New-Age Economic Partnership Agreement. Customs union (CU). An agreement between two or more parties in which tariffs and other trade barriers are eliminated on most or all trade. In addition, the parties adopt a common commercial policy toward third parties that includes the establishment of a common external tariff. Thus, products entering the customs union from third parties face the same tariff regardless of the country of entry. Examples are the Southern Cone Common Market (Mercosur, Mercado Común del Sur) and the agreement between the European Union (EU) and Turkey. Partial-scope agreement. An agreement between two or more parties that offer each other concessions on a selected number of products or sectors. Examples are the Asia-Pacific Trade Agreement (APTA) and the agreement between the Lao People’s Democratic Republic and Thailand. Economic integration agreement (EIA). An agreement covering trade in services through which two or more parties offer preferential market access to each other. Examples are the U.S.–Peru and Thailand–Australia PTAs. Typically, services provisions are contained in a single PTA that also covers goods. An EIA may be negotiated some time after the agreement covering goods; for example, the Caribbean Community (CARICOM) and the European Free Trade Association (EFTA) have negotiated separate services protocols. Preferential trade agreement (PTA). The generic term used in this study to denote all forms of reciprocal preferential trade agreements, including bilateral and plurilateral agreements. The figure shows the breakdown of PTAs covering trade in goods notified to the WTO and in force as of February 2010. FTAs are by far the most common type, accounting for 83 percent of all PTAs. Customs unions, a deeper form of integration, require significant policy coordination between their parties. They are more time consuming to negotiate, are less common, and make up only 10 percent of all PTAs. Partial-scope agreements account for the remaining 7 percent. Types of PTAs Notified to the WTO 7% 10% 83% free trade agreement customs unions partial-scope agreement Source: WTO Secretariat. trading relations and to complications for traders, a shift away from viewing the agreements as a means of exporters, and customs authorities alike. forging traditional regional partnerships among several The third is the changing structural configuration of geographically proximate countries and toward employing PTAs. Bilateral PTAs are increasingly becoming the norm. them, instead, as instruments for negotiating strategic, Such PTAs are concluded more quickly than those involving bilateral market access, often among countries in different multiple partners. More significantly, they are indicative of regions. Indeed, cross-regional PTAs account for two-thirds Landscape 39 of those currently under negotiation. Bilateral partnerships a manifestation of motivations that may not be addressed have the potential to generate further fragmentation of in multilateral global economic integration efforts (for global trading rules because each PTA maintains its own example, geopolitical concerns, the diffusion of social pref- distinct regulatory framework. Initiatives to alleviate frag- erences, and the establishment of regional public goods), mentation by harmonizing preferential rules of origin are the successful conclusion of the Doha Round trade little in evidence outside the pan-European system of negotiations—despite the inherent preference erosion that cumulation of origin. will result—may not be sufficient to diminish the appeal The fourth is the broadening and deepening of the regu- of these agreements. Figures 2.1–2.4 trace the chronologi- latory scope of PTAs. Increasingly, PTAs include a services cal development of PTAs within the WTO framework. component, in addition to the traditional exchange of pref- Figure 2.1 shows the number of PTA notifications erences on goods. On issues that fall under the current received by the GATT/WTO each year between 1948 and mandate of the WTO, some countries have elected to February 2010. In the five-year period 2000–04, 15 PTAs undertake bilateral commitments going beyond those were notified annually, on average. In 2005–07, notifica- they have accepted at the multilateral level (WTO+ provi- tions declined from the 2004 level, to an average of 24 per sions) and some are undertaking commitments on issues year. In 2009, 37 notifications were received, 20 covering that lie outside the current WTO mandate (WTO-extra trade in goods, and 17 covering trade in services. This was provisions). the highest number of notifications received in a single The fifth is the impact of PTAs. As discussed later in this year. chapter, trade flow data indicate that for a number of pluri- The growth in PTA notifications should be interpreted lateral PTAs, intra-PTA imports have increased as a share with caution because it reflects accession commitments of total imports, and growth in intra-PTA exports is associ- made by WTO members.2 Following its accession to the ated with growth in total exports. PTA partners in selected WTO in 2008, Ukraine notified 10 PTAs, some of which plurilateral PTAs trade more internally than would be had been in force for 10 years or more. Given that many of expected in the absence of a PTA, and the impact on extra- the countries in the WTO accession process (for example, PTA exports and imports is largely positive. the Russian Federation and the other successor states to the Soviet Union) are active PTA players, future accessions to the WTO will lead to periodic spurts in PTA notifications. The PTA Kaleidoscope In addition, recent efforts by the WTO membership to This section expands and updates an earlier study con- encourage notification of PTAs already in force but not yet ducted by the WTO in 2008 (Fiorentino, Crawford, and notified appears to be producing results, because several Toqueboeuf 2009). As in that study, we map PTA prolif- PTAs that had been in force for some time were notified in eration and examine the trends and characteristics of the course of 2009.3 PTAs with respect to their type, physical composition, Figure 2.2 shows the total number of PTAs notified to scope, and geographic spread. The focus is on PTAs noti- the GATT/WTO according to the year in which they fied to the WTO and in force, and on those currently entered into force or became inactive.4 As of February being negotiated. 2010, 457 PTAs had been notified, of which 266 (including accessions to existing agreements) are currently in force; of these, 191 are in the area of goods and 75 in services. The Quantifying and Qualifying the Proliferation of PTAs two significant dips in the cumulative active number of PTAs continue to be a prominent feature of most coun- PTAs shown in the figure are a result of the consolidations tries’ commercial policy, and we expect the current sharp of PTA networks in the European region following the upward trend in the number of new PTAs to continue for enlargements of the European Union in 2004 and 2007 and the foreseeable future. Although the multilateral tariff among Balkan countries in the enlarged Central European reductions that would accompany successful completion Free Trade Agreement (CEFTA). These periodic consolida- of the Doha Round of trade negotiations may dull coun- tions, while reducing the total number of active PTAs, are tries’ appetite for the negotiation of further PTAs in the not indicative of a decrease in the amount of trade that is medium term, we think it is likely that the number of subject to preferences; preferential trade continues to be PTAs will continue to increase in the short term as those conducted among the countries concerned, but under a PTAs already signed or under negotiation enter into force different relationship or configuration. A similar process of and those further down the pipeline, in the proposal or consolidation is expected to take place in Central America, study phase, come on line. Also, to the extent that PTAs are where bilateral agreements being concluded between 40 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Figure 2.1. Total PTA Notifications Received by the World Trade Organization, by Year, 1949–2009 40 Transparency 160 Mechanism 37 35 35 140 WTO 31 30 29 120 26 cumulative number number per year 25 100 22 22 21 20 19 19 80 18 17 16 15 15 14 14 60 10 9 40 8 77 6 66 5 5 4 4 20 2 3 3 3 3 2 2 2 2 2 2 2 1 1 11 1 1 1 111 1 1 0 0 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 01 03 05 07 09 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 total notifications per year cumulative number of GATT/WTO members Source: WTO Secretariat. Note: GATT, General Agreement on Tariffs and Trade; PTA, preferential trade agreement; WTO, World Trade Organization. Figure 2.2. All PTAs Notified to the GATT/WTO, by Year of Entry into Force, 1949–2009 105 500 100 95 450 90 85 400 80 75 350 cumulative number 70 number per year 65 300 60 55 250 50 45 200 40 35 30 150 25 20 100 15 10 50 5 0 0 19 9 19 1 19 3 19 5 57 59 19 1 19 3 65 67 19 9 19 1 73 75 19 7 19 9 19 1 83 19 5 19 7 89 91 93 19 5 19 7 99 20 1 03 20 5 07 09 4 5 5 5 6 6 6 7 7 7 8 8 8 9 9 0 0 19 19 19 19 19 19 19 19 19 19 19 20 20 20 notified PTAs (goods, services, and accessions) inactive PTAs cumulative PTA notifications cumulative active PTAs Source: WTO Secretariat. Note: GATT, General Agreement on Tariffs and Trade; PTA, preferential trade agreement; WTO, World Trade Organization. Landscape 41 Figure 2.3. PTAs Notified to the GATT/WTO and in Force, by Year of Entry into Force, 1959–2009 16 200 180 14 160 12 140 10 cumulative number number per year 120 8 100 80 6 60 4 40 2 20 0 0 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 01 03 05 07 09 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 PTAs cumulative physical PTAs Source: WTO Secretariat. Note: GATT, General Agreement on Tariffs and Trade; PTA, preferential trade agreement; WTO, World Trade Organization. Figure 2.4. PTAs Notified to the GATT (Pre-1995) and the WTO (Post-1995), by Legal Provision GATT (1947–94) WTO (1995–2009) 240 220 200 180 160 140 number 120 100 80 60 40 20 0 in force inactive in force inactive Article XXIV Article V Enabling Clause Source: WTO Secretariat. Note: GATT, General Agreement on Tariffs and Trade; PTA, preferential trade agreement; WTO, World Trade Organization. 42 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Chile, Panama, and individual Central American Common been superseded by newer ones between the same signa- Market (CACM) countries will be replaced by a series of tories or have been consolidated into larger geographic PTAs that link all CACM members with a respective Latin groupings. Of those that remain in force, a third were American partner. notified under the GATT Enabling Clause, which allows Asia, by contrast, exhibits a layering effect whereby preferential treatment among developing countries. Of countries are members of both plurilateral and bilateral the PTAs covering trade in goods notified to the WTO, PTAs. For example, a preexisting PTA between New 90 percent were notified under GATT Article XXIV, Zealand and Singapore coexists alongside the plurilateral which permits PTAs as an exception to most favored PTA linking the same parties, the Trans-Pacific Strategic nation (MFN) rules. Economic Partnership (SEP). Similarly, the PTA between Of the 334 PTAs notified to the WTO, 70 percent the Association of Southeast Asian Nations (ASEAN) and remain in force. As we shall see, much of the recent growth Japan coexists with bilateral PTAs between Japan and indi- of PTAs is accounted for by agreements among developing vidual ASEAN members such as Indonesia, Malaysia, the countries, many of which have notified the goods provi- Philippines, Singapore, Thailand, and Vietnam. This adds sions of their PTAs under GATT Article XXIV, rather than to the complexity of trading relations because agreements exercising their option to notify them under the Enabling may contain different schedules for tariff elimination, rules Clause. of origin, and regulatory provisions. The upward trend in the number of PTAs is evident in Continuing Evolution of the Composition of PTAs figure 2.2. Not only is the number of PTAs increasing, but the number of countries involved continues to diver- The past 15 years or so have witnessed changes in the sify. During the 1990s, much of the proliferation of PTAs dynamics of trading relationships between developed and took place in Europe and Central Asia as the countries developing countries.7 Figure 2.5 shows how participation of those regions forged new trading relationships follow- in PTAs has evolved over time. ing the breakup of the Soviet Union. Since 2000, PTA Among PTAs concluded since the establishment of the participation has become more diverse, spreading to all WTO, the number with exclusively developing-country geographic regions, particularly East Asia and the Pacific. members rose initially and has since remained fairly PTA activity is increasingly concentrated in developing steady; those exclusively among developed countries fell countries. during the first five years of the WTO and have risen Figure 2.3 distinguishes between the number of physical slightly since; and those between developed and develop- PTAs and the number of PTA notifications.5 The number ing countries show the most marked increase. In part, this of services PTAs has been increasing, particularly since mirrors the growing membership of developing countries 2000. For instance, of the 14 distinct PTAs that were noti- in the WTO and the fact that developing countries out- fied and entered into force in 2009, 11 had a services com- number developed countries in our classification by a ponent, and almost three-quarters of all PTAs in force and ratio of about 2:1. However, it also reflects the fact that notified to the WTO contain provisions on trade in serv- preferential trade relations between developed and ices. Also of note is the fact that developing countries are developing countries are increasingly becoming recipro- increasingly negotiating PTAs that include both goods and cal, in part because of the need to fulfill WTO legal obli- services components. gations. In addition, a growing number of developing A total of 183 physical PTAs have been notified to the countries are choosing to forge reciprocal trading rela- WTO (as of February 2010) and are currently in force. tionships with developed countries rather than rely on This figure does not include the hundred or so PTAs that nonreciprocal preferential trading relationships such as are currently in force but have not been notified to the GSP programs. WTO. Such PTAs are almost exclusively among developing Figure 2.6 analyzes the hundred or so PTAs under countries, and most are bilateral agreements involving two negotiation and signed (but not yet in force), based on the parties. parties’ level of development.8 The data shown in the fig- Figure 2.4 looks at the proliferation of PTAs chronolog- ure confirm our observation that North-South PTAs are ically, differentiating between the PTAs notified during the becoming increasingly prevalent. They constitute 69 per- GATT years and those notified since the establishment of cent of the PTAs under negotiation, whereas those exclu- the WTO, according to the relevant legal provision.6 Of sively between developing countries account for 22 percent the 123 PTAs notified during the GATT years, only a and those exclusively between developed countries account quarter remain in force. In many cases, older PTAs have for 9 percent. Landscape 43 Figure 2.5. Evolution of Notified PTAs in Force, by Type of Partner, 1958–2009 40 35 30 25 number 20 15 10 5 0 1958–94 1995–99 2000–04 2005–09 developed only developed-developing developing only Source: WTO Secretariat. Note: PTA, preferential trade agreement. Figure 2.6. Number of PTAs under Negotiation integration in the traditional sense—as with EFTA, and Signed, by Type of Partner, as of February 2010 ASEAN, and the Southern African Customs Union (SACU), which by definition are plurilateral partner- ships—and more as tools for negotiating strategic, bilat- eral, and more flexible market access. 9% A related development is the emergence of PTAs in 22% which all the parties are themselves members of PTAs. The first PTA of this kind notified to the WTO was that between EFTA and SACU, linking the four EFTA countries with the five-member SACU customs union. More PTAs of this type are currently under negotiation—for example, those between the European Union (EU) and the Southern Cone Common Market (Mercosur, Mercado Común del 69% Sur) and between the Gulf Cooperation Council (GCC) and Mercosur. Geographic Configuration of PTAs developed only developed-developing developing only Countries seeking preferential partners have tended Source: WTO Secretariat. recently to look beyond their regional neighbors and far- Note: PTA, preferential trade agreement. ther afield.10 As figure 2.8 shows, as of February 2010, cross-regional PTAs accounted for 28 percent of PTAs notified to the GATT and for 34 percent of those notified to the WTO. The tendency toward the negotiation of PTAs Structural Configuration of PTAs across regional boundaries is more pronounced for PTAs As can be seen in figure 2.7, which differentiates between currently being negotiated; cross-regional PTAs account bilateral and plurilateral PTAs notified and still in force, for two-thirds of the total in this group. plurilateral PTAs accounted for two-thirds of all PTAs Figure 2.9 presents the geographic regions represented notified during the GATT years.9 Since the establishment in PTAs that have been established over the past 10 years. of the WTO, bilateral PTAs have increasingly become the Countries in Europe and Central Asia, particularly mem- norm, making up more than 80 percent of all PTAs noti- bers of EFTA and of the Commonwealth of Independent fied during this period and roughly 90 percent of those States (CIS), were active PTA players during the period. currently under negotiation. This confirms the observation Also notable is the growing PTA participation of countries by Fiorentino, Crawford, and Toqueboeuf (2009) that PTAs in East Asia and the Pacific; no new PTAs came into force are less used as instruments for promoting intraregional in this region in 2000, but a yearly average of more than 44 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Figure 2.7. Bilateral versus Plurilateral PTAs Notified to the GATT/WTO GATT (1947–94) WTO (1995–2009) 200 150 number 100 50 0 bilateral plurilateral Source: WTO Secretariat. Note: GATT, General Agreement on Tariffs and Trade; PTA, preferential trade agreement; WTO, World Trade Organization. Figure 2.8. Cross-Regional and Intraregional PTAs Notified to the GATT/WTO GATT (1947–94) WTO (1995–2009) 200 150 number 100 50 0 intraregional cross-regional Source: WTO Secretariat. Note: GATT, General Agreement on Tariffs and Trade; PTA, preferential trade agreement; WTO, World Trade Organization. five new PTAs did so in the period 2005–09. In 2009, most PTAs consisting of parties in two or more geographic PTA activity was conducted by countries in the Americas regions account for the largest share, 67 percent of the and the Caribbean, led by Canada, Chile, Peru, and the total. Intraregional PTAs under negotiation among coun- United States. By contrast, South Asia and Sub-Saharan tries in the Americas and the Caribbean make up the sec- Africa, home of some of the world’s poorest countries, wit- ond largest group, followed closely by the East Asia and the nessed much less PTA activity during this time, indicating Pacific region. that they may risk becoming further marginalized in their Another view of PTAs in force and under negotiation for pursuit of PTA partners. a number of selected countries is shown in figure 2.11. The Figure 2.10 shows PTAs under negotiation and signed EU continues to be the dominant PTA player, with 29 PTAs but not yet in force, by geographic region. Cross-regional in force and another 14 under negotiation. Chile, EFTA, Landscape 45 Figure 2.9. PTAs, by Region and by Year of Entry into Force, 2000–09 25 20 15 number 10 5 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 East Asia and the Pacific Europe and Central Asia North America, Latin America, and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa Source: WTO Secretariat. Note: PTA, preferential trade agreement. An effort was made to classify each PTA according to the regions of its individual parties and to avoid the use of the cross-regional category. A degree of double-counting therefore arises because, for example, an intraregional PTA is counted once, whereas a PTA involving two or more regions is counted once for each region. Singapore, and Turkey constitute the second most active • Qualitatively different and new provisions lying outside group. Also of interest are Australia, Canada, the GCC, and the current WTO mandate (WTO extra). the Republic of Korea, which have only a handful of PTAs in force but are actively negotiating a number of others. PTA commitments that are also covered at the multilat- eral level include provisions concerning sanitary and phytosanitary (SPS) measures, technical barriers to trade Deepening Scope of PTAs (TBT), antidumping, state aid, and obligations already cov- PTAs are increasingly covering more than trade in goods ered by the General Agreement on Trade in Services (GATS), and services; they extend to rules and disciplines on vari- Trade-Related Aspects of Intellectual Property Rights ous regulatory border and behind-the-border policies.11 (TRIPS), and Trade-Related Investment Measures (TRIMS) For the purposes of this chapter, we define such agree- arrangements. WTO-extra obligations deal with environ- ments as deep PTAs. These might contain mental provisions, labor laws, and movement of capital and also with competition policy, intellectual property rights • Provisions that come under the current mandate of the (IPRs) not referenced in the TRIPS agreement, and so forth. WTO but only reaffirm the existing multilateral com- As figure 2.12 shows, the number of PTAs that include mitments deep provisions has been steadily increasing since the early • Provisions within the current mandate of the WTO but 2000s.12 The most prevalent of these provisions usually in which the parties to a deep PTA undertake commit- concern customs cooperation, IPRs, competition policy, ments beyond those accepted at the multilateral level TBT and SPS measures, government procurement, and (WTO+) investment. Assessing such evolution is far from an exact 46 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Figure 2.10. PTAs in Force and under Negotiation, science. The WTO regional trade agreement (RTA) data- by Region base currently includes only the commitments undertaken in PTAs notified to the WTO following the introduction of 3% the Transparency Mechanism in 2006 and, to a lesser extent, agreements notified before December 2006. (The Transparency Mechanism calls for early announcement 13% of negotiations to set up an RTA and early notification of the RTA’s creation.) As a result, the database currently covers about 45 percent of the PTAs notified to the WTO. 5% Figure 2.12 attempts to supplement the WTO information with other sources to provide a more accurate picture of the basic trends in the inclusion of additional commit- 11% ments in PTAs. Although the cumulative rise in agreements 68% that include such commitments is partly the outcome of the sample composition, the sheer volume of agreements containing such provisions (compared with the total num- ber of PTAs) in recent years indicates the level of interest in deeper integration in the context of preferential trading. Although it is difficult to be precise about the share of agreements containing provisions that go beyond existing cross-regional PTAs commitments at the multilateral level, research on recent East Asia and the Pacific bilateral PTAs signed by the United States and the EU Europe and Central Asia shows that there is an increasing tendency for their agree- North America, Latin America, and the Caribbean ments to exceed existing WTO commitments. The United Middle East and North Africa States and the EU are the main players on the international Source: WTO Secretariat. trade scene, and the bilateral agreements signed by them Note: PTA, preferential trade agreement. often constitute a benchmark for other PTAs. It has been Figure 2.11. PTAs in Force and under Negotiation by Selected Countries and Groupings, as of February 2010 45 40 35 30 number 25 20 15 10 5 0 Au N lia da M ile EU CC a n p. co e ey es TA ur di or pa EA Re at CO os ra Ch rk i na EF ex In G ap Ja St st Tu AS c Ca a, M er RI ng d re M CA ite Si Ko Un PTAs in force as of February 2010 forecast (PTAs currently under negotiation or signed) Source: WTO Secretariat. Note: ASEAN, Association of Southeast Asian Nations; CARICOM, Caribbean Community; EFTA, European Free Trade Association; EU, European Union; GCC, Gulf Cooperation Council; Mercosur, Southern Cone Common Market (Mercado Común del Sur); PTA, preferential trade agreement. Landscape 47 Figure 2.12. Issues Covered in Regional Trade Agreements, 1989–2009 140 120 100 80 number 60 40 20 0 09 07 08 06 04 05 03 01 02 00 99 97 98 94 95 96 92 93 91 90 8 9 20 20 19 19 20 20 20 20 20 20 19 20 19 19 19 19 19 20 19 19 19 customs IPRs competiton TBT SPS government procurement investment Source: WTO RTA database; World Bank Global Preferential Trade Agreements database; Global Economic Prospects (World Bank 2005); and additional observations based on Horn, Mavroidis, and Sapir 2010. Note: IPRs, intellectual property rights; SPS, sanitary and phytosanitary; TBT, technical barriers to trade. estimated that these two parties account for about 80 per- agreements goes beyond multilateral commitments. For cent of the rules that regulate the functioning of world example, in the case of customs administration, the EU markets (Sapir 2007). favors the establishment of a framework for negotiation Horn, Mavroidis, and Sapir (2010) review provisions in that aims at simplifying customs procedures and reducing 28 EU and U.S. PTAs with developed and developing coun- deadweight costs. The United States also seeks to establish a tries. Table 2.1, which is based in part on their findings, framework for cooperation in customs administration, typ- indicates a high degree of coverage of WTO+ areas in both ically requesting that the other party increase transparency EU and U.S. agreements. Provisions on customs coopera- and publish all customs-related laws and regulations. tion, TBT, and public procurement are included in most Enforceable provisions concerning SPS and TBT meas- EU and U.S. agreements. By contrast, provisions on trade ures appear in fewer than half of the EU agreements under in services are included in all but one of the U.S. agree- review. Typically, on top of reinforcing the commitments ments but in only four of the EU agreements. Similarly, of the WTO TBT and SPS agreements, the EU establishes a most U.S. agreements include obligations on TRIPS and on forum designed to promote unilateral or mutual recogni- regulation of export taxes, whereas no EU agreements tion of standards and conformity assessment. These com- include such provisions. mitments are deeper than in the case of U.S. PTAs, which It often proves in EU agreements that either the lan- usually reconfirm the parties’ WTO obligations. guage regarding WTO+ obligations is not sufficiently pre- In the area of services, the obligations can be quite sub- cise to be legally enforceable, or no dispute settlement stantial. In at least one case (the U.S. PTA with Chile), the mechanism is available for enforcing the commitment. United States has adopted regulatory provisions that do Areas that are often unenforceable because of imprecise not exist in the GATS. One such provision requires the par- language include public procurement, TBT and SPS provi- ties to communicate their services-related laws at the draft sions, and environmental laws; in U.S. agreements, SPS and stage, before they are actually enacted. Although the other competition provisions tend to have this shortcoming. party’s comments are not binding, an active integration Overall, however, the U.S. agreements contain substantially process is thus established. fewer areas with legally unenforceable language. The EU and U.S. PTAs contain a number of WTO-extra The depth of the commitments with respect to nontariff measures. In the area of competition, the EU generally measures (NTMs) is increasingly substantial and in most includes legally enforceable provisions in its PTAs, whereas Table 2.1. Deep Commitments in Selected EU and U.S. PTAs, by Type of Provision 48 Parties Customs Public Labor market Environmental to PTA cooperation SPS TBT procurement Competition IPRs Investment Services regulations laws EEA Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes EU–Albania Yes No Yes Yes Yes Yes Yes No No Yes EU–CARIFORUM Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes EU–Chile Yes Yes Yes Yes Yes Yes Yes Yes No Yes EU–Croatia Yes Yes Yes Yes Yes Yes Yes No No Yes EU–Egypt, Arab Rep. Yes No Yes Yes Yes Yes Yes No No Yes EU–Israel Yes Yes Yes Yes Yes Yes No No No Yes EU–Jordan Yes Yes Yes No Yes Yes Yes No No Yes EU–Macedonia, FYR Yes Yes Yes Yes Yes Yes Yes No No Yes EU–Mexico Yes Yes Yes Yes Yes Yes Yes Yes No Yes EU–Morocco Yes No Yes Yes Yes Yes Yes No No Yes EU–South Africa Yes No Yes Yes Yes Yes Yes No No Yes EU–Tunisia Yes No Yes Yes Yes Yes Yes No No Yes EU–Turkey Yes No Yes Yes Yes Yes No No No No Total provisions 14 8 14 13 14 14 12 4 2 13 Legally enforceable provisions 13 3 5 7 13 11 8 4 2 2 NAFTA Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes United States–Australia Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes United States–Bahrain Yes Yes Yes Yes No Yes No Yes Yes Yes United States–CAFTA-DR Yes Yes Yes Yes No Yes Yes Yes Yes Yes United States–Chile Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes United States–Colombia Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes United States–Israel No Yes No Yes No No No No No No United States–Jordan Yes No No Yes No Yes No Yes Yes Yes United States–Korea, Rep. Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes United States–Morocco Yes Yes Yes Yes No Yes Yes Yes Yes Yes United States–Oman Yes Yes Yes Yes No Yes Yes Yes Yes Yes United States–Panama Yes Yes Yes Yes No Yes Yes Yes Yes Yes United States–Peru Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes United States–Singapore Yes No Yes Yes Yes Yes Yes Yes Yes Yes Total provisions 13 12 12 14 7 13 11 13 13 13 Legally enforceable provisions 13 2 11 13 0 13 11 13 13 13 Source: World Bank 2005; Horn, Mavroidis, and Sapir 2010; WTO RTA database; Tuck Trade Agreements Database, Center for International Business, Dartmouth College (http://www.dartmouth.edu/~tradedb/ trade_database.html). Note: The inclusion of deeper provisions in PTAs is shown whether or not the commitments go beyond existing commitments under the relevant WTO agreements. In addition, such provisions may or may not be legally enforceable. CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; EEA, European Economic Area; CAFTA-DR, Central America Free Trade Agreement–Dominican Republic; IPR, intellectual property rights; NAFTA, North American Free Trade Agreement; SPS, sanitary and phytosanitary; TBT, technical barriers to trade. Landscape 49 only about half of the U.S. PTAs considered by the Horn, and protect investment, technical assistance, and so on.13 Mavroidis, and Sapir (2010) study contain such provisions, The U.S. agreements, by contrast, generally include legally and none is legally enforceable. Most EU PTAs prohibit enforceable rules whereby parties agree to extend MFN and agreements between enterprises that have the object or national treatment to each other and provide mechanisms effect of prevention, restriction, or prohibition of competi- for compensation in case of expropriation, as well as detailed tion, and most bar, as well, the abuse of a dominant posi- rules for arbitration in case of conflict. tion by one or more enterprises in activities affecting trade As for IPRs, all of the EU and U.S. PTAs listed in table 2.1 between parties. The agreements also stipulate that the contain legally binding clauses that oblige the parties to competition authorities of the PTA parties cooperate to become signatories to various intellectual property agree- ensure that such prohibitions are enforced. Many EU PTAs ments not covered by the TRIPS agreement. The obligations also prohibit public aid that distorts or might distort com- under the U.S. agreements tend to be more comprehensive petition by favoring certain enterprises or the production and to cover more aspects of intellectual property rights of certain goods. Legal enforceability varies among EU than do EU agreements. PTAs. For example, obligations of this kind in EU PTAs with Latin American countries are less far reaching than those signed with other countries. The EU–Mexico PTA Global Landscape of PTAs: State of Play and Future Regional Developments does not refer to prohibitions, as do other agreements, but simply mandates that the parties agree on the appropriate Figures 2.13 and 2.14 show the participation of individual measures for preventing distortions or restrictions of com- countries in PTAs covering trade in goods and services petition that could significantly affect trade between the that had been notified and were in force as of February EU and Mexico. 2010. As regards trade in goods, the EU is engaged in the In the case of investment, most EU and U.S. PTAs contain highest number of PTAs, followed by the United States, legally enforceable obligations, but of quite different kinds. Chile, Mexico, and the EFTA states, which are members of Typically, the EU agreement refers only to cooperation to 10 to 19 PTAs. Canada, Australia and most countries in promote investment between parties through the establish- Asia participate in five to nine PTAs, whereas most African ment of mechanisms to provide information on investment countries and some Latin American countries are involved rules, development of a bilateral legal framework to promote in one to four. The map does not show the hundred or so Figure 2.13. Participation in Notified PTAs as of February 2010 (Goods) Source: WTO Secretariat. Note: PTA, preferential trade agreement. 50 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Figure 2.14. Participation in Notified EIAs as of February 2010 (Services) Source: WTO Secretariat. Note: EIA, economic integration agreement. PTAs, mostly involving developing countries, that are in Europe and Central Asia force but have not been notified to the WTO. On the European continent, the largest network of PTAs The situation with regard to trade in services is some- revolves around the European Union. The EU itself, by what different. Most countries in Africa and the Middle virtue of successive enlargements (most recently, from 25 East have not yet notified PTAs involving trade in services. to 27 in 2007), has been part of a changing network of Interestingly, Chile, Mexico, Singapore, and the United PTAs in the region.14 In addition to its own enlargements, States are engaged in more PTAs involving services than are its longest-standing relations in the region are with the the EU and the EFTA states. EFTA states (Iceland, Liechtenstein, Norway, and Switzer- This section examines the PTAs in force and under land), beginning with PTAs in goods in the early 1970s and negotiation in each of five geographic regions. Figures in services in 1994, and with Turkey, with which it has had 2.15–2.18 show the networks of plurilateral groupings in a customs union since January 1996. The EFTA states and Europe and Central Asia; the Americas and the Caribbean; Turkey, by virtue of their association with the EU, have South Asia, East Asia, and the Pacific; the Middle East and continued to expand their own PTA networks both within North Africa; and Sub-Saharan Africa. What is immedi- and outside the region. ately clear from these diagrams is the degree of overlap- Since enlarging to 27 member states, the EU has contin- ping plurilateral PTAs, particularly in Central Asia and ued to expand its relationship with southeastern Europe Africa, but also increasingly in the Americas and Asia. For and with countries in the Mediterranean Basin. In south- ease of reference, bilateral relationships are not shown but eastern Europe, the EU has PTAs in force with Albania, are discussed in the text. Even when only plurilateral rela- Bosnia and Herzegovina, Croatia, the former Yugoslav tionships are shown, the complexity of trade relations Republic of Macedonia, and Montenegro. All these, faced by a number of countries is clearly evident. This together with Serbia, the United Nations Interim Adminis- complexity manifests itself in the resources necessary to tration Mission in Kosovo (UNMIK), and Moldova, are administer and implement the PTA; the burden placed on also members of the Central European Free Trade Agree- customs officials charged with applying differing tariff ment (CEFTA), which entered into force on May 1, 2007. schedules depending on the good and its origin; and the With the inauguration of that agreement, a number of dilemma facing exporters who supply their goods in dif- bilateral agreements between CEFTA’s members were ter- ferent markets and are confronted with different regula- minated.15 The EU launched negotiations with Ukraine in tory regimes such as rules of origin and product stan- February 2008. dards, according to the destination of the good. Landscape 51 Figure 2.15. Network of Plurilateral Groupings in Europe and Central Asia EAEC Russian Kyrgyz Federation EurAsEc Republic Belarus EEA CEZ Kazakhstan Tajikistan Ukraine CIS Uzbekistan CARICOM Georgia Turkmenistan ECO European Union OCTs Armenia Azerbaijan Iran, Islamic Rep. Moldova SACU EFTA Pakistan Norway UNMIK Iceland Serbia Afghanistan Liechtenstein Albania CEFTA Turkey Croatia Switzerland Macedonia, FYR Montenegro Bosnia and Herzegovina Euromed partners EU EFTA Morocco Algeria Tunisia Egypt, Arab Rep. Palestinian Auth. Israel Jordan Lebanon Syrian Arab Republic Source: WTO Secretariat. Note: CARICOM, Caribbean Community; CEFTA, Central European Free Trade Agreement; CEZ, Common Economic Zone; CIS, Commonwealth of Independent States; EAEC, Eurasian Economic Community; ECO, Economic Cooperation Organization; EEA, European Economic Area; EFTA, European Free Trade Association; EU, European Union; OCTs, Overseas Countries and Territories; SACU, Southern African Customs Union; UNMIK, United Nations Interim Administration Mission in Kosovo. Progress is being made toward the formation of a Euro- Authority, Syria, and Tunisia, and it is currently negotiating pean Union–Mediterranean PTA, which was expected to be agreements with Algeria, Jordan, and Lebanon. Once the completed by 2010 but is not yet complete with respect to PTA is completed, it will permit diagonal cumulation of geographic coverage or scope. The EU has signed Euromed origin across all the parties (the EU, EFTA, Turkey, and the association agreements with all its Mediterranean partners. Mediterranean partners) and the Faroe Islands. The most recent, with the Syrian Arab Republic, was Farther afield, the EU has increased its interest in PTAs expected to be completed in 2010 but is not yet in force.16 in recent years. In the Americas, negotiations on an agree- Efforts are also under way to deepen agreements with the ment with Canada, to include goods and services, were Arab Republic of Egypt, Israel, Lebanon, the Palestinian launched in May 2009, and negotiations have been ongoing Authority, and Syria, through the addition of services with Central American countries. In early March 2010, the chapters, and to enhance agricultural liberalization com- EU announced that it had concluded a PTA covering goods mitments in a number of the agreements. The EFTA and services with Colombia and Peru. Negotiations with states and Turkey, through their agreements with the EU, Mercosur, stalled since 2004, have recently restarted and are following suit. EFTA has thus far notified agreements may be concluded in 2010. Negotiations with Ecuador (in goods) to the WTO with all the Mediterranean part- were suspended in July 2009. ners except Algeria and Syria. Turkey has notified agree- In Asia, the EU’s PTA with Korea was initialed in ments in goods with Egypt, Israel, Morocco, the Palestinian October 2009. Negotiations with ASEAN, which began in 52 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Figure 2.16. Network of Plurilateral Groupings in the Americas and the Caribbean NAFTA EFTA Canada Mexico EU Mexico-Northern triangle EFTA LAIA/ALADI United States CAFTA-DR Honduras CAN Guatemala Venezuela, RB Bolivia El Salvador Colombia Cuba Nicaragua Ecuador Costa Rica Dominican Mercosur SACU Rep. Peru CACM Argentina Brazil Paraguay Panama CARICOM Uruguay EU EU Chile EFTA Trans-Pacific SEP Source: WTO Secretariat. Note: CACM, Central American Common Market; CAFTA-DR, Dominican Republic–Central America Free Trade Agreement; CAN, Andean Community; CARICOM, Caribbean Community; EFTA, European Free Trade Association; EU, European Union; LAIA/ALADI, Latin American Integration Association/ (Asociación Latinoamericana de Integración); Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement; SACU, Southern African Customs Union; Trans-Pacific SEP, Trans-Pacific Strategic Economic Partnership. July 2007, are currently on hold while the EU pursues bilat- But Arms (EBA) scheme would be able to export to the EU eral negotiations with each of the ASEAN member states, under the GSP. The EU is conducting EPA negotiations with beginning with Singapore and Vietnam. Negotiations with seven groups of countries: the Economic Community of India were launched in June 2007 and are currently in West African States (ECOWAS), plus Mauritania; the Eco- progress. nomic and Monetary Community of Central Africa Following the expiration of the WTO waiver for the 2000 (CEMAC, Communauté Économique et Monétaire de Cotonou trade preferences for the African, Caribbean, and l’Afrique Centrale), plus São Tomé and Principe and the Pacific (ACP) states, the EU has pursued negotiations for Democratic Republic of Congo; Eastern and Southern economic partnership agreements (EPAs) with these coun- Africa (ESA); the East African Community (EAC); the tries. On December 20, 2007, the EU adopted a market Southern African Development Community (SADC); access regulation to grant duty-free and quota-free access to CARIFORUM (CARICOM and the Dominican Republic); ACP countries that had concluded negotiations on agree- and 14 Pacific countries. Agreements between the EU and ments establishing or leading to the establishment of EPAs the CARIFORUM states, Cameroon, and Côte d’Ivoire are as of January 1, 2008, for all products except rice and sugar; already in force. Only the agreement between the EU and the latter two products are expected to become duty-free in the CARIFORUM states includes services. 2010 and 2015, respectively. Countries that have neither ini- The EFTA states have also been very active in PTA nego- tialed an agreement nor have access to the EU’s Everything tiations in recent years, having notified 14 agreements in Landscape 53 Figure 2.17. Network of Plurilateral Groupings in South Asia, East Asia, and the Pacific Mercosur APTA Japan SAFTA China India Bhutan Pakistan Sri Lanka Korea, Rep. EFTA Maldives Bangladesh Nepal Lao PDR PICTA ASEAN Cambodia Kiribati Myanmar Cook Islands Indonesia Malaysia Micronesia, Fed. Sts. Thailand Vietnam Niue Nauru Tuvalu Singapore Philippines Marshall Islands Samoa EFTA Tonga Solomon Islands Brunei Darussalam Vanuatu Fiji MSG Papua New Guinea Trans-Pacific Australia SEP PATCRA ANZCERTA New Zealand SPARTECA Source: WTO Secretariat. Note: ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Agreement; EFTA, European Free Trade Association; APTA, Asia-Pacific Trade Agreement; ASEAN, Association of Southeast Asian Nations; Mercosur, Southern Cone Common Market (Mercado Común del Sur); MSG, Melanesian Spearhead Group: PATCRA, Papua New Guinea–Australia Trade and Commercial Relations Agreement; PICTA, Pacific Island Countries Trade Agreement; SAFTA, South Asian Free Trade Arrangement; SPARTECA, South Pacific Regional Trade and Economic Cooperation Agreement; Trans-Pacific SEP, Trans-Pacific Strategic Economic Partnership. goods and 5 in goods and services to the WTO. In addition, successful, in that it has not been implemented by all the EFTA has signed agreements with Albania, Colombia, and parties. Preferential trade liberalization in the region has Serbia. It is involved in negotiations with Algeria; Hong therefore developed through a complex network of over- Kong SAR, China; India; Peru; Thailand; and Ukraine and lapping bilateral PTAs and plurilateral initiatives between is considering launching negotiations with Indonesia, the states of the region. For instance, Ukraine has notified Malaysia, Russia, and Vietnam. An interesting development PTAs with Armenia, Azerbaijan, Belarus, Georgia, Kaza- was the decision by Switzerland to break ranks and sepa- khstan, the Kyrgyz Republic, FYR Macedonia, Moldova, rately negotiate PTAs with Japan and China; the former Russia, Tajikistan, Turkmenistan, and Uzbekistan.17 entered into force on September 1, 2009, and the latter is In addition to the CIS PTA, other plurilateral agree- currently being negotiated. The other EFTA members have ments between the same parties include the Common Eco- also launched separate negotiations with China. nomic Zone (CEZ) between Belarus, Kazakhstan, Russia, In Central Asia, trade relations are still very much and Ukraine and the Eurasian Economic Community defined by historical linkages between the successor states (EAEC), made up of three CEZ parties (Belarus, Kaza- to the former Soviet Union and with the bordering coun- khstan, and Russia), plus the Kyrgyz Republic and Tajik- tries. The PTA set up by the CIS was negotiated in an istan. In addition, Belarus, Kazakhstan, and Russia have attempt to maintain these links, but it has not been very formed the EurAsEc Customs Union, which entered into 54 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Figure 2.18. Network of Plurilateral Groupings in Africa and the Middle East ECOWAS Qatar Kuwait Cape Verde Bahrain Oman Iraq Gambia, The United Arab Emirates Agadir Lebanon Guinea Saudi Arabia Yemen, Rep. Sierra Leone WAEMU/ AMU Syrian Arab Republic UEMOA GCC Liberia Palestinian Auth. Mali PAFTA Ghana Niger Tunisia Jordan Mauritania Morocco Nigeria Burkina Faso Egypt, Arab Rep. Senegal Algeria Libya Sudan Guinea-Bissau Benin Togo Chad Uganda Equatorial Guinea Ethiopia Kenya Côte d’Ivoire EAC Eritrea Rwanda Central African Republic Comoros Burundi Gabon Congo, Rep. Djibouti CEMAC Zambia COMESA Tanzania Cameroon Congo, Dem. Rep. Zimbabwe Malawi Madagascar Mauritius Seychelles SADC Namibia Swaziland Angola Botswana Mozambique Lesotho EFTA SACU Mercosur EU South Africa Source: WTO Secretariat. Note: AMU, Arab Maghreb Union; CEMAC, Economic and Monetary Community of Central Africa (Communauté Économique et Monétaire de l’Afrique Centrale); COMESA, Common Market for Eastern and Southern Africa; EAC, East African Community; ECOWAS, Economic Community of West African States; EFTA, European Free Trade Association; EU, European Union; GCC, Gulf Cooperation Council; Mercosur, Southern Cone Common Market; PAFTA, Pan-Arab Free Trade Area; SACU, Southern African Customs Union; SADC, Southern African Development Community; WAEMU/UEMOA, West African Economic and Monetary Union/Union Économique et Monétaire Ouest-Africaine. force in January 2010 and will be implemented over a five- Bahrain, Israel, Jordan, Morocco, and Oman in the Middle year transition period. There is also an overlap between the East and North Africa (MENA) region; and Australia and EAEC, the CEZ, and the CIS, as some parties to these Singapore in the East Asia and the Pacific region. Agree- agreements are also members of the Economic Coopera- ments with Colombia, Korea, Panama, and SACU have tion Organization (ECO), which consists of Afghanistan, been signed but have not yet entered into force pending Azerbaijan, the Islamic Republic of Iran, Kazakhstan, the congressional approval. Negotiations appear to be in Kyrgyz Republic, Pakistan, Tajikistan, Turkmenistan, progress with Malaysia, Thailand, and the United Arab Turkey, and Uzbekistan. Emirates, and an enlargement of the Trans-Pacific Strategic Economic Partnership (SEP) between Brunei Darussalam, Chile, New Zealand, and Singapore to include the United The Americas and the Caribbean States has been proposed. The Americas and the Caribbean continue to be actively Canada has been less involved for a number of years but involved in PTAs. In addition to NAFTA, the United States has recently stepped up its participation in PTAs. Agree- has agreements with numerous countries, including Chile ments with EFTA and with Peru have been notified and and the Central America Free Trade Agreement plus the entered into force in 2009. Agreements with Colombia and Dominican Republic (CAFTA-DR) in its own hemisphere; Jordan have been signed but have yet to enter into force, Landscape 55 and negotiations with Panama were concluded in August negotiating with Korea. Both Colombia and Peru have 2009. Canada is also currently involved in negotiations recently completed PTA negotiations with the EU.20 with CARICOM, the Dominican Republic, four Central Ecuador currently has agreements in force with Chile, America countries, Singapore, Korea, and the EU, while Cuba, and Mercosur within the LAIA/ALADI framework. PTAs have been proposed with India, Mercosur, Morocco, Mercosur has signed several framework agreements and Ukraine. aimed at establishing of PTAs but has only one partial- Mexico continues to expand its already substantial net- scope agreement in force, that with India, in addition to work of PTAs.18 In addition to countries within the hemi- agreements under the LAIA/ALADI framework.21 The sphere, it has PTAs with EFTA, the EU, Israel, and Japan. group is currently negotiating agreements with Canada, Further expansion of its PTA network is planned, with the EU, and Turkey. negotiations in progress with Korea and Singapore. The Chile has for several years been among the most active Central and South American regions maintain complex participants in PTA negotiations, and its agreements span intraregional and extraregional relations. Within the all the continents. It has agreements in force with Australia, region, there are four customs unions at various stages of Canada, China, Colombia, Costa Rica, El Salvador, EFTA, completion: CACM in Central America, CARICOM in the the EU, Honduras, Japan, Korea, Mexico, Panama, Trans- Caribbean, and the Andean Community (CAN) and Mer- Pacific SEP members (Brunei Darussalam, New Zealand, cosur in South America. In addition, a large number of and Singapore), and the United States. It also has a partial- bilateral agreements, representing varying degrees of inte- scope agreement in force with India and several agree- gration, have been negotiated within the Latin American ments under the LAIA/ALADI framework.22 In addition to Integration Framework (LAIA; in Spanish, ALADI, Aso- these, it has signed an agreement with Guatemala in the ciación Latinoamericana de Integración). CACM members context of its agreement with Central American countries have also concluded PTAs with other countries in the and has extended the agreement with China to include region. Chile has bilateral protocols in force with all the services. Negotiations are ongoing with Malaysia, members of the CACM except Guatemala and Nicaragua, Nicaragua (under the Central American agreement), Thai- where negotiations are still going on. Panama’s individual land, Turkey, and Vietnam. bilateral protocols with CACM members are in force, and the country also has agreements in force with Chile, South Asia, East Asia, and the Pacific Colombia, the Dominican Republic, Mexico, Singapore, and Taiwan, China.19 Agreements with Canada and the After several years of resistance to signing PTAs, the Asia United States have been signed but have yet to enter into and the Pacific region has been playing catch-up and has force, and Panama is considering negotiations with Merco- become one of the most active regions in PTA negotiations, sur. In the Caribbean, CARICOM, in addition to its PTAs both among countries in the region and with extraregional with Colombia, Costa Rica, Cuba, the Dominican Repub- partners. East Asia has been particularly active, with Japan lic, and the República Bolivariana de Venezuela, is negotiat- and China taking the lead. Japan alone has 11 agreements ing with Canada and is considering an agreement with in force, 8 of which have taken effect since 2007. Most of Mercosur. CARICOM members together with the Domini- these agreements are with ASEAN and its members can Republic are parties to the EU–CARIFORUM EPA, (Brunei Darussalam, Indonesia, Malaysia, the Philippines, which became effective at the end of November 2008. Singapore, Thailand, and Vietnam), but agreements are In South America, two of the main regional blocs, also in force with Chile, Mexico, and Switzerland. An Mercosur and the Andean Community, are pursuing a agreement with Peru entered into force on March 1, PTA. Individual Andean Community members are also 2010, and Japan is negotiating with Australia, the GCC, negotiating PTAs, both within and outside the region. As of India, and Korea. China currently has nine agreements in 2009, Peru had PTAs in force with Canada, Singapore. and force: with ASEAN; with the Asia-Pacific Trade Agreement the United States, and its agreement with China entered (APTA), which includes Bangladesh, India, Korea, the into force on March 1, 2010. It also has agreements under Lao People’s Democratic Republic, and Sri Lanka; and the LAIA/ALADI framework with Chile, Cuba, Mercosur, with Chile; Hong Kong SAR, China; Macao, China; New and Mexico and is currently negotiating with EFTA, Japan, Zealand; Pakistan; Peru; and Singapore. Negotiations with Korea, and Thailand. Colombia, in addition to having Australia, Costa Rica, the GCC, Iceland, Norway, and agreements in force with CARICOM, Chile, Costa Rica, El Switzerland are in progress. Korea, in addition to being a Salvador, Guatemala, and Panama, has signed agreements party to APTA, has agreements with ASEAN, Chile, EFTA, with Canada, EFTA, and the United States and is currently and Singapore, has signed agreements with the United States 56 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard and the EU, and has launched PTA negotiations with a developed a significant list of preferential partners. Australia number of parties, including Australia, Canada, Colombia, currently has PTAs with Chile, Singapore, Thailand, and the GCC, India, Japan, Mexico, New Zealand, and Peru. the United States; a trilateral agreement with ASEAN and Taiwan, China, is expanding its network of PTAs, having New Zealand has recently entered into force. Negotiations notified agreements with Nicaragua and Panama to the are being held with China, the GCC, Japan, Korea, and WTO in 2009; agreements are in force with Guatemala and Malaysia, and entry into the Trans-Pacific SEP is being dis- with Honduras–El Salvador. An agreement with the cussed. New Zealand’s PTAs are with Australia, China, Dominican Republic is currently being negotiated, and dis- Singapore, and Thailand, and it is a party to the Trans- cussions on a PTA with China began recently. Pacific SEP. It has signed an agreement with Malaysia and In Southeast Asia, the major trading bloc, ASEAN, is has concluded agreements with the GCC countries and with working toward the creation of an East Asian Economic Hong Kong SAR, China. Negotiations are ongoing with Community by 2015. When fully implemented, this India and Korea. Australia and New Zealand are renego- scheme is expected to form a single market in goods, serv- tiating their nonreciprocal agreement, the South Pacific ices, and investment. ASEAN members are also negotiating Regional Trade and Economic Cooperation Agreement PTAs with other parties, both as individual members and (SPARTECA), with the Pacific Island countries, with the aim as a group. ASEAN itself has agreements in force with of replacing it with a reciprocal PTA (PACER Plus). In the China, India, Japan, and Korea and with Australia and meantime, the Pacific Island Countries Trade Agreement New Zealand. It is currently negotiating with the EU. (PICTA), which was notified to the WTO in August 2008, is Individually, Singapore has led the way, with 11 PTAs in being implemented by most of the parties. The PICTA coun- force (with Australia, China, EFTA, Japan, Jordan, Korea, tries are negotiating an EPA with the EU, but to date, only India, New Zealand, Panama, the Trans-Pacific SEP, and Fiji and Papua New Guinea have initialed interim EPAs. the United States), and it is negotiating another 7 (with Canada, Costa Rica, the EU, the GCC, Mexico, Pakistan, The Middle East and North Africa and Ukraine). Other active ASEAN members are Thailand, which has four agreements in force and five under In the Middle East and North Africa, the key plurilateral negotiation, and Malaysia, with two in force and another agreements are the Agadir Agreement between Egypt, six under negotiation.23 Jordan, Morocco, and Tunisia, in force since 2007; the Gulf PTAs in South Asia have largely been confined to agree- Cooperation Council customs union, in force since 2003; ments among neighboring countries, although this is chang- and the Pan-Arab Free Trade Agreement (PAFTA), which ing rapidly. India and Pakistan have both expanded their has been in force since January 1, 1998, and includes negotiations to countries outside the immediate region. In members of the GCC and the Agadir Agreement, as well as addition to agreements within the region, such as the South other countries in the region. The Arab Maghreb Union Asian Free Trade Area (SAPTA, which includes Bangladesh, (AMU) includes Agadir parties Tunisia and Morocco, as Bhutan, Maldives, Nepal, Pakistan, and Sri Lanka) and PTAs well as Algeria, Libya, and Mauritania. In addition, a with Afghanistan, Bhutan, Nepal, Pakistan, and Sri Lanka, crisscrossing network of bilateral agreements exists. For India has PTAs with APTA, Singapore, and, more recently, instance, Jordan has agreements with Bahrain, Egypt, ASEAN. It also has partial-scope agreements with Chile and Israel, Morocco, the Palestinian Authority, Sudan, Syria, Mercosur and is currently in negotiations with EFTA, the Tunisia, and the United Arab Emirates and is negotiating EU, the GCC, Japan, Korea, Mauritius, Thailand, and the with the GCC. Tunisia has agreements with Egypt, Iraq, Bay of Bengal Initiative for Multi-Sectoral and Economic Jordan, Libya, and Morocco. Egypt has agreements with Cooperation (BIMSTEC). Pakistan, in addition to being a Iraq, Jordan, Lebanon, Libya, Morocco, the Palestinian party to SAPTA, the Protocol relating to Trade Negotiations Authority, Syria, and Tunisia and has proposed negotia- among Developing Countries (PTN), and ECO, has notified tions with India. agreements with China, Malaysia, and Sri Lanka to the Links with countries outside the region are also expand- WTO; PTAs with Mauritius and the Islamic Republic of ing. Several countries are part of the Euromed process of Iran are also in force. Pakistan is negotiating agreements agreements with the EFTA and EU. The United States has with the GCC and Singapore and is in negotiations to agreements with Bahrain, Israel, Jordan, Morocco, and expand the ECO. Oman, and is currently negotiating with the United Arab In the Pacific region, Australia and New Zealand, in Emirates. The GCC as a group is also negotiating with a addition to their long-standing Closer Economic Rela- large number of partners across the world; an agreement tions Agreement (ANZCERTA), in effect since 1983, have with Lebanon is in force, and agreements have been signed Landscape 57 with EFTA, Singapore, and Syria. The GCC’s extensive PTA members and with the rest of the world. Several factors negotiating agenda includes Australia, China, the EU, can contribute to the success or the failure of a PTA in India, the Islamic Republic of Iran, Japan, Jordan, Korea, stimulating trade flows. Coverage and the degree of liberal- Mercosur, New Zealand, Pakistan, and Turkey. ization are of crucial importance; clearly, agreements that cover substantially all trade, including agricultural prod- ucts and services, and those that incorporate significant Sub-Saharan Africa tariff and quota reductions are more likely to lead to higher Regional integration in Sub-Saharan Africa has, for the trade flows among their members. If the barriers to trade most part, taken the form of PTAs among geographically with the rest of the world are kept low, as well, the risk of contiguous countries. SACU, the world’s oldest customs trade diversion is minimized, and trade with third parties is union, is engaged in negotiating PTAs and recently notified likely to be created as a result of the PTA. an agreement with EFTA. Other efforts at creating intrare- The proliferation of PTAs that leads to overlapping gional and extraregional partnerships have fallen short of agreements with varying rules of origin and diverse tariff their ambitious statements of intent. In several cases, mem- schedules may complicate integration into global value bership of regional groupings is defined by political chains and prove detrimental to trade. Hence, a successful alliances rather than market access goals, resulting in PTA is likely to be associated with nonrestrictive rules of overlapping memberships that create difficulties in imple- origin. In addition, trade facilitation measures, because of mentation.24 Negotiations for an economic partnership their effect on the costs of trade, are important to a success- agreement with the EU, although intended to strengthen ful PTA. Finally, the comprehensiveness of the PTA is of regional integration, have created further confusion in importance. The inclusion of behind-the-border regula- eastern and southern Africa because memberships of the tory measures that foster increased cross-border competi- EPA groups and the regional agreements are different.25 tion, including competition in services, and that establish With regard to the current state of play of the EPA negotia- rules governing investment and IPRs suitable to the part- tions, in June 2009, an interim EPA was signed between ners’ level of development contributes to the success of the the EU and Botswana, Lesotho, and Swaziland (part of PTA. Finally, even the best-designed agreements, if not the SADC EPA); Mozambique joined soon afterward. An implemented in full, will not bear the expected fruits. interim EPA was initialed (but not yet signed) between Often, the most effective PTAs have been those designed to the EU and the Seychelles, Zambia, and Zimbabwe in complement a general program of economic reform. November 2007 and with the Comoros, Madagascar, and This section surveys trade flows for a selection of major Mauritius in December 2007 for the Eastern and Southern plurilateral PTAs; both intra-PTA and extra-PTA trade Africa (ESA) EPA. An interim agreement was initialed flows are included. Trade between members of the selected between the EU and Uganda in November 2007 for the PTAs is equivalent to roughly 40 percent of world trade. A East African Community (EAC) EPA. gravity model is developed to quantify the importance of In West Africa, the main regional groups are the West these PTAs in stimulating trade among their members and African Economic and Monetary Union (WAEMU; in with third countries. French, Union Économique et Monétaire Ouest-Africaine, UEMOA), ECOWAS, and CEMAC, all three of which are Trade Developments for Selected Plurilateral PTAs customs unions in force or in the making. The eight WAEMU members are all members of ECOWAS. The EU is negotiat- The most intuitive indication of the success of a PTA is the ing EPAs with ECOWAS and CEMAC. With regard to the increase in trade among PTA partners as a share of total ECOWAS EPA, only Côte d’Ivoire and Ghana have initialed trade.26 Even when the share of regional trade does not the interim agreement, in December 2007; the EU and Côte increase, it is possible that the volume of intra-PTA trade d’Ivoire notified the interim EPA in goods to the WTO in will rise if liberalization with respect to third partners leads December 2008. In the case of the CEMAC EPA, the EU noti- to expansion of total trade. fied the WTO of the provisional application of the interim As figure 2.19 shows, for most of the PTAs under con- agreement in goods with Cameroon in October 2009. sideration, intra-PTA imports as a share of total imports increased following the introduction of the PTA. (Extra- PTA trade and possible trade diversion are discussed Impact of PTAs on Trade below.) The most pronounced increases in the shares This next-to-last section examines the extent to which of intra-PTA imports were observed in the EU, ASEAN, PTAs have contributed to the expansion of trade among the Andean Community (CAN), SAFTA, and, initially, 58 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Figure 2.19. Evolution of the Share of Intra-PTA Imports in Total Imports, 1970–2008 ASEAN CACM 30 20 25 15 20 percent percent 15 19% 10 18% 10 17% 5 8% 5 0 0 19 0 19 3 19 6 19 9 19 2 19 5 19 8 91 19 4 20 7 00 20 3 06 08 19 0 73 19 6 19 9 19 2 19 5 19 8 19 1 19 4 20 7 00 03 2006 08 7 7 7 7 8 8 8 9 9 0 7 7 7 8 8 8 9 9 9 19 19 20 20 19 19 20 20 COMESA ECOWAS 10 20 8 15 percent percent 6 10 4 11% 5 2 3% 0 0 19 0 19 3 19 6 19 9 82 19 5 19 8 19 1 19 4 20 7 00 03 2006 08 19 0 19 3 19 6 19 9 82 19 5 19 8 19 1 19 4 20 7 00 20 3 2006 08 7 7 7 7 8 8 9 9 9 7 7 7 7 8 8 9 9 9 0 19 19 20 20 19 19 20 EU Mercosur 70 25 60 66% 64% 20 50 62% 62% percent percent 40 53% 15 30 10 14% 20 5 10 0 0 19 0 19 3 19 6 19 9 19 2 19 5 19 8 91 94 20 7 00 20 3 2006 08 19 0 73 19 6 19 9 19 2 19 5 19 8 19 1 19 4 20 7 00 20 3 20 6 08 7 7 7 7 8 8 8 9 0 7 7 7 8 8 8 9 9 9 0 0 19 19 19 20 19 19 20 NAFTA WAEMU/UEMOA 50 12 40 10 8 10% percent percent 30 39% 6 20 4 10 2 0 0 19 0 19 3 19 6 19 9 82 19 5 88 19 1 19 4 20 7 00 20 3 06 08 19 0 19 3 19 6 19 9 19 2 19 5 19 8 19 1 94 20 7 00 03 2006 08 7 7 7 7 8 9 9 9 0 7 7 7 7 8 8 8 9 9 19 19 19 20 20 19 19 20 20 Source: International Monetary Fund (IMF), Direction of Trade Statistics (DOT). Note: ASEAN, Association of Southeast Asian Nations; CACM, Central American Common Market; COMESA, Common Market for Eastern and Southern Africa; ECOWAS, Economic Community of West African States; EU, European Union; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement; WAEMU/UEMOA, West African Economic and Monetary Union/Union Économique et Monétaire Ouest- Africaine. The dot on the plot line in each panel indicates the date of entry into force of the agreement (or enlargements, in the case of the EU). Landscape 59 NAFTA. The share of intra-NAFTA trade began to fall in Next, we consider external protection as one of the the early 2000s, largely because increased imports from likely factors determining whether the presence of PTAs China replaced Mexican exports on the U.S. market might be an obstacle to growth of trade with external part- (Batista 2008) and also crowded out domestic production ners. Low external tariffs reduce the potential for trade in Mexico. The average share of intra-ASEAN imports in diversion, ensure access to competitively priced inputs, and total imports increased from 17 percent in the 1980s to increase competition in the domestic market. The external 25 percent in the 2000s; for intra-CAN trade during that protection of all PTAs under consideration has been falling period, the average share increased from 5 to 14 percent. over the past decade (see figure 2.21).27 Several PTAs The intra-Mercosur share increased substantially, from (EFTA, the EU, the GCC, and NAFTA) have an average 9 percent in the 1980s to slightly more than 20 percent in MFN tariff of about 5 percent or lower. Most PTAs impose the late 1990s. The rise, however, preceded the creation of MFN tariffs that are, on average, less than 10 percent; these the PTA by several years, and so it is uncertain to what include ASEAN, CACM, CAN, CEFTA, Mercosur, SACU, extent Mercosur was responsible for increasing trade flows and SADC. Preferential agreements with third parties and among its members. In many cases, increases in regional multilateral liberalization have also led to growth in the trade have been associated with unilateral or multilateral proportion of imports entering PTA markets tariff-free liberalization, as well, and with growing economic rela- (see figure 2.22). The share in total imports of imports sub- tionships with PTA partners. ject to zero MFN rates (calculated as an average for all For several PTAs, the share of intra-PTA imports in total members) has increased since 1995 for all PTAs except imports has been falling or stagnant. These include EFTA, ECOWAS and WAEMU. On average, more than half of all where the share of intra-PTA imports decreased from imports enter the markets of developed countries at zero 22 percent in the 1980s to 17 percent in the early 2000s, and MFN tariffs. Several developing-country PTAs (ASEAN, the CACM, with a drop from 12 to 7 percent over the same CACM, EAC, SACU, and SADC) also have relatively open period. Several other PTAs, such as ANZCERTA, CEMAC, trade regimes. the GCC, the Papua New Guinea–Australia Trade and In Latin America and Africa, only a low share of imports Commercial Relations Agreement (PATCRA), and SADC, enters existing PTAs free of duty (figure 2.22). PTAs for experienced an increase in intra-PTA trade flows in the which this is true, including CAN, CEMAC, the GCC, and 1990s, followed by a decline in the early 2000s. The decrease WAEMU, are likely to perform worse with respect to cre- in intra-PTA trade may be less worrying if the value of total ation of trade with nonmembers than PTAs that are more trade is increasing. Figure 2.20 indicates that the share of open. This result is consistent with the findings of Global intraregional trade (imports and exports) in gross domestic Economic Prospects 2005 (World Bank 2005), in which, on product (GDP) has indeed been growing in most of the the basis of the gravity model, the authors concluded that sampled PTAs since the early 1970s. EFTA was the only PTA several PTAs (CEMAC, CIS, COMESA, EAC, ECOWAS, in which, on average, the share of intraregional trade in SADC, and WAEMU) registered lower-than-expected GDP was higher in the 1980s than in the early 2000s. overall exports. Moreover, for all the PTAs under review, the growth in intra-PTA exports has been associated with growth in total Question of Intra- and Extra-PTA Trade Creation exports (table 2.2). The coefficient of correlation is positive for most PTAs and is quite high for several of them, such as A simple gravity model of trade could help insulate the ASEAN, EFTA, the EU, the GCC, and NAFTA. In the case effect of PTAs from all the other factors in play in explain- of some African PTAs (Common Market for Eastern and ing trade developments. Bilateral trade between any two Southern Africa [COMESA], ECOWAS, and WAEMU), the countries depends on their market sizes measured by correlation is either negative or low, indicating that these GDP (the equivalent of mass) and the distance between may not have been successful in stimulating extra-PTA them. Because of their empirical robustness, gravity exports. These statistics, however, are mainly useful for models have been extensively used to explain bilateral understanding some basic associations; they do not permit trade between countries and to estimate the impact of inferences about the direction of causality between growth PTAs. Although early applications of gravity models have of intra-PTA trade and exports in general, or the impor- been criticized for their lack of theoretical foundations, tance of PTAs themselves in stimulating exports. We will later studies have shown that with special assumptions, a return to the importance of intra-PTA trade below, using simpler version of the gravity model can be derived from econometric analysis to establish the relationship between the factor proportions model (Deardorff 1995), from PTA membership and trade flows. increasing returns to scale and product differentiation 60 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Figure 2.20. Evolution of the Share of Intraregional Trade in Gross Domestic Product, 1970–2008 ASEAN CACM 40 7 6 30 5 percent percent 4 20 3 10 2 1 0 0 19 0 19 3 19 6 19 9 19 2 85 88 19 1 19 4 20 7 00 20 3 20 06 08 70 19 3 76 79 82 19 5 88 19 1 94 20 7 00 03 2006 08 7 7 7 7 8 9 9 9 0 7 8 9 9 19 19 19 20 19 19 19 19 19 19 19 20 20 EFTA EU 14 50 12 40 10 percent percent 8 30 6 20 4 10 2 0 0 19 0 19 3 19 6 79 19 2 19 5 88 19 1 94 20 7 00 20 3 2006 08 19 0 19 3 19 6 79 19 2 19 5 88 19 1 94 20 7 00 20 3 2006 08 7 7 7 8 8 9 9 0 7 7 7 8 8 9 9 0 19 19 19 19 20 19 19 19 19 20 Mercosur NAFTA 5 14 12 4 10 percent percent 3 8 2 6 4 1 2 0 0 19 0 19 3 19 6 19 9 19 2 19 5 88 19 1 19 4 20 7 00 20 3 2006 08 70 73 76 79 82 85 19 8 19 1 19 4 97 00 03 2006 08 7 7 7 7 8 8 9 9 9 0 8 9 9 19 19 20 19 19 19 19 19 19 19 20 20 20 Source: International Monetary Fund (IMF), Direction of Trade Statistics (DOT). Note: ASEAN, Association of Southeast Asian Nations; CACM, Central American Common Market; EFTA, European Free Trade Association; EU, European Union; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement. models, or from a combination of both (Shelburne 2000; exporter and importer time dummies. In a refinement of Evenett and Keller 2002). the model, Baldwin and Taglioni (2006) suggest that biases Typically, a gravity model equation explains the trade from unobserved pairwise characteristics could be signifi- between two partners by using several factors: their income cant and proposed the inclusion of pair dummies to reduce levels; a vector of explanatory variables that depend on the omitted-variables bias. Following this approach, the the specific country pair but are constant over time (distance estimated gravity equation includes country-pair dummies among trading partners, dummies for a common land bor- and a time dummy instead of exporter and importer time der, a common language, a common colonizer, a current dummies. It is worth noting that the inclusion of these colonial relationship, a past colonial relationship, and an dummies precludes the use of country-pair-specific vari- index of religious similarity); a set of time- and country- ables such as distance between countries, contingency, com- pair-varying explanatory variables (membership in the mon language, and colonial relationships.28 The estimated same PTA, membership in the same currency union); and coefficients of PTA dummies are presented in table 2.3. Landscape 61 Table 2.2. Correlation between Intra-PTA and Total Export Table 2.3. Estimation Results of the Gravity Model of the Growth Rates, 1970–2008 Average Trade between Two Partners, Selected PTAs PTA Correlation coefficient Intra-PTA Extra-PTA Extra-PTA trade exports imports ANZCERTA 0.76 ASEAN 0.88 AFTA 1.556*** 0.860*** 0.745*** CACM 0.30 ANZCERTA 0.162 –0.0832** –0.0191 CAN 0.63 CACM –0.324 0.120*** 0.314*** CARICOM 0.30 CAN 0.588*** 0.120*** 0.0254 CEFTA 0.31 CARICOM –0.381*** –0.439*** –0.395*** CEMAC 0.12 CEFTA 0.0212 –0.110*** –0.0304* CIS 0.51 CEMAC 0.364 0.164*** 0.295*** COMESA –0.05 CIS –0.701*** 0.208*** 0.152*** EAC 0.33 COMESA –0.155*** –0.318*** –0.269*** ECOWAS 0.03 EAC 1.221*** 0.0545** 0.137*** EFTA 0.96 ECOWAS 0.514*** 0.0484** 0.171*** EU 0.98 EFTA 0.503*** 0.149*** 0.0768*** Euromed 0.89 EU 0.472*** 0.108*** 0.00367 GCC 0.91 Euromed 0.124*** 0.346*** 0.217*** Mercosur 0.55 GCC 0.374*** 0.413*** 0.303*** NAFTA 0.86 Mercosur 0.689*** 0.741*** 0.645*** PATCRA 0.48 NAFTA 0.878*** 0.192*** 0.175*** SADC 0.33 PATCRA 0.339 0.158*** 0.182*** SAFTA 0.33 SADC 0.842*** –0.134*** –0.0332 WAEMU/UEMOA –0.08 SAFTA 0.466*** 0.506*** 0.542*** WAEMU/UEMOA 0.465*** 0.123*** 0.00836 Source: International Monetary Fund (IMF), Direction of Trade Statistics (DOT); authors’ calculations. Source: Authors’ calculations. Note: ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Note: AFTA, ASEAN Free Trade Area; ANZCERTA, Australia–New Zealand Agreement; ASEAN, Association of Southeast Asian Nations; CACM, Closer Economic Relations Trade Agreement; ASEAN, Association of Central American Common Market; CAN, Andean Community; CARICOM, Southeast Asian Nations; CACM, Central American Common Market; CAN, Caribbean Community; CEFTA, Central European Free Trade Agreement; Andean Community; CARICOM, Caribbean Community; CEFTA, Central CEMAC, Economic and Monetary Community of Central Africa European Free Trade Agreement; CEMAC, Economic and Monetary (Communauté Économique et Monétaire de l’Afrique Centrale); CIS, Community of Central Africa (Communauté Économique et Monétaire de Commonwealth of Independent States; COMESA, Common Market for l'Afrique Centrale); CIS, Commonwealth of Independent States; COMESA, Eastern and Southern Africa; EAC, East African Community; ECOWAS, Common Market for Eastern and Southern Africa; EAC, East African Economic Community of West African States; EFTA, European Free Trade Community; ECOWAS, Economic Community of West African States; EFTA, Association; EU, European Union; Euromed, European Union–Mediterranean European Free Trade Association; EU, European Union; Euromed, European Free Trade Area; GCC, Gulf Cooperation Council; Mercosur, Southern Cone Union–Mediterranean Free Trade Area; GCC, Gulf Cooperation Council; Common Market (Mercado Común del Sur); NAFTA, North American Free Mercosur, Southern Cone Common Market (Mercado Común del Sur); Trade Agreement; PATCRA, Papua New Guinea–Australia Trade and NAFTA, North American Free Trade Agreement; PATCRA, Papua New Commercial Relations Agreement; SADC, Southern African Development Guinea–Australia Trade and Commercial Relations Agreement; SADC, Community; SAFTA, South Asian Preferential (Free) Trade Arrangement; Southern African Development Community; SAFTA, South Asian Preferen- WAEMU/UEMOA, West African Economic and Monetary Union/Union tial (Free) Trade Arrangement; WAEMU/UEMOA, West African Economic Économique et Monétaire Ouest-Africaine. and Monetary Union/Union Économique et Monétaire Ouest-Africaine. * p < 10% ** p < 5% *** p < 1%. Overall, we find that the impact on intra-PTA trade of infer whether it was the PTA alone that led to those poli- the PTAs covered in this study is positive. Almost all the cies. Again, AFTA seems to be the most successful, with the PTA members seem to trade more with each other than highest impact on bilateral trade (see figure 2.23). By con- would otherwise be expected from a typical trading rela- trast, CARICOM, CIS, and COMESA members seem to be tionship between countries with similar incomes and other trading with each other less than a normal trading relation- characteristics. Our statistical analysis indicates that AFTA, ship would predict. In the case of COMESA, the negative CAN, EAC, ECOWAS, EFTA, EU, GCC, Mercosur, NAFTA, impact is rather small. Intra-COMESA trade is estimated to SADC, SAFTA, and WAEMU trade more internally than have been 14 percent smaller than expected under “normal� would be expected in the absence of a PTA.29 The countries trading conditions.30 The strong negative impact on CIS that belong to those PTAs seem to have adopted policies bilateral trade can be understood in the light of the disman- more conducive to bilateral trade expansion, but we cannot tling of the Soviet Union and increasing openness to the rest 62 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard Figure 2.21. Most Favored Nation Applied Tariffs, Trade-Weighted Average of All PTA Members Selected Periods 25 20 15 percent 10 5 0 N CM N M CO C EC AC AS 7 M CC CU U / DC A A TA TA A r su -2 FT ES O A EA CA CO W EF AF M E co G SA M SA EU CA M CE AS O CE N UE er RI CA M AE W 1995–99 2000–04 2005–08 Source: World Bank, World Trade Indicators database. Note: ASEAN, Association of Southeast Asian Nations; CACM, Central American Common Market; CAN, Andean Community; CARICOM, Caribbean Community; CEFTA, Central European Free Trade Agreement; CEMAC, Economic and Monetary Community of Central Africa (Communauté Économique et Monétaire de l’Afrique Centrale); COMESA, Common Market for Eastern and Southern Africa; EAC, East African Community; ECOWAS, Economic Community of West African States; EFTA, European Free Trade Association; EU-27, European Union after 2007 (27 members); GCC, Gulf Cooperation Council; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement; SACU, Southern African Customs Union; SADC, Southern African Development Community; WAEMU/UEMOA, West African Economic and Monetary Union/Union Économique et Monétaire Ouest-Africaine. Figure 2.22. Proportion of Tariff-Free Imports as a Share of Total Imports, All Goods, Selected PTAs and Periods 90 80 70 60 percent 50 40 30 20 10 0 N CM CA CAN M CO AC C AS 5 7 CC CU U/ DC CE A A TA TA A r su -1 -2 FT ES O EA EA CO W EF AF M co G SA M M SA EU EU CA M CE AS O N UE er RI EC M AE W 1995–99 2000–04 2005–08 Source: World Bank, World Trade Indicators database. Note: ASEAN, Association of Southeast Asian Nations; CACM, Central American Common Market; CAN, Andean Community; CARICOM, Caribbean Community; CEFTA, Central European Free Trade Agreement; CEMAC, Economic and Monetary Community of Central Africa (Communauté Économique et Monétaire de l’Afrique Centrale); COMESA, Common Market for Eastern and Southern Africa; EAC, East African Community; ECOWAS, Economic Community of West African States; EFTA, European Free Trade Association; EU-15, European Union before the 2004 enlargement (15 members); EU-27, European Union after 2007 (27 members); GCC, Gulf Cooperation Council; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement; SACU, Southern African Customs Union; SADC, Southern African Development Community; WAEMU/UEMOA, West African Economic and Monetary Union/Union Économique et Monétaire Ouest-Africaine. Tariffs include ad valorem rates and ad valorem equivalents of specific tariffs. Landscape 63 Figure 2.23. Percentage Changes in Trade from Entry into Force of a Preferential Trade Agreement to 2008, Based on Gravity Model Estimates 4.0 3.5 3.0 2.5 2.0 percent 1.5 1.0 0.5 0 –0.5 –1.0 CA N CM RI N M AC CO S C AS ro EU ed RO M CC AS A CE A A TA PA TA RA M S C UE TA A N r su CI T FT ES O EA D EA CA CO -M W ER EF AF TC U/ AF M co EU G M SA M CE O ZC er EC CA Eu AN AE W intra-PTA trade extra-PTA exports extra-PTA imports Source: Authors’ calculations. Note: ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Agreement; ASEAN, Association of Southeast Asian Nations; CACM, Central American Common Market; CAN, Andean Community; CARICOM, Caribbean Community; CEFTA, Central European Free Trade Agreement; CEMAC, Economic and Monetary Community of Central Africa; CIS, Commonwealth of Independent States; COMESA, Common Market for Eastern and Southern Africa (Communauté Économique et Monétaire de l’Afrique Centrale); EAC, East African Community; ECOWAS, Economic Community of West African States; EFTA, European Free Trade Association; EU, European Union; Euromed, European Union–Mediterranean Free Trade Area; GCC, Gulf Cooperation Council; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement; PATCRA, Papua New Guinea–Australia Trade and Commercial Relations Agreement; SADC, Southern African Development Community; SAFTA, South Asian Preferential (Free) Trade Arrangement; WAEMU/UEMOA, West African Economic and Monetary Union/Union Économique et Monétaire Ouest-Africaine. The bars show the magnitude of the impact of the respective PTAs on intra-PTA trade, calculated on the basis of statistically significant dummy variables from table 2.3. of the world. Figure 2.23 translates the estimates presented trade diversion.31 These results are broadly consistent in table 2.3 into the percentage impact on trade following with the findings of the previous sections and Baldwin the introduction of the PTA, up to 2008. For example, it is and Freund, ch. 6 in this volume indicating that PTAs estimated that trade between AFTA members from its with higher external barriers are likely to be associated founding to 2008 was three times higher than would have with trade diversion. been predicted by a typical trading relationship between countries with similar income and other characteristics. Conclusions We also analyze the impact of the introduction of PTAs on extra-PTA exports and imports and find that Today’s multilateral trading system is characterized by a most of the PTAs considered here have been trade creat- multiplicity of overlapping and intersection preferential ing. The agreements whose members trade more with trade agreements. The number of PTAs has grown the rest of the world than would have been expected markedly since 2000. PTAs offer members a means of from a normal trading relationship include AFTA, securing enhanced market access and attracting invest- Mercosur, and SAFTA. Members of CARICOM, CEFTA, ment, but at the cost of rendering the trading landscape COMESA, and SADC, however, are trading less with the less transparent and more unpredictable. Each PTA tends rest of the world than our gravity model would have pre- to create its own web of regulatory rules that coexist along- dicted in the absence of PTAs, indicating some degree of side multilateral rules. 64 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard The PTA landscape shows a number of distinctive different regions. This development risks generating fur- features: ther fragmentation of global trading rules, because each PTA maintains its own distinct regulatory framework. • Ubiquity. PTAs have spread to most geographic regions, • Scope. The regulatory scope of PTAs is broadening and especially East Asia and the Pacific, and North-South deepening. Increasingly, PTAs include a services com- preferential partnerships are often chosen over unilat- ponent, as well as commitments that exceed those they eral programs, such as the GSP. have accepted at the multilateral level or that lie outside • Consolidation. In some cases, bilateral relationships are the current WTO mandate. being replaced by plurilateral PTAs among the same • Impact. Analysis indicates that for a number of plurilat- partners. eral PTAs, PTA partners trade more internally than • Nature. Bilateral PTAs are increasingly the norm. This would be expected in the absence of a PTA and the movement corroborates the observation of a shift away impact on extra-PTA trade is largely positive. Similar from the use of PTAs to forge traditional regional partner- calculations to determine the effects of bilateral PTAs ships among geographically proximate countries and would be useful. In addition, the effects of PTAs on toward their employment as instruments for negotiating stimulating trade in services and investment is an area strategic, bilateral market access, often among countries in worthy of future study. Annex Annex Table 2A.1. Membership of Selected Plurilateral Preferential Trade Agreements Abbreviation Name of PTA Members AFTA ASEAN Free Trade Area Brunei Darussalam, Cambodia, Indonesia, Lao People’s Democratic Republic, Malaysia, Myanmar, Philippines, Singapore, Thailand, Vietnam ANZCERTA Australia–New Zealand Closer Economic Australia, New Zealand Relations Trade Agreement APTA Asia-Pacific Trade Agreement Bangladesh, China, India, Republic of Korea, Lao PDR, Sri Lanka (Bangkok Agreement) ASEAN Association of Southeast Asian Nations Brunei Darussalam, Cambodia, Indonesia, Lao PDR, Malaysia, Myanmar, Philippines, Singapore, Thailand, Vietnam CACM Central American Common Market Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua CAFTA-DR Central America–Dominican Costa Rica, Dominican Republic, El Salvador, Guatemala, Honduras, Republic–United States Nicaragua, United States CAN Andean Community Bolivia, Colombia, Ecuador, Peru CARICOM Caribbean Community Antigua and Barbuda, The Bahamas, Barbados, Belize, Dominica, Grenada, Guyana, Haiti, Jamaica, Montserrat, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, Suriname, Trinidad and Tobago CEFTA Central European Free Trade Agreement Albania, Bosnia and Herzegovina, Croatia, former Yugoslav Republic of Macedonia, Moldova, Montenegro, Serbia, United Nations Interim Administration Mission in Kosovo (UNMIK) CEMAC Economic and Monetary Community of Cameroon, Central African Republic, Chad, Republic of Congo, Central Africa/Communauté Économique Equatorial Guinea, Gabon et Monétaire de l’Afrique Centrale CEZ Common Economic Zone Belarus, Kazakhstan, Russian Federation, Ukraine CIS Commonwealth of Independent States Armenia, Azerbaijan, Belarus, Kazakhstan, Kyrgyz Republic, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine, Uzbekistan COMESA Common Market for Eastern and Burundi, the Comoros, Democratic Republic of Congo, Djibouti, Southern Africa Arab Republic of Egypt, Eritrea, Ethiopia, Kenya, Libya, Madagascar, Malawi, Mauritius, Rwanda, the Seychelles, Sudan, Swaziland, Uganda, Zambia, Zimbabwe EAC East African Community Burundi, Kenya, Rwanda, Tanzania, Uganda (continued next page) Landscape 65 Annex Table 2A.1. (continued) Abbreviation Name of PTA Members EAEC Eurasian Economic Community Belarus, Kazakhstan, Kyrgyz Republic, Russia, Tajikistan ECO Economic Cooperation Organization Afghanistan, Azerbaijan, Islamic Republic of Iran, Kazakhstan, Kyrgyz Republic, Pakistan, Tajikistan, Turkey, Turkmenistan, Uzbekistan ECOWAS Economic Community of West Benin, Burkina Faso, Cape Verde, Côte d’Ivoire, The Gambia, African States Ghana, Guinea, Guinea-Bissau, Liberia, Mali, Niger, Nigeria, Senegal, Sierra Leone, Togo EEA European Economic Area European Union, Iceland, Liechtenstein, Norway EFTA European Free Trade Association Iceland, Liechtenstein, Norway, Switzerland EU European Union Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal, Romania, Slovak Republic, Slovenia, Spain, Sweden, United Kingdom GCC Gulf Cooperation Council Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, United Arab Emirates LAIA/ALADI Latin American Integration Association/ Argentina, Bolivia, Brazil, Chile, Colombia, Cuba, Ecuador, Mexico, Asociación Latinoamericana de Integración Paraguay, Peru, Uruguay, República Bolivariana de Venezuela Mercosur Southern Cone Common Market/ Argentina, Brazil, Paraguay, Uruguay Mercado Común del Sur MSG Melanesian Spearhead Group Fiji, Papua New Guinea, Solomon Islands, Vanuatu NAFTA North American Free Trade Agreement Canada, Mexico, United States OCT Overseas Countries and Territories Anguilla, Aruba, British Antarctic Territory, British Indian Ocean Territory, British Virgin Islands, Cayman Islands, Falkland Islands, French Polynesia, French Southern and Antarctic Territories, Greenland, Mayotte, Montserrat, Netherlands Antilles, New Caledonia, Pitcairn, Saint Helena, Saint Pierre and Miquelon, South Georgian and South Sandwich Islands, Turks and Caicos Islands, Wallis and Futuna Islands PAFTA Pan-Arab Free Trade Area Algeria, Bahrain, Egypt, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Palestinian Authority, Qatar, Saudi Arabia, Sudan, Syrian Arab Republic, Tunisia, United Arab Emirates, Republic of Yemen PATCRA Papua New Guinea–Australia Trade and Australia, Papua New Guinea Commercial Relations Agreement PICTA Pacific Island Countries Trade Agreement Cook Islands, Fiji, Kiribati, Federated States of Micronesia, Nauru, Niue, Papua New Guinea, Samoa, Solomon Islands, Tonga, Tuvalu, Vanuatu PTN Protocol relating to Trade Negotiations Bangladesh, Brazil, Chile, Egypt, Israel, Republic of Korea, Mexico, among Developing Countries Pakistan, Paraguay, Peru, Philippines, Romania, Tunisia, Turkey, Uruguay, former Yugoslavia SACU Southern African Customs Union Botswana, Lesotho, Namibia, South Africa, Swaziland SADC Southern African Development Community Angola, Botswana, Democratic Republic of Congo, Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Seychelles, South Africa, Swaziland, Tanzania, Zambia, Zimbabwe SAPTA/SAFTA South Asian Preferential (Free) Trade Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, Sri Lanka Arrangement SPARTECA South Pacific Regional Trade and Economic Australia, Cook Islands, Fiji, Kiribati, Marshall Islands, Federated States of Cooperation Agreement Micronesia, Nauru, New Zealand, Niue, Papua New Guinea, Samoa, Solomon Islands, Tonga, Tuvalu, Vanuatu Trans-Pacific Trans-Pacific Strategic Economic Partnership Brunei Darussalam, Chile, New Zealand, Singapore SEP WAEMU/ West African Economic and Monetary Benin, Burkina Faso, Côte d’Ivoire, Guinea-Bissau, Mali, UEMOA Union/Union Économique et Monétaire Niger, Senegal, Togo Ouest-Africaine Source: Authors’ compilation. Notes reviews; and other public sources. The information may not be exhaustive because, whereas it is possible to account accurately for all notified PTAs, 1. The information used in this study is based on several sources: information on nonnotified PTAs, agreements under negotiation, and notifications to the WTO; documentation submitted to the WTO’s Com- agreements in the proposal stage is often scarce or inconclusive. mittee on Regional Trade Agreements (CRTA) or the Committee on Trade 2. There is, of course, no obligation for nonmembers of the WTO to and Development (CTD); WTO accession documents; trade policy notify their PTAs, but PTAs that involve both members and nonmembers 66 Rohini Acharya, Jo-Ann Crawford, Maryla Maliszewska, and Christelle Renard of the WTO should be notified by the respective WTO member or 16. The Mediterranean parties are Algeria, the Arab Republic of members. Egypt, Israel, Jordan, Lebanon, Morocco, the Palestinian Authority, the 3. According to the WTO’s Transparency Mechanism for Regional Syrian Arab Republic, and Tunisia. The agreement with Syria was adopted Trade Agreements (WTO document WT/L/671), notification is to take by the EU Council on October 27, 2009; signature of the agreement awaits place as early as possible, no later than directly following the parties’ rati- further notice from Syria (European Commission, Overview of PTA and fication or application, and before the application of preferential treat- other Trade Negotiations, February 5, 2010). ment between the parties. 17. Ukraine is also negotiating PTAs with EFTA, the EU, and Singa- 4. There is sometimes a significant time lag between the entry into pore, and an agreement with Canada has been proposed. force of a PTA and its notification. For instance, of the 37 PTA notifica- 18. In addition to its plurilateral agreements such as NAFTA, tions received in 2009, 12 relate to PTAs that entered into force before LAIA/ALADI, the Protocol relating to Trade Negotiations among Devel- 2009. Delays in notification may be caused by lengthy ratification proce- oping Countries, the Global System of Trade Preferences among Develop- dures, countries’ lack of knowledge of notification obligations, indecision ing Countries (GSTP), and the Mexico-Northern Triangle agreement about which legal provision (for example, Article XXIV or the Enabling (with El Salvador, Guatemala, and Honduras), Mexico has eight PTAs and Clause) to invoke (in the case of PTAs covering trade in goods), or recent another eight agreements under the LAIA/ALADI framework. accession to the WTO, creating notification obligations. 19. Panama’s agreements with Guatemala and Nicaragua are in force 5. WTO statistics are based on notification requirements that do not but have not yet been notified to the WTO. reflect the physical numbers of PTAs and so tend to overstate the total 20. Colombia also has agreements with Cuba, Mercosur, and Mexico number of PTAs. Notifications to the WTO include those made under under the LAIA/ALADI framework. GATT Article XXIV, General Agreement on Trade in Services Article V, 21. The agreements are with the Andean Community, Bolivia, Chile, and the Enabling Clause, as well as accessions to existing PTAs. The noti- Colombia, Cuba, Ecuador, Mexico, Peru, and República Bolivariana de fication requirements contained in WTO provisions require that PTAs Venezuela. covering trade in goods and services be notified separately. For a com- 22. The agreements are with Bolivia, Cuba, Ecuador, Mercosur, Peru, plete list of PTAs notified to the GATT/WTO and in force, see the WTO’s and the República Bolivariana de Venezuela, Regional Trade Agreements Database (http://rtais.wto.org/). The total 23. Thailand has agreements in force with Australia, Japan, the Lao number of notified PTAs in force minus economic integration agree- PDR, and New Zealand. Negotiations are ongoing concerning the ments in services and accessions to existing PTAs yields the number of Bangladesh, India, Myanmar, Sri Lanka, and Thailand Economic Cooper- physical agreements. ation (BIMSTEC) and with Chile, EFTA, India, and the United States. 6. Before the establishment of the WTO in 1995 and the negotiation Malaysia has agreements in force with Japan and Pakistan and is currently of the General Agreement on Trade in Services (GATS), there was no obli- negotiating with Australia, Chile, EFTA, India, New Zealand, and the gation to notify PTAs with services provisions. United States. 7. The World Bank’s classification of developed countries includes all 24. Nowhere is overlapping membership more problematic than in countries defined as high income, whether members of the Organisation simultaneous membership of multiple customs unions. For instance, for Economic Co-operation and Development (OECD) or non-OECD. Tanzania is a member of the East African Community (a customs union) This differs from the WTO classification, in which developing-country and of SADC, a group that plans to become a customs union. Similarly, status is based on WTO members’ self-selection. several countries are members of the SADC and of the Common Market 8. Information on PTAs under negotiation and signed is based on for Eastern and Southern Africa (COMESA), both of which plan to available figures and is accurate to the best of our knowledge. Infor- become customs unions. mation on some geographic regions, such as Central Asia and the Mid- 25. For example, Tanzania is negotiating under the EU-SADC EPA, dle East, is difficult to obtain and, thus, can only be considered a best although it is a member of the EAC, and a number of SADC and estimate. COMESA members are negotiating together under the EU-ESA EPA. 9. Bilateral PTAs may include more than two countries when one of (The Comoros, Djibouti, Eritrea, Ethiopia, and Sudan are COMESA the parties is itself a PTA. For example, for our purposes, EFTA-Chile is members; Madagascar, Mauritius, the Seychelles, Zambia, and Zimbabwe considered a bilateral PTA, although it involves the four countries of are members of the SADC.) EFTA plus Chile. A plurilateral PTA refers to one in which the number of 26. This simple measure ignores the issue of preference utilization; constituent parties exceeds two countries, such as Mercosur, CARICOM, that is, in some instances intra-PTA trade might not be conducted on a and the SADC. preferential basis. Furthermore, the growth of intra-PTA imports might, 10. The World Bank regions used in this study are as follows: East in some cases, be associated with trade diversion. Asia and Pacific; Europe and Central Asia; North America, Latin America, 27. The reduction of external protection could be associated with the and the Caribbean; Middle East and North Africa; South Asia; and Sub- creation of this particular PTA or with other bilateral or plurilateral liber- Saharan Africa. The regional classification of countries differs somewhat alization. from that used in the WTO framework. 28. The sample covers all 179 countries available in the International 11. This section relies heavily on Horn, Mavroidis, and Sapir (2010). Monetary Fund Direction of Trade statistics over the period 1970–2008. 12. Figure 2.12 shows the inclusion of such provisions in RTAs GDP data originate from the IMF’s World Economic Outlook database. whether or not the commitments go beyond existing commitments under Dummies are included for the membership of the following PTAs: the relevant WTO agreements. In addition, such provisions may or may ASEAN, CACM, CAN, CARICOM, CEFTA, CEZ, CIS, COMESA, EAC, not be legally enforceable. ECOWAS, EFTA, EU, GCC, Mercosur, NAFTA, PATCRA, SADC, SAFTA, 13. Until the conclusion of the Lisbon agreement, EU law did not and WAEMU/UEMOA. cover investment provisions with third countries. Instead, this was a bilat- 29. These results are consistent with the gravity estimates of the eral issue like double-taxation agreements. impact of PTAs on intraregional trade found in Global Economic Prospects 14. The enlargement led to the termination of a number of bilateral 2005 (World Bank 2005). Although the World Bank analysis covers an ear- agreements between the EU-25 (referring to the EU before the latest lier time period (1960–2000) and the gravity model specification is differ- enlargement) and the new member states, and between the two new ent, the relative values of the coefficients on intra-PTA trade are similar to member states (Bulgaria and Romania) and third parties. The termina- our findings. tion of these agreements was notified to the WTO (WT/REG/GEN/N/4). 30. The impact on trade is equal to e–0.155 –1 = –14 percent. 15. The termination of these agreements was notified to the WTO 31. The result for CEFTA is not easy to interpret because membership (WT/REG/GEN/N/5). of the bloc has changed dramatically with EU enlargements. Landscape 67 References Trading System, ed. Richard Baldwin and Patrick Low. Cambridge, U.K.: Cambridge University Press. Baldwin, Richard, and Daria Taglioni. 2006. “Gravity for Dummies and Horn, Henrik, Petros C. Mavroidis, and André Sapir. 2010. Beyond the Dummies for Gravity Equations.� NBER Working Paper 12516, WTO? An Anatomy of EU and US Preferential Trade Agreements. The National Bureau of Economic Research, Cambridge, MA. World Economy 33(11), 1565–88. Batista, Jorge Chami. 2008. “Competing for the US Import Market: NAFTA Sapir, André. 2007. “Europe and the Global Economy.� In Fragmented and Non-NAFTA Countries.� In Economic Integration in the Americas, Power: Europe and the Global Economy, ed. André Sapir. Brussels: ed. Joseph A. McKinney and H. Stephen Gardner. London: Routledge. Bruegel. http://www.bruegel.org/publications/show/publication/ Deardorff, A. V. 1995. “Determinants of Bilateral Trade: Does Gravity fragmented-power-europe-and-the-global-econom.html. Work in a Neoclassical World?� NBER Working Paper 5377, National Shelburne, R. C. 2000. Intra-Industry Trade, the Gravity Model and Simi- Bureau of Economic Research, Cambridge, MA. larity in Endowments and Country Size. Papers and Proceedings of the Evenett, S. J., and W. Keller. 2002. “On Theories Explaining the Success of International Trade and Finance Association, San Diego State Univer- the Gravity Equation.� Journal of Political Economy 110: 281–316. sity, San Diego, CA. Fiorentino, Roberto V., Jo-Ann Crawford, and Christelle Toqueboeuf. World Bank. 2005. Global Economic Prospects 2005: Trade, Regionalism, 2009. “The Landscape of Regional Trade Agreements and WTO Sur- and Development. Washington, DC: World Bank. veillance.� In Multilateralizing Regionalism: Challenges for the Global 3 ECONOMICS Richard Baldwin Preferential trade agreements (PTAs) have been important preferences (Viner’s ambiguity). Issues related to scale features of the world trade system since the inception in economies and procompetitive effects are then explored. 1947 of the General Agreement on Tariffs and Trade Finally, the discussion turns to the additional benefits that (GATT), which was succeeded in 1995 by the World Trade PTAs may offer in terms of learning effects, increased for- Organization (WTO). For the first 50 years of the eign investment, the provision of regional public goods GATT/WTO, PTAs dealt mainly, and often exclusively, with such as transport networks, and other positive externali- preferential tariff reduction. It is therefore natural that ties. The annexes to the chapter contain technical explana- most economic analyses of PTAs focused on preferential tions for interested readers. tariffs. This view, however, has come to be outdated, as dis- cussed by Chauffour and Maur, ch.1 in this volume. The GATT/WTO’s success in cutting developed coun- Basic Economic Effects of Regionalism tries’ most favored nation (MFN) tariffs, coupled with rapid The terminology behind the basic economics of discrimi- unilateral MFN tariff reductions by developing countries natory liberalization is marked by a conceptual tangle of (Baldwin 2010b), means that today’s margins of tariff pref- conflicting, overlapping, and competing terminologies. To erences are relatively small. Carpenter and Lendle (2010), set the stage for discussion and analysis, we cover the three for instance, show that whereas almost half of world trade elemental effects at work. All have been known at least is between countries that share a PTA, only 16 percent is since 1950. To avoid creating yet another set of terms, we eligible for preferences, and most of these preferences are label them according to their intellectual fathers. small; only 2 percent of world trade outside the EU has margins over 10 percent. The reason is that much of world trade is now granted MFN duty-free status, so that Smith’s Certitude, Haberler’s Spillover, and preference margins are zero by definition. Moreover, the Viner’s Ambiguity preference margins that do exist are typically small. Less Early contributions to the theory of preferential trade than 2 percent of world imports, excluding intra-EU trade, include Adam Smith, Robert Torrens, and Frank Taussig is eligible for preference margins that exceed 10 percent (see Pomfret 1997 for discussion and references). One of (Carpenter and Lendle 2010). As a result, analysis based on the most robust findings in the field is what might be called the traditional, preference-centric perspective is no longer “Smith’s certitude�: sufficient. Nevertheless, the traditional view remains the core of received wisdom on PTAs, and so the bulk of this When a nation “exempt[s] the good of one country from chapter deals with the economics of tariff preferences. duties to which it subjects those of all others . . . the mer- Specifically, this chapter examines the economics and chants and manufacturers of the country whose commerce is so favored must necessarily derive great advantage.� political economy of preferential trade liberalization, start- (Pomfret 1997, quoting Smith 1776) ing with the elemental effects of preferential tariff cutting: the gain to partners receiving the preference (Smith’s Much later, Gottfried Haberler (1936, 384) asserted certitude); the loss to third countries that must compete that all members of a preferential trade agreement must with the preference-receiver (Haberler’s spillover); and the gain, while third nations must lose. We now know that ambiguous welfare impact on the nation granting the the first part of the assertion is wrong, but what might be 69 70 Richard Baldwin called “Haberler’s spillover�—the part about third nations and factors, given tastes, technology, and endowments). losing—turns out to be almost as robust as Smith’s certitude. The action of the free market, according to the first wel- Haberler’s spillover and Smith’s certitude are the linchpins of fare theorem of Walrasian economics, guarantees an the political economy of preferential tariff cutting. increase in economic efficiency, and lump-sum transfers The only basic element added in the postwar period within the PTA ensure welfare gains for all. Third nations came with Jacob Viner’s famous 1950 book, The Customs are not affected because their trade vectors do not change. Union Issue. His key finding was that preferential tariff lib- Dixit and Norman (1980) generalize the analysis, show- eralization has ambiguous welfare effects on the preference- ing that the Kemp-Wan improvement can be obtained granting nation (“Viner’s ambiguity�). Viner’s ambiguity without lump-sum transfers; intra-PTA commodity taxes is quite general, yet one is hard pressed to see this from and subsidies are sufficient. the analysis in his book. A preferential tariff is nothing Of course, real-world PTAs do not adjust external tariffs more than a special case of nonuniform commodity taxa- in a Kemp-Wan manner, nor do they have access to large tion, but Viner did not have the benefit of modern economic lump-sum transfers. Nevertheless, the theorem is impor- tools for tax analysis. Rather, he relied on the enduring but tant from a policy perspective; it proves that PTAs are not imprecise concepts of “trade diversion� and “trade cre- necessarily bad for world welfare. Moreover, it helps us ation.� These terms are misleading because they suggest that think about why the combination of multilateral and pref- trade volumes are the key, even though Viner’s words clearly erential tariff cutting, which has been in operation since indicate that cost changes are what matter. Moreover, they the 1950s, has had a relatively benign to date. fail to cover all the effects generated by discriminatory tariff liberalization, even in a simple Walrasian setting. Given Illustration of Basic Economic Effects these shortcomings, and the decades-long debate on “what Viner really meant� (a debate in which Viner himself partic- Smith’s certitude, Haberler’s spillover, and Viner’s ambiguity ipated, without notable effect), it is curious that the terms capture most of the basic economics of PTAs and, together have enjoyed such enduring success.1 with the Kemp-Wan logic, most of the political-economy The generality of Viner’s ambiguity is glaringly obvious reasoning in the “big-think� regionalism literature.2 It is to readers schooled in the theory of the second-best (pref- possible to deal with these mathematically. However, to erential liberalization induces new distortions while demonstrate the basic interactions among the elemental removing others), but Viner’s book was a landmark. The effects and to facilitate the subsequent exposition of the theory of the second-best was unknown in 1950, and many logic of the big-think regionalism literature, a graphical of Viner’s contemporaries—Haberler, for example—were analysis may be useful. muddled about the essential differences between general The simplest framework that meets the requirements is and preferential liberalization. a Walrasian three-nation model (Home, Partner, and Rest A fourth elemental effect in the regionalism literature of the World [RoW]) with three goods, numbered 1, 2, and concerns the interaction between preferential and multilat- 3. Each nation exports two goods and imports the other eral tariff cutting. It is not really a basic economic effect good (figure 3.1). Since each nation has two sources of but, rather, a specific combination of effects brought to prominence by the fact that the most important regional Figure 3.1. Trade Pattern for a Simple Preferential Trade Association liberalizations over the past 60 years have been accompa- nied by multilateral liberalization. When thinking about this teaming of multilateral and regionalism liberalization, the Good 1 guiding light is the Kemp-Wan logic, which is discussed next. Good 2 Partner Home James Meade (1955) introduced analysis that led to one of the few general statements that can be made about Good 1 Good 2 PTAs—the Kemp-Wan theorem. Kemp and Wan (1976) demonstrated that PTAs could be designed to be Pareto improving for every member of the PTA and for the world at large. The logic is elegant. Assume that two nations sign Good 3 Good 3 a PTA and alter their external tariffs to freeze their external trade flows. The external trade flows can then be treated as RoW part of the bloc’s endowment. Removal of all intra-PTA barriers thus shifts the two-nation bloc from a second-best situation to a first-best situation (i.e., laissez-faire in goods Source: Author’s elaboration. Economics 71 imports, tariff discrimination can be a real issue in all exporter and Partner is the importer. Nothing happens in markets. To rule out Meade’s secondary effects, tastes are the market for Good 3, where RoW is the importer: RoW assumed to be identical across nations and additively sepa- maintains its MFN tariff, and the strong separability rable in all goods.3 For simplicity’s sake, the three nations assumptions rule out Meade’s secondary and tertiary are symmetric in size and with respect to the most favored effects. nation (MFN) tariff they initially impose.4 We see Smith’s certitude and Haberler’s spillover imme- The two trading equilibriums (regionalism versus diately in figure 3.2. Smith’s certitude shows up as Partner multilateral free trade) in a typical market (Good 1) can gains a + b from the higher border price and the expansion be worked out with the help of the PTA diagram shown of exports. Since the FTA is reciprocal and nations are sym- in figure 3.2. The analyses for imports of Good 2 (into metric, Home gains the same in Good 2. Haberler’s Partner) and Good 3 (into RoW) are isomorphic because spillover shows up in the fact that RoW loses area e because of the strong symmetry. of the drop in the border price it faces (from P – T to P' – T) The diagram shows the export supply curves (marked and the reduction of its exports to Home and Partner. XS, with the appropriate superscript to indicate the origin The preference rent. A critical observation, as far as nation) for Home’s two potential suppliers. The horizontal big-think regionalism is concerned, touches on a decom- sum of the XS curves is shown as MSFT, along with Home’s position of Smith’s certitude—that is, how PTA-based import demand curve, MD. Under global free trade, the exporters gain from two distinct features of their improved domestic and border price is PFT, as shown in all nations market access. First, the removal of the intra-PTA tariff for all goods. Assuming that all nations impose a specific boosts their market access directly. Second, PTA-based tariff T on an MFN basis, the internal price in Home exporters benefit from the reduction in RoW exports is driven up to P, while the border price is driven down to induced by the tariff discrimination. The second part of P – T for both suppliers. Home imports drop, with the the gain—area a in figure 3.2—could be called the prefer- reduction divided equally between the two suppliers. ence rent because if the tariff cutting were multilateral From MFN tariffs to preferential trade agreements. If a instead of preferential, PTA partners would gain only b, PTA is formed between Home and Partner, the total not a + b. This preference rent a is vulnerable to so-called import supply curve becomes the kinked MSPTA curve.5 preference erosion and thus plays a leading role in countries’ The resulting internal price falls to P', but there are now resistance to moving from preferential tariff liberalization to two border prices. The border price facing Partner-based multilateral liberalization. exporters rises from P – T to P', while the border price fac- On the import side (figure 3.3), Home gains a trade- ing RoW firms falls from P – T to P' – T. Partner exports volume effect, equal to area A, from expanding its imports; expand, and RoW exports contract. Identical things hap- that is, from replacing high-cost domestic production with pen in the market for Good 2, but here Home is the lower-cost imports. Home also gains from a border-price Figure 3.2. Trading Equilibriums in a Preferential Trade Association border price border price internal price MSMFN XSR XSP MSPTA P P’ P’ MSFT c a P FT P FT d b P–T P–T e P’ – T MD T XR’ RoW XP’ Partner Home exports exports M exports Source: Author’s modification of figure in Baldwin and Wyplosz 2009, ch. 5. 72 Richard Baldwin Figure 3.3. Ambiguous Net Welfare Effects Standards and regulations can influence economic activity and the structure of industry in ways too numer- border price domestic price ous to count. The World Bank, for instance, has conducted XSP A a large multiyear project studying the impact of standards D2 P on trade.7 To focus on essentials, however, we abstract from P’ many details and model TBTs as frictional barriers; that is, D1 C1 barriers that raise the marginal cost of selling a good inter- P–T nationally.8 The effect of such frictional barriers is similar B P’ – T C2 to that of a tariff except that no tariff revenue or other form MD of rent is generated. The gap between the prices of the good inside the importing nation and inside the exporting NB: C1 = D1 nation is burnt up by costly activities that are required to satisfy the TBT. X X’ PTA XR’ XR M Home To keep things simple, consider again a three-nation exports imports world with TBT liberalization between two countries, Source: Author’s elaboration. Home and Partner; the third nation (RoW) remains out- side the arrangement. We start by assuming that all three nations impose TBTs that have a specific-tariff equivalent of T. The policy change to be studied is a lowering of T to effect—the terms-of-trade improvement against RoW zero on all trade between Home and Partner, with no (area B)—while losing from the terms-of-trade loss against change in the barriers on trade between RoW and Home or Partner (area C1 + C2). Home’s terms-of-trade gain on the between Partner and RoW. export side partly offsets the terms-of-trade loss on the The price and quantity effects of the preferential liberal- import side (D1 = C1), so Home’s net welfare change is A + B ization are very similar to those discussed above in the case + D2 – C2.6 As drawn, it looks as though Home and Partner of tariff liberalization; the only change concerns the border gain, but whether they in fact do depends on elasticities and price. With frictional barriers, the domestic price is the the initial MFN tariff. In general, Viner’s ambiguity holds in border price for the importing nation, so the liberalization this framework. As can be demonstrated mathematically, the lowers Home’s border price. At the same time, the exporter PTA reduces welfare when the MFN tariff is sufficiently high that benefits from the liberalization receives a higher price (Baldwin and Venables 1995). for its exports, and so the exporter’s border price rises. The The net welfare impact on RoW is unambiguously neg- point is made concretely in figure 3.4. ative (Haberler’s spillover). RoW experiences no change on Starting from the situation in which the TBT is applied the import side but twice loses area e (shown in the left- to both Partner and RoW imports (i.e., when the equilib- most panel in figure 3.2)—once on its exports of Good 1 to rium price is P'), the preferential liberalization shifts Home, and once on its exports of Good 2 to Partner. The Home’s import supply curve (MS) to MSPTA, from MSMFN. Haberler spillover is an externality as far as the global trade The internal Home price falls to P", and this also system is concerned, and, as such, it plays a central role in becomes the price in Partner. The price that RoW the big-think regionalism literature. exporters receive (net of the cost of the TBT) falls to P" – T. Partner exports expand and those of RoW contract, but the former contract less than the latter expand, so A Note on Nontariff Barriers overall, Home’s imports rise. The welfare implications The traditional view of regionalism focuses almost are shown in figure 3.5. exclusively on preferential tariffs, but many PTAs also As with preferential tariff cutting, the liberalization cre- include liberalization of nontariff barriers. Here, we turn ates a positive trade volume effect for Home equal to area to the basic economics of technical barriers to trade A. Since, however, the price that Home actually pays for its (TBT) liberalization, or, more generally, frictional bar- imports from Partner and RoW is Home’s internal price rier liberalization. “Frictional,� in this sense, means that (rather than the external price, as in the case of tariff pro- the barriers impose a wedge between domestic and bor- tection), the liberalization leads to an unambiguous terms- der prices, but the wedge is “burnt up� rather than col- of-trade gain. Home gets its imports—both from RoW and lected as a rent by the government (as with a tariff) or by a from Partner—for less, that is, for P" instead of P'. This quota holder (as with a quota). gain corresponds to area F in figure 3.5. Thus, although the Economics 73 Figure 3.4. Effects of Preferential Frictional Barrier Liberalization on Prices and Imports RoW Partner Home border price border price domestic price MSMFN MSPTA XSR XSP P’ P’’ T P’ – T P’’ – T T P* MD RoW Partner Home XR’’ XR’ exports XP’ XP’’ exports M’ M’’ exports Source: Author’s elaboration. Figure 3.5. Welfare Effects of Preferential TBT Liberalization: Viner’s Ambiguity Vanishes euros euros XS A P’ F P’’ D P’ – T P’’ – T MD XP ’ XP’’ exports XR’’ XP’ M’ M’’ imports Source: Author’s elaboration. Note: TBT, technical barriers to trade. discriminatory application of the TBT leads to supply Thus, in the market under study, the PTA partners switching from RoW to Partner, this “trade diversion� has unambiguously gain. Notice that Viner’s ambiguity has no welfare consequences for Home. disappeared. RoW unambiguously loses, for the usual As usual, Partner unambiguously gains from the pref- reasons; that is, the heightened competition in Home’s erences because it enjoys a positive trade volume effect market induces RoW firms to sell less and to charge a (its exports expand) and a positive terms-of-trade effect. lower net price. 74 Richard Baldwin Dynamic Gains of Regionalism however, liberalization can unleash a virtuous circle of more competition, lower prices, increased sales, and higher One of the principal justifications for PTAs, especially employment. those involving developing countries, is the belief that uniting small economies will make regional firms more efficient and more competitive by allowing them access to a The Break-Even–Competition Curve bigger market. The boost in efficiency would allow the Diagram in a Closed Economy firms to lower prices, raise quality, and increase their com- To study the logic of the effect of integration on scale and petitiveness, both internally and externally. competition, we need a simple yet flexible framework Plainly, this logic cannot work in the supply-and- that allows for imperfect competition. The break-even demand-diagram world considered in the preceding sec- (BE)–competition curve (COMP) diagram employed here tion, since that worldview assumes that a firm’s efficiency assumes a knowledge of simple imperfect-competition has nothing to do with market size. This section offers a models. (Annex B provides an introduction for readers window into the dynamic gains of regional integration and who are not familiar with such models.) explains the logic of how regional integration could lead to To study the effects of integration on firm size and effi- a situation of fewer, larger firms operating on a more effi- ciency, number of firms, prices, output, and the like, it is cient scale and facing more effective competition. useful to consider a set of diagrams in which all these things are determined. To keep things simple, we begin with the case of a closed economy. Liberalization, Defragmentation, and The key relationship is shown in figure 3.6, where the Industrial Restructuring number of firms and the profit-maximizing price-cost National markets are separated by a very long list of barri- margin are determined. As usual, the equilibrium will be ers. These include tariffs and quotas, but also multiple the intersection of two curves, the BE curve and the COMP technical, physical, and fiscal impediments that make it curve. To explain the economics of the diagram, we start easier for companies to sell in their local market than in with the COMP curve. other markets. The typical outcome of these explicit and hidden barriers is that national firms can often be success- Figure 3.6. Competition (COMP) and Break-Even (BE) Curves ful in their home markets while being marginal players abroad. This state of affairs, known as market fragmenta- markup (μ) tion, reduces competition, which, in turn, raises prices and keeps too many firms in business. Keeping firms in busi- μmono ness is not, of course, a bad thing in itself. The problem is that it results in an industrial structure marked by too many inefficient small firms that can get away with charg- ing high prices to cover the cost of their inefficiency. Tearing down regional barriers defragments the mar- μduo kets and generates extra competition. This procompetitive BE (break-even) curve effect puts pressure on profits, and the market’s response is “merger mania�: the least efficient firms are squeezed out by the competition, prompting an industrial restructuring whereby weaker firms enter into mergers or are bought up. μ’ Schematically, the successive steps can be summarized as liberalization → defragmentation → procompetitive effect COMP curve → industrial restructuring. The result is fewer, bigger, more efficient firms that face more effective competition from each other. All this means improved material well-being for all, as prices fall and output rises. In some industries, restructuring may be accompanied by a sizable reallocation n=1 n=2 number n’ of employment: firms cut back on redundant workers and of firms close inefficient plants and offices (a painful process for Source: Author’s elaboration. workers who have to change jobs). In other industries, Note: The superscript mono refers to monopoly; duo, to duopoly. Economics 75 The competition (COMP) curve. It is easy to under- even when the markup is high. That is, taking the markup stand that imperfectly competitive firms charge a price as given, the BE curve shows the number of firms that can that exceeds their marginal cost; they do so in order to earn enough to cover their fixed cost—say, the cost of set- maximize profit. But how wide is the gap between price ting up a factory. and marginal cost, and how does it vary with the number Equilibrium prices, output, and firm size. It is important of competitors? These questions are answered by the to note that firms are not always on the BE curve, since they COMP curve. can earn above-normal or below-normal profits for a If there is only one firm, the price-cost gap—the markup while. In the long run, however, firms can enter or exit the of price over marginal cost—will equal the markup that a market, and so the number of firms rises or falls until the monopolist would charge. If more firms are competing in typical firm earns just enough to cover its fixed cost. By the market, competition will force each firm to charge a contrast, firms are always on the COMP curve, since firms lower markup. This “competition-side� relationship between can change prices quickly in response to any change in the the markup and the number of firms is shown in figure 3.6 number of firms. as the COMP curve. It is downward sloping because com- With this in mind, we are ready to work out the equilib- petition drives the markup down as the number of com- rium markup, number of firms, price, and firm size in a petitors rises, as explained above. We denote the markup by closed economy, using figure 3.7. The right-hand panel the Greek letter μ, (“mu,� an abbreviation for markup). The combines the BE curve with the COMP curve. The inter- size of the markup is an indicator of how competitive the section of the two defines the equilibrium markup and the market is. long-run number of firms. More specifically, the COMP The break-even (BE) curve. The markup and the num- curve tells us that firms would charge a markup of μ' when ber of firms are related in another way, summarized by the there are n' firms in the market, and the BE curve tells us BE curve. that n' firms could break even when the markup is μ'. The When a sector exhibits increasing returns to scale, there equilibrium price is, by the definition of the markup, just is only room for a certain number of firms in a market of a the equilibrium markup plus the marginal cost, MC. Using given size. Intuitively, more firms will be able to survive if the MC curve from the left-hand panel, we see that the the price is far above marginal cost, that is, if the markup is equilibrium price is p' (μ' plus MC). The middle panel high. The curve that captures this relationship is called the shows the demand curve, and this allows us to see that the zero-profit curve, or the break-even (BE) curve (figure total level of consumption implied by the equilibrium 3.6). It has a positive slope because more firms can break price is C'. Figure 3.7. Prices, Output, and Equilibrium Firm Size in a Closed Economy Home market euros price markup (P + MC) demand curve BE E’ E’ p’ p’ μ' E’ AC COMP MC n’ n� number of firms sales C’ total x’ per firm sales Source: Author’s elaboration. 76 Richard Baldwin The left-hand panel helps us find the equilibrium firm words, the new BE curve must pass through point 1 in the size, that is, sales per firm, which we denote x. This panel figure. At that point, the markup is μ', the number of firms shows the average and marginal cost curves of a typical is 2n', and logic tells us that this combination of μ and n firm. A typical firm’s total profit is zero when price equals would result in all firms breaking even. Point 1, however, is average cost. (When price equals average cost, total revenue merely an intellectual landmark used to determine how far equals total cost.) Since we know that total profits are zero out the BE curve shifts. It is not where the economy would at the equilibrium and that the price is p', the equilibrium be right after liberalization, since the markup would imme- firm size must be is x', which is where the firm’s size diately be pushed down to μA. implies an average cost equal to p'. Because the increase in competition would immediately In summary, this three-panel figure lets us determine the push down the markup to μA, the two newly integrated equilibrium number of firms, markup, price, total con- markets will initially be at a point below the BE curve. We sumption, and firm size, all in one diagram. With this in know that all firms will be losing money at point A because hand, we are now ready to study how regional integration— the actual markup (μA) is less than what would be needed or indeed, any source of increase in market size—could lead to have all 2n' firms break even. This loss of profit is not a to industrial restructuring. problem in the short run because firms only need to break even in the long run. Indeed, the profit losses are what would trigger the process of industrial restructuring that The Impact of Regional Liberalization eventually reduces the number of firms. Regional integration removes trade barriers gradually, but The corresponding effect on prices is shown in the mid- the basic economic effects can be better illustrated by con- dle diagram, as the move from E' to A and then to E ". sidering a much more drastic liberalization—taking a Before we explain this, observe that the middle panel shows completely closed economy and making it a completely the demand curve for Home only, so the no-trade-to-free- open economy. To keep things simple, we suppose that trade liberalization does not shift the demand curve. The there are only two nations, Home and Partner, and that Foreign market has an identical demand, but since exactly they are identical. Since they are identical, we could trace the same thing goes on in Foreign, we omit the Foreign the effects by looking at either market, but for convenience, demand curve to reduce the diagram’s complexity. we focus on Home’s market. As mentioned above, the initial impact of the extra The immediate impact of the no-trade-to-free-trade competition (2n' firms, rather than n', selling to the Home liberalization is to provide each firm with a second market market) pushes the equilibrium markup down to μA, so the of the same size and to double the number of competitors price falls to pA. Thus, during the industrial restructuring in each market. How does this change the outcome? phase, the price would rise to p" (from pA), but this rise The competition aspect of the liberalization can be does not take the price all the way back to its preliberaliza- simply traced. The increased number of competitors in tion level of p'. each market makes competition tougher. In reaction, the The impact of this combination of extra competition typical firm will lower its markup in each market to point and industrial restructuring on a typical firm is shown in A in figure 3.8. the left-hand panel. As prices fall, firms that remain in the The doubling of the market size facing each firm also market increase their efficiency—that is, lower their aver- has an important effect. The liberalization adds a new mar- age cost—by spreading their fixed cost over a larger num- ket for each firm, so it makes sense that more firms will be ber of sales. Indeed, since price equaled average cost able to survive. To see how many more firms can survive, before the liberalization and equals average cost in the we work out the impact of the liberalization on the BE long run after liberalization, we know that the price drop curve and find that the curve is shifted to the right, specifi- is exactly equal to the efficiency gain. In the left-hand cally, to BEFT, as shown in the rightmost panel of figure 3.8. panel, this is shown as a move from E ' to E ". Increasing The shift means that at any given markup, more firms can returns to scale are at the root of this efficiency gain. As break even. This is true because, as market size increases, the equilibrium scale of a typical firm rises from x ' to x ", sales per firm increase, providing a higher operating profit average costs fall. per firm at any given level of markup. To summarize, the no-trade-to-free-trade liberaliza- The size of the rightward shift is determined without tion results in fewer, larger firms. The resulting scale difficulty. If there were no change in the markup, a dou- economies lower average cost and thus make these firms bling of the number of firms could break even, since each more efficient. The extra competition ensures that these firm would be selling the same number of units. In other savings are passed on to consumers, so prices are lower. Economics 77 Figure 3.8. Prices, Output, and Equilibrium Firm Size with Integration Home market only euros price markup demand curve BE BE FT E’ E’ E’ 1 p’ p’ μ' E� p� E� E� p� A μA A pA AC COMP MC n’ n� 2n’ number of firms x� sales C’ C� total x’ per firm sales Source: Author’s elaboration. It is useful to think of the integration as taking place in two E ". During this process, firms enlarge their market steps, as follows. shares, the markup rises somewhat, and profitability is restored. 1. Short term: Defragmentation and the procompetitive effect (from E' to A). We start with the short-term The welfare effects of this liberalization are straightfor- impact, that is to say, the impact before the number of ward. The four-sided area C marked out by p', p", E', and firms can adjust. Before the liberalization, each market E" in the middle panel of figure 3.9 corresponds to the was extremely fragmented in the sense that firms in gain in the Home consumer surplus. As usual, this gain each country had a local market share of 1/n' and a zero can be broken down into the gain to consumers of paying share in the other market. After the liberalization, the a lower price for the units they bought prior to the liberal- market share of each firm is the same in each market, ization and the gains from buying more (C " versus C '). that is, 1/2n'. This elimination of market fragmentation Exactly the same gain occurs in the Foreign market (not has a procompetitive effect, which is defined as a shown in the figure). decrease in the price-cost markup, shown in the right- As it turns out, the four-sided region labeled C in fig- hand panel of figure 3.8 as a move from E' to A. The ure 3.9 is Home’s long-term welfare gain because there is short-term impact on prices and sales can be seen in the no offsetting loss to producers and there was no tariff rev- middle panel as a drop from p' to pA. enue to begin with. Firms made zero profits before liberal- 2. Long term: Industrial restructuring and scale effects (A to ization, and they earn zero profits after liberalization. Note, E"). Point A is not a long-term equilibrium because the however, that this long-term calculation ignores the operating profit earned by a typical firm is insufficient medium-term adjustment costs. to cover the fixed cost. We see this by noting that point A The discussion above has shown that integration ini- is below the BE curve, telling us that the markup is too tially leads to big price reductions and large profit losses. low to allow 2n' firms to break even. To restore a normal These profit losses are eliminated as the number of firms level of profitability, the overall number of firms has to falls and as profits are restored to normal levels. During fall from 2n' to n". In some cases, this process occurs via this industrial restructuring process, prices rise slightly. This mergers and buyouts; in others, the number of firms is sequence of steps, sometimes called industrial consolidation reduced by bankruptcies. As this industrial consolida- or an industry shakeout, is relevant to some industries. In tion occurs, the economy moves from point A to point other industries, firms anticipate the increased competition 78 Richard Baldwin Figure 3.9. Welfare Effects of Complete Liberalization Home market only euros price markup demand curve BE BEFT E’ E’ E’ p’ p’ μ’ C E� p� E� E� μ� p� AC COMP MC number n’ n� of firms x’ x� sales C’ C� total per firm sales Source: Author’s elaboration. and undertake mergers and acquisitions quickly enough to activity, whereas dispersion forces discourage such concen- avoid big losses. In figure 3.8, this would look like a move tration. The spatial distribution of economic activity at any from E' directly to E". moment in time depends on the balance of the proconcen- tration (agglomeration) forces and the anticoncentration (dispersion) forces. Agglomeration and the The main question in this section is, how does trade New Economic Geography integration affect the equilibrium location of industry? To Industrialization and deindustrialization are core concerns set the stage for the equilibrium analysis, we first consider of developing-country policy makers around the world. dispersion and agglomeration forces in isolation. For better Or, to put it differently, policy makers care about the loca- understanding of how trade arrangements affect profitabil- tion of industry. Although a whole host of policies affects a ity, and thus industrialization, it is convenient to employ a nation’s industrialization, trade policy has proved to be a simple analytical framework, one constructed by Puga and critical element in industrialization in almost all countries. Venables (1998). It focuses on four forces: two dispersion This section considers an analytical framework that per- forces (factor-market competition and local-market com- mits us to think logically about several of the key forces petition), and two agglomeration forces (input-cost link- affecting industrialization and how they interact with trade ages and demand linkages). barriers. The framework is often called the new economic geography, following the terminology of 2008 Nobel laureate Dispersion and Agglomeration Forces Paul Krugman. The basic focus of the new economic geography is on Dispersion forces favor the geographic spreading out of whether firms would enter or exit a particular market. economic activity. Land prices are the classic example. The The key determinant, in this simplified view of the price of land, and therefore the price of housing, office world, is the firms’ profitability. If setting up a new firm space, and so on, is usually higher in built-up areas such as in a particular country would be profitable, then the firm central London than in rural areas such as northern Wales. is created. If production in a particular country becomes If everything else were equal, firms and workers would pre- unprofitable, the firm ceases production. The entry or fer to locate in less built-up areas. (Of course, we know that exit decision rests on the balancing of two sets of forces: other things are not equal.) The forces that make built-up agglomeration forces and dispersion forces. Agglomeration areas more attractive are called agglomeration forces; we set forces promote the spatial concentration of economic them aside for the moment to concentrate on dispersion Economics 79 forces. Dispersion forces counteract agglomeration forces require many intermediate inputs—parts and compo- by increasing the attractiveness of less developed regions. nents. When these parts and components are produced In addition to land prices, there are several other forms of locally, they tend to be cheaper and can be supplied in a congestion-based dispersion forces, including wages, timelier manner. Demand linkages reflect the attractive- which tend to be higher in built-up areas. ness of a country that has easy access to customers, In this section, we focus on only two dispersion forces. whether local or in a trading partner. Factor-market competition captures the way that industrial- Demand-linked circular causality rests on market-size ization tends to push up wages. (That is, the competition is issues—hence its name. Firms want to locate where they for labor and operates between industry and other sectors have good access to a large market such as Japan or the within a single country.) Local-market competition reflects United States. If a firm locates in the big market, it incurs the fact that the presence of many industrial firms in a par- shipping costs to sell to other markets, but its costs of sell- ticular country tends to increase the degree of competition ing to big-market customers are low. (It is cheaper to sell to for customers in the local market. An important point is nearby customers.) Since there are more customers in the that local-market competition depends on trade barriers. If big market, firms can reduce their shipping by moving a country’s markets are perfectly open to international there. This is where the circular causality of demand link- competition, the competition for local customers is a ages starts. Other things being equal, firms want to be in global competition. At the other extreme, if the country is the big market. completely closed off, local competition depends solely The causality becomes circular because the movement on the number of local firms. As trade barriers fall, the of firms from the small market to the big market makes the nature of competition shifts gradually from fully local to big market bigger and the small market smaller. The reason fully global. is that firms buy inputs from other firms. Thus, firms mov- Always assuming that other things are equal, factor- ing to the big market create more demand in the big mar- market competition means that countries with little indus- ket and less in the small one. We call this an agglomeration try tend to have low wages and are thus more attractive to force, since spatial concentration of economic activity cre- industry. Local-market competition means that countries ates forces that encourage further spatial concentration. with little industry are markets in which the extent of com- The basic idea is illustrated in figure 3.10. It is useful to petition is rather low, and they too are attractive to indus- separate two things that are closely related: market size (big try. Both types of competition are dispersion forces that market as a share of total market, or the spatial distribution tend to make firms want to avoid spatial concentration. of demand), and firm location (share of firms in the big Agglomeration forces exist when the spatial concentra- market, or the spatial distribution of firms). tion of economic activity creates forces that encourage fur- Starting from the left-hand arrow, we see that market ther spatial concentration. This definition is more circular size affects the location of firms. The logic rests on firms’ than the straight-line chain of causes and effects usually desire to minimize shipping costs. The right-hand arrow presented in economics, but its circularity is the heart of shows that the location of firms affects relative market size. the subject. The logic is simply that firms tend to buy inputs locally. There are many agglomeration forces, but some of them operate only on a very local scale. These explain, for instance, why banks tend to group together in one part of a Figure 3.10. Demand-Linked Circular Causality city such as Paris or London and theatres cluster in another part of the city. The study of agglomeration at this level—it Share of firms in the big is called urban economics—is fascinating, but it is not the market level of agglomeration that interests us. Trade arrange- When firms To reduce move to big ments affect agglomeration at the level of countries, and at market, local shipping PRODUCTION EXPENDITURE that national level, city-level agglomeration forces are costs, firms SHIFTING SHIFTING purchases makes the big unimportant. The third and fourth forces we consider in prefer to locate in market bigger this section are agglomeration forces that tend to encour- and the small big market. age the clustering of industry in a single country. market smaller. Input-cost linkages capture the fact that a well-devel- Size of big oped local network of suppliers can boost a country’s market as share of world attractiveness as a site for industrial production. This idea turns on the real-world fact that manufacturing firms Source: Author’s elaboration. 80 Richard Baldwin If no dispersion forces were in operation, this circular presence of many suppliers in the big market and how the causality would continue until the small market was movement of firms to the big market widens the range of entirely empty of industrial firms. supplies and thus makes the big market even more attrac- This brings us to the second major type of agglomera- tive from a cost-of-production point of view. This sort of tion force. Input-cost-linked circular causality works in a logic is sometimes called “cluster economics.� fashion that is similar to demand-linked circular causality, but it involves production costs rather than market size. Locational Effects of Liberalization In the modern economy, firms buy plenty of things from other firms. These range from raw materials and Trade integration affects the balance of agglomeration and machinery to specialized services such as marketing, dispersion forces in complex ways. Such complexity is accounting, and information technology (IT) services. important for understanding the real world, since trade Since it is cheaper to find and buy such inputs from firms liberalization has produced very different results in differ- that are nearby, the presence of many firms in a location ent developing countries. The best way to understand this tends to reduce the cost of doing business in that location. complex logic is to follow the principle of progressive com- Thinking this through, we can see that a similar circular plexity. We start with a set of simplifying assumptions that causality will encourage agglomeration (see figure 3.11). allow us to focus on the critical logical relationships. Once The figure separates two things that are closely related we have understood this logic in a setting reduced to essen- but are worth keeping distinct: firm location (share of tials, we add back complicating factors. firms in the big market, or the spatial distribution of A simple diagram such as figure 3.12 helps with the firms), and the cost advantage of producing in the big mar- study of the balance of the agglomeration and dispersion ket (cost of producing in the big market, or the spatial dis- forces. In the figure, the strength of agglomeration and dis- tribution of production costs). persion forces is plotted on the vertical axis, and the hori- Starting from the left-hand arrow, we note that if many zontal axis shows the share of all firms that are located in firms are already in the big market, then doing business in the big region—here, in the North. the big market will, all else being equal, be cheaper than doing business in the small market. This production-cost • The agglomeration force line rises because of circular differential influences the location of firms. The right- causality—that is, agglomeration forces become stronger hand arrow shows how the relocation of firms tends to as industry agglomerates. improve the business climate in the big market and • The dispersion force line rises because the benefit of worsen it in the small market, at least in terms of the range staying in the small region increases as more firms move of available inputs. Again, if there were no dispersion to the northern market. For example, as more industry forces, this circular causality would empty out the small moves to the North, the wage gap widens, and so does market entirely. In other words, cost-linked circular causal- the difference in local competition. ity describes the way in which firms are attracted by the The locational equilibrium in the left-hand panel is shown by point E; this is where the share of firms in the Figure 3.11. Input-Cost-Linked Circular Causality North rises to the point at which incentives to agglomerate are just balanced by incentives to disperse. It is instructive Share of firms to consider why other points are not the equilibrium. For in the big market example, consider the point at which half the firms are in When firms the North. For this equal distribution of firms, the strength To take move to the big market, of the agglomeration force is shown by point a, and the advantage of PRODUCTION COST the wider the range of strength of the dispersion force is shown by point b. SHIFTING SHIFTING range of intermediate goods expands Because a is greater than b, we know that the agglomera- intermediate goods, firms while tion force—the force that tends to make firms want to move to the contracting move to the North—is stronger than the dispersion force, big market. in the small market. which tends to make firms want to move to the South. As a Cost of consequence, the situation in which only half the firms are producing in in the North cannot be the equilibrium; with agglomera- the big market tion forces stronger than the dispersion forces, some firms Source: Author’s elaboration. will move from South to North. As firms move northward, Economics 81 Figure 3.12. Locational Equilibrium Diagram Some industry in All industry in both nations one nation strength of the strength of the agglomeration and agglomeration and dispersion forces dispersion forces dispersion force D E A A D E a agglomeration force b B S share of share of 1/2 1 1/2 1 firms in firms in North North Source: Author’s elaboration. the gap between the agglomeration force and the disper- costless trade, industry would spread out much more than sion force narrows. The location equilibrium is where the is the case today. two forces just offset each other, that is, point E. This global statement—that lower trade costs foster the The example in the left-hand panel shows an equilib- industrialization of poor countries—must be tempered, rium with some industry in both nations. The right-hand since partial removal of trade costs (e.g., the lowering of panel shows the situation in which agglomeration forces tariffs without removing natural trade barriers) can shift are so strong that all industry ends up in the North (full the locational equilibrium in either direction or may have agglomeration). This is not a bad approximation of the sit- no effect at all. What matters is whether the agglomeration uation that confronts many developing countries today forces fall sufficiently relative to the dispersion forces. (that is, they have essentially no competitive industry) and The basic idea is shown in figure 3.13. A trade liberaliza- that faced almost all developing countries before the emer- tion will shift the agglomeration forces line down. If the lib- gence of the newly industrializing countries in the 1980s. eralization has a big impact on the agglomeration forces, Finally we come to the main subject of this section: how the line could shift to A2; if the effect is more modest, the does tighter economic integration affect the location of shift could be only to A1. The liberalization also erodes the industry across countries? Here, we view trade integration dispersion forces, and this is shown as the shift to D1. (We as simply reducing trade barriers such as tariffs and other could also consider large and small effects on the D curve, restrictions. How do we show the trade-cost reduction in but that would clutter the diagram without adding insight.) the locational equilibrium diagram? The key is that, in the case of A1 and D1, the liberaliza- As the discussion above suggests, lower trade costs tion has had no effect on industry location; industry is all between the two nations in our simple framework will in the North, before and after. In the A2 and D1 case, how- weaken both the agglomeration forces and the dispersion ever, the liberalization has sparked some industrialization forces. After all, at the extreme of costless trade, there is in the South; as the new equilibrium, E', implies a share of no advantage to being in any particular market, with the industry in the North (S') that is less than one. sole exception of that conferred by the factor-market competition dispersion force. The other three forces rely Trade Arrangements on differences that are created by costly trade. Factor- market competition, however, has nothing to do with We now employ the tools presented above to organize our trade costs; all that matters is how much industry is in each thinking about how preferential trade arrangements could nation. The conclusion, then, is that if the world went to affect the location of industrial production. Here we follow 82 Richard Baldwin Figure 3.13. Locational Equilibrium Diagram with Trade Figure 3.14. Trade Arrangements and Industrialization Liberalization share of share of All industry in one nation industry in South industry in South strength of the agglomeration and dispersion forces S 1 = S2 E S1 A A1 D D1 S1 A2 S2 E’ freer freer B trade trade Source: Puga and Venables 1998. Note: The notation S1 = S2 denotes that the two lines are identical and that both countries have an equal share of world industry. S’ share of 1/2 1 firms in North Source: Author’s elaboration. lower wages outweighs the demand and cost advantages in the North, and some firms move south. the thinking of Puga and Venables (1998) who assume four Importantly, cluster economics continues to operate in nations: two northern nations with free trade in manufac- the South, so industry initially only starts operating in one tured goods between them, and two southern nations of the southern countries. In this exercise, the two southern whose exports face equal tariffs in all markets. In this countries are identical, so the choice of which starts first is model, initially, all industry is in the two northern nations. entirely a matter of chance; to be concrete, this nation is (Four nations are required if the intrinsic economic mass labeled southern nation 1, and its industry share is marked of the integrating area is to equal the intrinsic mass of the S1. The reason for this lumpy industrialization is that the third nations when all nations are symmetric.) first firms to set up create cost and demand linkages to The Puga and Venables framework is too complex to other firms in the same country. They also raise wages, solve with paper and pencil, and so the effects are simu- but the linkage effects are stronger, so what we see is a sec- lated with the use of a computer. The outcome is plotted in ond industrial agglomeration forming in just one of the a diagram (figure 3.14) in which the share of world indus- developing countries. This implication of cluster eco- try is on the vertical axis and the level of tariffs is on the nomics is both quite in line with real-world experience and horizontal axis. Because the focus is on the industry in the somewhat unexpected for economists familiar with more South, the paths of the industry share of the two southern neoclassical, marginalist reasoning, where things tend to nations are plotted as S1 and S2. change smoothly. Notice that when the industrialization process spreads, the emergence of the second southern industry Global Multilateral Tariff Liberalization is initially harmful to the first because of the extra com- To set the stage, consider the impact of a global reduction petition. (Think of the effect of the emergence of China of tariffs (remember that North-North tariffs are already on the exports of the Philippines and Thailand.) This is zero before starting). As global tariffs fall, the tendency another implication of cluster economics that is not at for industry to disperse rises. At some point, it becomes first obvious. profitable for some firms to relocate to the South. The local-market competition dispersion force is weakened, Unilateral Liberalization the factor-market dispersion force is unaffected, and both agglomeration forces are weakened as the advantages of Next, consider unilateral liberalization in this simple Puga- being near customers (demand linkages) weaken and the Venables setup. We examine three types: only one southern advantages of being near suppliers (cost linkages) also nation liberalizes; both southern nations liberalize; and weaken. At some point, the attractiveness of the South’s both northern nations liberalize. Economics 83 Liberalization in a single southern economy. Here, only sufficiently high, industrialization will start in one south- southern nation 1 engages in unilateral import tariff liber- ern nation. As before, the logic of cluster economics tells us alization, with all other barriers held constant. The solid that the process begins in only one southern nation but line marked S1 in the right-hand panel of figure 3.14 then shifts to the other. shows the outcome, with the unilateral opening leading to industrialization. This is not always what happens, but it is Preferential Liberalization certainly one possibility, and the one that actually occurs under the specific Puga-Venables setup. If the two southern nations sign a PTA and lower tariffs The point is that although extra import competition between themselves, something like multilateral liberaliza- from the North has a negative effect on industrial location tion occurs. As long as the two markets are not too small, in the South (the product-market local competition effect), the liberalization will cause industry to become established the cheaper supply of imported intermediate goods in the South, but the mechanism is completely different: becomes the dominant force at some point, and industry the driving force here is the effective market enlargement becomes established in southern nation 1. This result is not caused by reducing intra-South barriers. This is, of course, general; for example, if southern nation 1 were very small the classic argument made in the 1960s and 1970s for and faced high export barriers, unilateral liberalization South-South PTAs. would not induce industrialization. But where this is not As in the multilateral case, the spread of industry to the case, the combination of low wages and low-cost developing countries is uneven, initially taking place in one intermediates (resulting from import liberalization) country and only spreading to the second when trade bar- eventually leads to industrialization, as is shown in the riers are lower. Indeed, this sort of uneven development right-hand panel. did occur in some early South-South PTAs. For example, in Unilateral liberalization by northern economies: The gen- the East African Community, industry started to grow in eralized system of preferences (GSP). Next, consider the case Kenya at the expense of Uganda and Tanzania. The key dif- in which the North makes a gesture to the South by unilat- ference is that the countries do not benefit from better erally removing tariffs on imports from the South. The access to northern markets or to North-produced interme- outcome will be similar to that shown in the right-hand diate inputs. panel. The northern tariff cutting improves the prospects The impact of a North-South PTA is particularly inter- of locating industry in the South because it erodes the esting. Here, the southern nation obtains better access to demand-linked causality that favors the North to begin the big northern market and benefits from lower-cost with. However, since the South is not lowering its tariffs and inputs, but in each case only with respect to the partner. much industry will remain in the North, the input-cost The liberalizing southern economy suffers from more linkage continues to strongly favor a northern location. competition from northern firms, but because its wages are Moreover, as mentioned above, if the southern market is lower, the balance of better reciprocal market access is in small enough, or the input-cost linkages are strong enough favor of the South. This spread of industry is associated (or both), the North’s unilateral tariff cutting may have no with a large decline in the North’s share of industry. The effect on industrialization. In terms of figure 3.13, we could loser is the other southern economy, which does not attract have a situation like A1 and D1 where, despite the shift in any industry and now has to contend with industrial clus- attractiveness toward the small region, the balance of forces ters in both North and South. It is not difficult to see how a still favors full agglomeration in the North. Given how little single North-South PTA such as the U.S.–Mexico agree- most GSP programs have done to promote southern ment proposed in 1991 or the one between Japan and industry, this is not a case worth keeping in mind. Malaysia could trigger a spate of requests from other Liberalization by both southern economies. We now look southern nations. at MFN liberalization by southern economies in tandem. All southern import tariffs are reduced in the same way, so Liberalization of Parts and Components only northern tariffs against southern exports remain. Ini- But not Final Goods tially, the outcome is like that of unilateral liberalization by a single southern nation. The lower southern tariffs A very common liberalization strategy among developing heighten the anti-industrialization product-market com- countries is to reduce tariffs unilaterally on inputs but petition from the North, but it encourages southern pro- not, or to a lesser degree, on final goods. This evokes the duction by lowering the cost of inputs. If the southern old measures of the effective rate of protection, whereby markets are not too small and the importance of inputs is the actual protection provided by a nominal tariff of, say, 84 Richard Baldwin 10 percent on automobiles can be vastly larger than 10 per- final good. Of course, this might not work if the southern cent if the tariff on imported intermediates is zero. A nation is too small. numerical example will illustrate. Suppose a country can The members of the Association of Southeast Asian buy autos at US$10,000 and adds a 10 percent tariff, so that Nations (ASEAN) are working on a slight variant of this the auto sells for US$11,000 on the internal market. Fur- uneven liberalization strategy. They are lowering regionally thermore, suppose the country charges no tariff on the the within-ASEAN tariffs—the common effective prefer- parts needed to assemble an auto and can buy these parts ence tariffs (CEPTs)—on final goods, as well as on parts for US$8,000. This implies that the cost of assembly, when and components, but preferences only emerge on final it is done most efficiently, is worth US$2,000. But now we goods as they are lowering their MFN applied tariffs on see that assembly of autos inside the country will be prof- parts and components in line with their preferential cuts. itable as long as it costs less than US$3,000. Thus, in some This has made sales of ASEAN-assembled autos very sense, the effective rate of protection on the assembly activ- attractive within the ASEAN group. ity is 50 percent, not 10 percent as the tariff on autos sug- The point can be seen in table 3.1. Apart from Malaysia gests. This is an old story, and countries around the world and Thailand, which have tried to build up more complete still keep up the fiction of having an automobile industry auto industries, the main ASEAN countries have lowered by maintaining a high tariff on autos and a low tariff on their MFN tariffs on engines to zero. More important, all of completely knocked down (CKD) autos—basically, kits them also have lowered their preferential tariffs on engines that are opened and assembled like IKEA furniture. to zero. Tariffs on autos are still very high (30 to 80 percent) Recently, however, the situation has become far more on an MFN basis but are much lower (5 percent) on a pref- subtle. Many developing countries, especially in East Asia, erential basis. Although these tariff cuts are impelled by the are industrializing on the basis of parts and components logic of supply-chain manufacturing in East Asia (much of manufacturing rather than final goods manufacturing. In this driven by Japanese firms), the cuts have also fostered essence, they industrialize by becoming part of the global such trade. From the perspective of the Puga-Venables supply chain. (Actually few of these supply chains are analysis, this combination of regional preferences is a way global; apart from some electronics, they are regional, in of favoring agglomeration of parts and final goods indus- order to reduce transport cost and delays.) try at the regional level. For a wide range of countries, the import and export of parts and components are much more important than the Regional Interindustry Trade export of final goods (Fukunari and Ando 2005). For the Philippines, for instance, 60 percent of the country’s International trade occurs when a product is made in one machinery exports consists of parts, as does 45 percent of country and purchased in another. In traditional trade the- its imports in this category. Plainly, the Philippines’ indus- ory, a country’s production is thought of as stemming from try is in the business of importing parts, adding some its advantages, or, to be more precise, comparative advan- value, and then exporting the parts. tages. Countries with abundant farmland would produce In terms of the Puga-Venables analysis, this uneven lib- more cereal than they could consume and would thus be eralization of parts and final goods fosters southern exporters of cereals, and the same logic would apply to industry in that it reduces the cost of inputs without the export patterns of countries that are especially well increasing the competition from northern industry in the endowed with unskilled labor, high technology, or oilfields. Table 3.1. ASEAN Tariffs on Engines and Automobiles, Most Favored Nation (MFN) Tariffs and Common Effective Preference Tariffs (CEPTs), 2008 (percent) Malaysia Indonesia Philippines Thailand Item MFN CEPT MFN CEPT MFN CEPT MFN CEPT Small-auto engines 25 0 0 0 0 0 10 0 Medium-size-auto engines 30 0 0 0 0 0 10 0 Automobiles 30 5 55 5 30 5 80 5 Source: ASEAN Secretariat Web site, http://www.aseansec.org/12025.htm. Note: ASEAN, Association of Southeast Asian Nations. Small-auto and medium-size-auto engines correspond to ASEAN tariff lines 8407.31.00.00 and 8407.32.00.10. Automobiles correspond to 8703.21.29.10. Economics 85 Much of the world’s trade, however, involves two-way was strengthened as the latter—Hong Kong SAR, China; trade in similar goods—usually, manufactured goods— the Republic of Korea; Singapore; and Taiwan, China,— and often between countries that have similar comparative experienced their own “hollowing out� and followed the advantages. To account for this, Paul Krugman and a num- lead of Japanese manufacturing companies by offshoring ber of other theorists in the late 1970s and early 1980s the most labor-intensive production stages to East Asian developed the so-called new trade theory. This took a quite countries that had a comparative advantage in such tasks different view of the determinants of a country’s exports. (that is, countries whose low wages more than compensated The point of departure is firm-level scale economies for their low labor productivity). that explain why the production of each good is spatially Information technology advances and the falling costs concentrated, not just in one country, but even in one fac- of transportation, especially air freight, facilitated and tory. If the production of most goods is spatially concen- accelerated the development of Factory Asia by making trated but the customers for that good are spread across complex production structures easier and cheaper to man- many countries, goods are made in one country but are age, while at the same time making them more flexible and bought in another, generating international trade. More- more reliable. China’s opening up was another huge spur, over, scale economies tell us that no country can efficiently bringing as it did something like a half-billion low-wage, produce all goods, so each country specializes in a subset of low-productivity workers to the gates of Factory Asia. This the world’s goods. Trade in similar products arises very nat- accelerated the erosion of the headquarter nations’ com- urally between similar countries. France and Germany, for parative advantage in labor-intensive production processes example, both produce autos, and they both consume them. while simultaneously increasing the attractiveness of the Since the production of each model of auto is spatially con- offshoring solution. centrated but the customers are spatially dispersed, the Figure 3.15 shows the number of plants from Japanese result is two-way trade in autos between countries that have electrical machinery and automobile industries that set up quite similar comparative advantages. in the listed East Asian economies. (Plainly, the number of Until recently, the focus of this intraindustry trade was plants is only a rough indicator of the actual degree of off- between developed countries, to a large extent because they shoring.) The offshoring process started gradually but were the only ones that had significant and competitive picked up speed in the late 1980s. Between 1975 and 1990, manufacturing sectors. Since the mid-1980s, however, the total number rose three and a half times, with almost the so-called second unbundling phase of globalization half of this increase coming between 1985 and 1990. The has meant a rapid rise of industry in developing coun- figure also shows that China was not a major player in the tries. A great deal of this new industry does not involve competition for plant locations before 1990. The plants full production of goods; rather, it focuses on parts and components for export. The general trend is discussed in Brulhart (2009). Figure 3.15. Location of Japanese Auto and Electrical A good way to introduce the topic of this trade in parts Machinery Plants in East Asia, 1975–2004 and components is to tell the tale of how it started in East Asia. The phenomenal growth of Japanese incomes and 900 wages in the 1980s and 1990s eroded the country’s compara- 800 tive advantage in manufacturing. Japanese businesses reacted 700 by seeking lower-cost manufacturing sites for labor-intensive 600 stages of production. The obvious solution was to offshore number these stages to nearby East Asian economies. 500 This tendency started the development of what is 400 called “Factory Asia.� Instead of Japanese goods being 300 made in Japan and sold in the United States or Europe, a 200 new pattern of “triangle trade� emerged. Firms that were 100 headquartered in Japan would produce certain high- technology parts in Japan, ship them to factories in 0 1975 1980 1985 1990 1995 2000 2004 ASEAN nations for labor-intensive stages of production China Thailand Malaysia (including assembly), and then ship the final products to Indonesia Vietnam Western markets or back to Japan. This division of East Asia into headquarters economies and factory economies Source: Baldwin 2008. 86 Richard Baldwin Table 3.2. Intraindustry Trade as a Share of Internal and undermining the WTO. The new perspective, the so-called External Trade of PTAs twenty-first-century regionalism, is really about defining (percent) the new international disciplines necessary to underpin Internal External twenty-first century international commerce—that is, EU-15 46.6 24.8 international production networks and the flows of people, CEMAC 1.2 0.1 things, and information that they generate (Baldwin WAEMU 0.9 0.4 2010a). Here, twenty-first century regionalism is defined as EAC 0.3 0.4 consisting of three parts: (a) the system of deep North- SACU 0.3 9.0 South agreements signed by the United States, the EU, and Source: Brulhart 2009. Japan; (b) the system of bilateral investment treaties and Note: CEMAC, Economic and Monetary Community of Central Africa other facilitating agreements such as double-taxation (Communauté Économique et Monétaire de l’Afrique Centrale); EAC, East African Community; EU-15, European Union before the 2004 enlargement treaties; and (c) the massive unilateral policy reform (15 members); SACU, Southern African Customs Union; WAEMU, West undertaken by emerging nations (of which unilateral tariff African Economic and Monetary Union. cutting is the most easily measured part). This chapter has focused on the economics of the tradi- generated new intraindustry trade, almost all of it in tional view of PTAs: that they are mostly about preferential machinery and much of it in parts and components. tariffs. As tariffs are just a form of tax, all of the traditional- Intraindustry trade, however, is not universally impor- view economics can be thought of as applied tax econom- tant in PTAs. The two strongest contrasts are the EU and ics. A great part of the deeper disciplines involved in the African PTAs. As table 3.2 shows, the degree of two-way twenty-first-century regionalism, however, turns on regu- trade in similar products inside the EU is very high, almost lation economics, and the regionalism literature has not 50 percent, but the share is much lower for the EU’s exter- focused much on this type of economics when it comes to nal trade. Much of this intraindustry trade involves parts PTAs. There is still much to be done. and components, and such trade tends to be regionalized in order to reduce the costs and difficulties of managing complex manufacturing processes at great distances. Annex A. Policy Pitfalls of Classic but Incomplete Diagrammatic Analyses Among the African PTAs, two show higher intraindustry trade measures within the regional arrangement, and two Until the 1990s, the main points in the economic literature exhibit the opposite. In all cases, the figures are quite low, were presented using diagrammatic analysis. Two diagrams reflecting the marginal role of manufacturing in these were particularly pivotal. The fact that they ignored some countries’ exports. of the three elemental effects discussed in this chapter (Smith’s certitude, Haberler’s spillover, and Viner’s ambi- guity) distorted the direction of the literature and, with it, Conclusions academic trade economists’ perceptions of PTAs. Since For most of the history of the GATT/WTO, regionalism these older, incomplete diagrams occasionally enter today’s was all about tariff preferences. The literature has therefore regionalism debate, it is worth presenting them briefly and focused primarily on the economics of preferential tariff highlighting their shortcomings. The first is the Johnson liberalization. The first key concern—the issue facing diagram that is still used in most undergraduate textbooks; nations in the early postwar period in Europe and Latin the second is the Shibata “small PTA� diagram. America—was whether an individual nation would gain from joining a PTA. The answer is unclear. Discriminatory The Johnson (1960) Diagram and the liberalization is a synonym for preferential liberalization; Johnson-Cooper-Massell Proposition this is both “liberalization�— which removes some price wedges and thus tends to improve economic efficiency— Although economists have been well aware of the three and “discrimination,� which introduces new price wedges effects since Meade (1955), this analysis was not integrated and thus tends to harm efficiency and welfare. From the into mainstream trade theory, in part because it was mar- 1990s on, concern shifted to the more systemic question of ginal, whereas trade economists were interested in studying whether regionalism was good or bad for the multilateral the discrete liberalization implied by PTAs. Viner (1950) trading system but the focus on tariffs remained. provided no diagrams. Accordingly, customs union theory, More recently, fundamental changes in international as it was known at the time, was a distinctly wordy subject commerce have led to a dramatic reduction in tariff until Johnson (1960b) introduced his famous diagram preferences, but no reduction in worries that PTAs are that illustrated Viner’s ambiguity in a manner that was Economics 87 immediately transparent to all economists (figure 3A.1). identical to MFN free trade. (Both before and after, all For policy analysis, Johnson’s diagram is fatally flawed as it imports would come from A.) Home’s domestic price falls assumes that FTAs and the like only have welfare effects on from PA + T to PA, and the net welfare gain is B + G + the import side—export effects are assumed away. D + H, in the right-hand panel.10 In the diagram, Home imports can come from partner Readers will immediately note that Smith’s certitude country A or B. Home’s demand is an infinitely small and Haberler’s spillover are missing. Third nations are share of world demand, so it faces perfectly elastic export entirely unaffected by the trade policy of an infinitely small supply curves from both sources (labeled XSA and XSB). nation such as Home. In Johnson’s diagram, the partner We start with Home’s imposing an MFN specific tariff of nations care no more about Home’s trade policy than a T, so that all imports come from the low-cost supplier, perfectly competitive firm does about gaining or losing one country A. The domestic price is PA + T, and the border atomistic buyer. This omission was an attractive feature price is PA. when the pivotal question was whether a single nation Home can form a customs union with country A or B, would gain from joining a customs union, but it renders so we consider both. The customs union with B would the diagram useless for consideration of systemic issues. remove the tariff only on imports from B (the high-cost The diagram assumes that Home’s decision to form a PTA supplier), leading to supply switching. Home switches has no systemic effects at all. Also missing from the dia- from importing everything from A to importing every- gram is an analysis of the preferential access that Home’s thing from B. Home’s domestic price falls from PA + T to exports win in its partner’s market. PB. Assuming a utilitarian metric, the net welfare effects are For two decades, the Johnson diagram dominated eco- (B + D) minus E, which may be negative or positive nomic analyses of PTAs to such an extent that Smith’s cer- depending on elasticities and the height of the initial tariff; titude and Haberler’s spillover came to be largely forgotten this is Viner’s ambiguity.9 by academic trade economists. This went so far that many The customs union with country B was termed “purely mainstream trade theorists came to view PTAs as econom- trade diverting,� yet if the initial tariff was high and the PB – ically irrational—a view encapsulated in the Johnson- PA border-price gap was small, it can be welfare improving Cooper-Massell proposition stating that a small nation for Home. This result—a welfare-improving but purely should always prefer unilateral MFN liberalization to any trade-diverting customs union—seemed to contradict PTA (Cooper and Massell 1965). The point is easily illus- Viner’s reasoning, and it produced the first of what was to trated in figure 3A.1: cutting T to zero on imports from A be a long series of ivory-tower debates over terminology; and B will always yield net welfare gains that are at least as this one pitted Meade (1955) against Johnson (1960b) and high as those from any customs union. Corden (1965). From the modern perspective, Johnson’s analysis seems If Home chooses to form a customs union with A, the impossibly simplistic, and the disconnect between aca- ambiguity disappears. Such a customs union is unam- demic and real-world thinking is truly astounding. For biguously welfare improving, since its positive effects are instance, when the United Kingdom submitted its first Figure 3A.1. Johnson’s Diagram, Small Home and Partner Countries euros euros D S PB + T XSB + T PA + T XSA + T B+D A B C D PB XSB E F G E H PA XSA MD M M’ imports quantity Source: For right-hand panel, Johnson (1960b); for left-hand panel, author’s elaboration. Note: The right-hand panel is the standard open-economy supply-and-demand diagram in price-quantity space for an infinitely small nation (Home). The left-hand panel transcribes the analysis into a more compact diagram in price-import space. 88 Richard Baldwin application for EU membership in 1961, better market The diagram presumes that the two PTA partners, access for U.K. exporters was the key concern, but aca- Home and Partner, import the same good from the RoW. demic economists were working with the Johnson diagram Home and Partner are “small� with respect to RoW and that assumed this away. Moreover, the main preferential so face a perfectly elastic RoW export supply curve, trade agreement in existence at the time, the European XSRoW. This sets the initial border price to PR in both Economic Community (EEC), accounted for a substantial nations. Home has a higher MFN tariff than Partner to fraction of world imports, and the key nations—France, start with (TH, as opposed to TP), so the pre-PTA price is Germany, and the United Kingdom—were far from higher in Home. atomistic. As Pomfret (1997) points out, a number of When Home and Partner form their PTA, Partner- frameworks were developed at the time that would have based firms initially see a higher price in Home and so allowed the necessary extension, including Johnson begin exporting to Home. In equilibrium, all post-PTA (1957, 1958), Humphrey and Ferguson (1960), and Black- Home imports, M'H, are supplied by Partner firms. Part- hurst (1972), but the Johnson diagram’s hold on the liter- ner’s internal price remains at PR + TP, so its consumption ature was so firm that the early efforts were obliged to and production are unchanged, which means that the new stick with his small-country fiction.11 exports to Home are replaced, one for one, by new Partner imports from RoW. In the case illustrated in figure 3A.2, Partner is large enough relative to Home to ensure that The “Small PTA� Diagram Home’s entire demand can be satisfied by Partner’s pro- An important analytical extension of the Johnson diagram ducers at PR + TP. In terms of welfare, the PTA results in a came with the “small PTA� analysis (Shibata 1967). It positive trade volume effect for Home but a negative bor- allowed for Smith’s certitude, although it still assumed der price effect. (Home pays PR + TP for its imports instead away Haberler’s spillover. The diagram continues to be of PR.) Partner expands its imports across the tariff wedge, used even today (e.g., Grossman and Helpman 1995), so it and this results in a positive trade volume effect equal to TR is worth presenting briefly. times the expanded imports (M'H). The small PTA diagram shows somewhat different Although it seems an odd objective from today’s per- assumptions concerning the pattern of comparative advan- spective, where rules of origin (ROOs) are a major barrier tage and the size of the two partners. The various combina- to trade, Shibata’s goal was to illustrate the irrelevance of tions of assumptions yield a range of results that have been rules of origin. His point was that ROOs only prevent bla- covered by three decades of literature. (See Panagariya tant trade deflection. Because goods from Partner and 1999 for a comprehensive survey of literature using the RoW are fungible, the equilibrium is the same with and small PTA diagram in recent decades.) Here we study a without rules of origin, as long as Partner’s supply is suffi- fairly standard case and illustrate the diagram’s properties cient. If Partner’s supply were not large enough to supply by demonstrating two classic results in the regionalism lit- all of Home’s imports at PR + TP, the PTA with ROOs erature (figure 3A.2). would have somewhat higher prices than one without. Figure 3A.2. The Small PTA Diagram: A Simple Case PARTNER HOME euros euros trade trade volume price D S gain gain PR + TH trade P R + TP price loss PR XSRoW D S quantity quantity M’H M’H Source: Author’s modification of a diagram in Pomfret (1997). Economics 89 The use of this exercise in the big-think regionalism unit brings in extra revenue (represented by areas D + E), literature comes in the form of “imported liberalization/ but it also depresses the price received for all units sold ini- protection� (Grossman and Helpman 1995). tially, lowering revenue by an amount equal to area A. The Another application that found popularity in the aca- net change in revenue, termed “marginal revenue,� is given demic literature but seems odd today is the proposition by areas D + E minus area A. The change in cost, termed that PTAs will always break down. Using a diagram similar “marginal cost,� is area E. Plainly, profit only increases if to figure 3A.2, Vousden (1990, 234) argues that Home the extra revenue D – A exceeds the extra cost E. As it is would be tempted to lower its MFN tariff to just under that drawn, D – A + E appears to be negative, so marginal rev- of Partner in order to recapture the tariff revenue and that enue is less than marginal cost at Q' + 1. This means that Partner would have an incentive to reply. The resulting raising output from Q' would lower profits, and so the ini- race-to-the-bottom tariff cutting was viewed as making tial guess of Q' turned out to be too high. PTAs “unsustainable.� Vousden (1990) did not attract To find the profit-maximizing level using this trial-and- much attention until Richardson (1995) extended and error method, we would consider a lower guess—say, Q' popularized his proposition. These two results (irrelevance minus four units—and repeat the procedure applied of rules of origin and unsustainability of PTAs) are classic above. At the profit-maximizing level, marginal revenue examples of how academic thinking on regionalism often just equals marginal cost. This level must be optimal, since followed literature-driven paths that had little relevance to any increase or decrease in sales will lower profit. Increas- real-world policy concerns. ing sales beyond this point will increase cost more than revenue, while decreasing sales would lower revenue more than cost. Both would reduce profit. Annex B. A Brief Review of Imperfect The right-hand panel of figure 3B.1 presents an easier Competition Models way of finding the point at which marginal revenue equals We start with the simplest problem: the decision faced by a marginal cost. The diagram includes a new curve, the mar- firm that has a monopoly. The monopoly case is easy ginal revenue curve, which shows how marginal revenue because it avoids strategic interactions. When a firm is the declines as the level of sales rises. (It declines because area only seller of a product, it can choose how much to sell and A from the left-hand panel becomes very small for low lev- what price to charge without considering the reaction of els of sales.) At the sales level marked Q*, marginal revenue other suppliers. The only restraint a monopolist faces is the just equals marginal cost. The firm charges the most it can, demand curve. A downward-sloping demand curve is a P*, at this level of sales. These are the profit-maximizing constraint because it forces the monopolist to confront a levels of sales and price. trade-off between price and sales; higher prices mean lower Several aspects of imperfect competition come through sales. When considering the impact of regional integration even in the monopoly case. First, in setting up the problem, on imperfectly competitive firms, we need to determine we had to make assumptions about the firm’s beliefs con- how various policy changes will alter prices and sales. The cerning the behavior of other economic agents. In this case, first step in this direction is to see what determines a the monopolist is assumed to believe that consumers are monopolist’s price and sales in a closed economy. The nat- price takers and that the trade-off between prices and sales ural question then is, what is the profit-maximizing level of depends only on the demand curve rather than, for exam- sales for the monopolist? ple, on the reaction of firms in other markets. Second, the An excellent way to proceed is to make a guess at the critical difference between perfect and imperfect competi- optimal level—say, Q' in the left-hand panel of figure 3B.1. tion comes out clearly. As part of the definition, perfectly Almost surely, this initial guess will be wrong, but what we competitive firms are assumed to take the price of their want to know is whether it is too low or too high. To this output as given. (A classic example is a wheat farmer who end, we calculate the profit earned when Q' units are sold at cannot set his own price but just sells at the current market the highest obtainable price, P'. The answer is A + B, since price.) This means that such firms are assumed to be igno- the total value of sales is price times quantity (area A + B + rant of the fact that selling more will depress the market C) minus cost (area C). price. In terms of the diagram, perfectly competitive firms Would profits rise or fall if the firm sold an extra unit? ignore area A, so they maximize profits by selling an Of course, to sell the extra unit, the firm will have to let its amount at which price equals marginal cost. Of course, any price fall a bit, say, to P". The change in profit equals the increase in sales would have some negative impact on change in revenue minus the change in cost. Consider first price, so it is best to think of perfect competition as a sim- the change in revenue. This has two parts. Selling the extra plifying assumption that is close to true when all firms have 90 Richard Baldwin Figure 3B.1. Monopoly Profit Maximization price price P’ P* A P� marginal demand B D revenue curve curve marginal cost curve C E Q’ Q’ + 1 sales Q* sales Source: Author’s elaboration. market shares which are close to zero. By stepping away that economists call duopoly. For simplicity, we assume from this simplification, imperfect competition allows that the firms have the same marginal cost curves. Taking firms to explicitly consider the price-depressing effect— firm 2’s sales as given at Q2, firm 1 has a monopoly on the area A—when deciding how much to sell. residual demand curve labeled RD1. Firm 1’s optimal out- The monopoly case is instructive but not very realistic; put in this case is x1', since at point A1 the residual marginal most firms face some competition. Taking account of this revenue curve, RMR1, crosses the marginal cost curve, MC. reality, however, brings us up against the strategic consider- The right-hand panel shows the same sort of analysis for ations discussed above. The convention we adopt to sort firm 2. Taking firm 1’s output as fixed at Q1, firm 2’s opti- out this interaction is the so-called Cournot-Nash equilib- mal output is x2'. rium that won John Nash a Nobel prize. That is, we assume Note that the situation in the figure is not an equilib- that each firm acts as though the other firms’ outputs were rium. To highlight the importance of the difference fixed. The equilibrium we are interested in is that in which between expected and actual outcomes, the diagram each firm’s expectations of the other firms’ outputs turn shows the solutions of the two firms when expectations out to be correct; that is, no one is fooled. This no-one- about the other firm’s output do not match the reality. The fooled notion proves to be somewhat difficult to compre- consistent-expectations outcome (the Nash equilibrium) is hend in the abstract, but, as we shall see below, it is easy in shown in the next figure, but we first consider why figure specific applications. 3B.2 is not an equilibrium. As drawn, this is not a Cournot-Nash equilibrium because the firms’ actual output levels do not match Residual Demand Curve Shortcut expectations: firm 1 produces x1', which is greater than Since firms take as given the sales of other firms, the only what firm 2 expected (that is, Q1), and similarly, firm 2 constraint facing a typical firm is the demand curve produces x2', which is greater than what firm 1 expected shifted to the left by the amount of sales of all other (Q2). We can also see the problem by observing that the firms. In other words, each firm believes it is a monopo- implied prices are not equal. If these quantities were actu- list on the shifted demand curve. (We called the shifted ally produced by the firms, then firms would not be able to demand curve the residual demand curve.) This realiza- charge the prices they expected to charge. In other words, tion is handy because it means that we can directly apply this is not an equilibrium because the outcome is not con- the solution technique from the monopolist’s problem; sistent with expectations. the only change is that we calculate the marginal revenue curve on the basis of the residual demand curve instead Finding the Expectations-Consistent Equilibrium of the demand curve. This trick is shown in figure 3B.2 for a competition The easiest way to find the expectation-consistent set of between two firms producing the same good—a situation outputs is to exploit the assumed symmetry of firms. In Economics 91 Figure 3B.2. Duopolist as Monopolist on Residual Demand: Example of a Nonequilibrium price price ’s firm 1 expectation of firm 2’s expectation of sales by firm 2, Q2 sales by firm 1, Q1 p 1’ demand demand curve (D) curve (D) p2’ residual demand curve, firm 1 (RD1) residual demand curve, firm 2 (RD2) MC MC A1 A2 x1’ firm 1 sales x2’ firm 2 sales residual marginal revenue residual marginal revenue curve, firm 1 (RMR1) curve, firm 2 (RMR2) Source: Author’s elaboration. Figure 3B.3. Duopoly and Oligopoly: Expectation-Consistent Outputs DUOPOLY OLIGOPOLY price typical firm’s expectation price typical firm’s expectation of the other firm’s sales of the other firm’s sales p* D D p** RD RD’ A MC A MC RMR RMR’ x* 2x* sales x** sales 3x** Source: Author’s elaboration. the symmetric equilibrium, each firm will sell the same to note are that (a) the optimal output for a typical firm is amount. With this fact in mind, a bit of thought reveals x*, given by the intersection of RMR and MC; (b) total that the residual demand curve facing each firm must be sales to the market are 2x*, and at this level of sales the half of the overall demand curve. This situation is shown in overall market price (given by the demand curve D) is con- the left-hand panel of figure 3B.3 for a duopoly. Some facts sistent with the price each firm expects to receive, given the 92 Richard Baldwin residual demand curve, RD; and (c) the outputs of the the market equilibrium. Modeling TBTs in such “sexy� industries will cer- tainly be the subject of much future work, but in this chapter we focus on identical firms are equal in equilibrium. mundane industries in which TBTs act by raising the costs of foreign Although allowing for two firms is more realistic than firms more than the costs of local firms. allowing for only one firm, studying the impact of integra- 9. The left-hand panel of figure 3A.1 translates the effects into tion on mergers and acquisitions requires us to allow for an Meade’s two-part framework: B + D is the trade volume effect (related to the change in the volume of imports), and E is the trade price effect arbitrary number of firms. In economists’ jargon, such a (related to the change in the border price). situation is called an oligopoly. As it turns out, this situa- 10. This contrast is the source of the rule of thumb that a PTA with tion is straightforward in dealing with the case in which your main trading partners is more likely to be welfare improving, since you are giving preferences to the partners that have demonstrated them- firms are symmetric. As more firms compete in the market selves to be the low-cost suppliers by winning the largest market share in (consider three instead of two), the residual demand curve an even competition with other suppliers. facing each shifts inward, so that the residual marginal rev- 11. The early 1980s saw a number of widely read studies that sought to reverse the Johnson-Cooper-Massell (JCM) proposition while retain- enue curve also shifts inward. The implications of this shift ing the small-country framework. These efforts (e.g., Wonnacott and for prices are clear. The new RMR = MC point occurs at a Wonnacott 1981; Berglas 1983) strike the modern reader as awkward lower level of per-firm output, implying a lower price. In because of the small-nation assumption and the intricate diagrammatic equilibrium (i.e., where outcomes match expectations), analysis. each of the three firms produces an identical amount and charges an identical price. What happens is that as the num- Bibliography ber of firms continues to rise, each increase in the number of competitors shifts the RD facing each firm. This will Balassa, Bela. 1967. “Trade Creation and Trade Diversion in the European Common Market.� Economic Journal 77: 1–21. inevitably lead to lower prices and lower output per firm. Baldwin, Richard. 1989. “The Growth Effects of 1992.� Economic Policy Of course, this analysis simply formalizes what most 9: 247–82. readers would expect. If one adds more competitors to a ———. 2008. “Managing the Noodle Bowl: The Fragility of East Asian Regionalism.� Singapore Economic Review 53 (3): 449–78. market, prices will fall, along with the market share of ———. 2010a. “21st Century Regionalism: Filling the Gap between 21st each firm. As is so often the case, the brilliant concepts Century Trade and 20th Century Trade Governance.� http://eiitf.iift are simple. .ac.in/Richard%20Baldwin%20-%2021st%20Century%20Regionalism .pdf. ———. 2010b. “Unilateral Tariff Liberalisation.� NBER Working Paper Notes 16600, National Bureau of Economic Research, Cambridge, MA. Baldwin, Richard, Simon Evenett, and Patrick Low. 2009. “Beyond Tariffs: 1. The basic problem was that the profession found the simple trade Multilateralizing Non-Tariff RTA Commitments.� In Multilateralizing creation–trade diversion paradigm to be effective in communicating the Regionalism: Challenges for the Global Trading System, ed. Richard crucial welfare-ambiguity result, but the words did not fully capture all Baldwin and Patrick Low. Geneva: World Trade Organization. the basic economic effects. Arvind Panagariya suggests that the terms per- http://us-cdn.creamermedia.co.za/assets/articles/attachments/19838 sist because they are “highly effective tools of focusing policy makers’ _multila_region_e.pdf. attention on the ambiguous welfare effects of [regional trade associa- Baldwin, Richard, and Anthony Venables. 1995. “Regional Economic tions]� (Panagariya 1999). Integration.� In Handbook of International Economics, vol. 3, ed. Gene 2. Of course, when considering the full economic impact, one must Grossman and Kenneth Rogoff, 1597–1644. Amsterdam: North consider scale economies, procompetitive effects, variety effects, location Holland. effects, and growth effects (Baldwin and Venables 1995). Most of these, Baldwin, Richard, and Charles Wyplosz. 2009. The Economics of European however, are not critical in the “big-think� regionalism literature. Integration. 3rd ed. New York: McGraw-Hill. 3. Readers may mentally insert a fourth untaxed good that enters the Berglas, Eitan. 1983. “The Case for Unilateral Tariff Reductions: utility function linearly, to formally eliminate Meade’s tertiary effects. Foreign Tariffs Rediscovered.� American Economic Review 73 (5): 4. The PTA diagram can be thought of as a modification of the Black- 1141–42. hurst (1972) diagram. Bhagwati, Jagdish. 1971. The World Trading System at Risk. Princeton, N.J.: 5. Given the symmetry, an FTA is automatically a customs union. Princeton University Press. 6. Area C2 might be called the “trade diversion� effect, and D2 + A Blackhurst, Richard. 1972. “General versus Preferential Tariff Reduction might be call the “trade creation� effect, but, as usual, the trade creation- for LDC Exports: An Analysis of the Welfare Effects.� Southern Eco- trade diversion dichotomy is incomplete. Here, it omits the third-nation nomic Journal 38: 350–62. terms-of-trade gain, B. Brulhart, Marius. 2009. “An Account of Global Intra-Industry Trade, 7. For the results, see World Bank, Trade, http://www.worldbank.org/ 1962–2006.� World Economy 32 (3): 401–59. trade/standards. Carpenter, Theresa, and Andrea Lendle. 2010. “How Preferential Is World 8. In certain industries, the impact of TBTs is radically more complex. Trade?� CTEI Working Paper. In industries with network externalities, such as mobile telephones, stan- Cooper, C. A., and B. F. Massell. 1965. “A New Look at Customs Union dards can be manipulated to throw up barriers against nonlocal firms. In Theory.� Economic Journal 75: 742–47. industries with patent races (e.g., pharmaceuticals) a regulation that Corden, W. Max. 1965. “Recent Developments in the Theory of Interna- merely delays the introduction of foreign goods can radically alter the tional Trade.� Special Papers in International Economics 7, Interna- market outcome in favor of home firms. In industries with learning tional Finance Section, Princeton University, Princeton, N.J. curves, product standards that apply to only a fraction of the market— Dixit, Avinash, and Victor Norman. 1980. Theory of International Trade. government or military purchases, for example—can have large effects on Cambridge, U.K.: Cambridge University Press. Economics 93 Fukunari, Kimura, and Mitsuyo Ando. 2005. “Two-Dimensional Frag- ———. 1999. “Preferential Trade Liberalization: The Traditional Theory mentation in East Asia: Conceptual Framework and Empirics.� Inter- and New Developments.� Journal of Economic Literature 39 (2): national Review of Economics and Finance 14 (3): 317–48. 287–331. Grossman, Gene M., and Elhanan Helpman. 1995. “The Politics of Free- Panagariya, Arvind, and Ronald Findlay. 1994. “A Political-Economy Trade Agreements.� American Economic Review 85: 667–90. Analysis of Free Trade Areas and Customs Unions.� Policy Research Haberler, Gottfried. 1936. The Theory of International Trade with Its Appli- Working Paper 1261, World Bank, Washington, DC. cations to Commercial Policy. New York: Macmillan. Pomfret, Richard. 1997. The Economics of Regional Trading Arrangements. Humphrey, D. C., and C. E. Ferguson. 1960. “The Domestic and World Oxford, U.K.: Oxford University Press. Benefits of a Customs Union.� Economia Internazionale 13: 197–213. Puga, Diego, and Anthony Venables. 1998. “Trading Arrangements and Johnson, Harry G. 1953. “Optimum Tariffs and Retaliation.� Review of Industrial Development.� World Bank Economic Review 12 (2, May): Economic Studies 21 (2): 142–53. 221–49. ———. 1957. “Discriminatory Tariff Reduction: A Marshallian Analysis.� Richardson, Martin. 1995. “Tariff Revenue Competition in a Free Trade Indian Journal of Economics 5: 39–47. Area.� European Economic Review 39: 1429–37. ———. 1958. “Marshallian Analysis of Discriminatory Tariff Reduction: Shibata, Hirofumi. 1967. “The Theory of Economic Unions: A Compara- An Extension.� Indian Journal of Economics 6: 177–82. tive Analysis of Customs Unions, Free Trade Areas and Tax Unions.� ———. 1960a. “The Cost of Protection and the Scientific Tariff.� Journal In Fiscal Harmonization in Common Markets, ed. Call S. Shoup, of Political Economy 68: 327–45. vol.1, 145–264. New York: Columbia University Press. ———. 1960b. “The Economic Theory of Customs Unions.� Pakistan Smith, Adam. 1776. The Wealth of Nations. Repr., Chicago, IL: University Economic Review 10: 14–30. of Chicago Press, 1977. ———. 1965. “An Economic Theory of Protectionism, Tariff Bargaining Taussig, Frank. 1892. “Reciprocity.� Quarterly Journal of Economics 7: and the Formation of Customs Unions.� Journal of Political Economy 26–39. 73: 256–83. Torrens, Robert. 1844. The Budget: On Commercial and Colonial Policy. Kemp, Murray C., and Henry Y. Wan, Jr. 1976. “An Elementary Proposi- London: Smith, Elder. tion Concerning the Formation of Customs Unions.� Journal of Inter- Viner, Jacob. 1950. The Customs Union Issue. New York: Carnegie Endow- national Economics 6 (1): 95–97. ment for International Peace. Meade, James E. 1955. The Theory of Customs Unions. Amsterdam. North- Vousden, Neil. 1990. The Economics of Trade Protection. Cambridge, U.K.: Holland. Cambridge University Press. Panagariya, Arvind. 1996. “APEC and the United States.� CIES Discussion Wonnacott, Paul, and Ronald J. Wonnacott. 1981. “Is Unilateral Tariff Paper 96/08, Center for International Economic Studies (CIES), Reduction Preferable to a Customs Union? The Curious Case of the University of Adelaide, Adelaide, Australia. Missing Foreign Tariffs.� American Economic Review 71: 704–14. 4 NORTH-SOUTH Preferential Trade Agreements Bernard Hoekman This chapter argues that if preferential trade agreements The next section provides a brief context for the trade- (PTAs) are to be development-friendly, they must focus on related reform agenda that confronts developing countries. complementing liberalization in trade goods with behind- Some rationales for and challenges facing North-South the-border regulatory reforms that are supported through PTAs are then discussed from a development perspective, development assistance instruments and that engage the and the evolving status quo is described. Suggestions for private sector. Such an extension of the PTA agenda to reg- further steps are presented, proceeding from the assump- ulatory issues can be beneficial to developing countries. Yet tion that the focus of North-South trade agreements is to deep and comprehensive PTAs between developed and promote development. Given that preferential market developing countries bring new risks. The proposed norms access liberalization is a second-best exercise from a global may limit policy freedom in inappropriate ways or may welfare perspective, proposals are developed to encourage result in the allocation of resources to areas that offer few the pursuit of nondiscriminatory liberalization and a immediate benefits or that are complex and costly to much more targeted focus on the key constraints that pre- implement. A redesign of the approach to PTAs—to liber- vent developing countries from benefiting more from trade alize and expand market access, to build in policy flexibility opportunities. Current approaches toward PTAs with and accountability, and to broaden technical assistance—is developing countries that are being pursued by the Euro- needed. pean Union (EU) are assessed in light of these proposals. A basic premise is that a shift in objective from market The intention is not to be comprehensive or to single out access or market integration to development means that the EU; a similar analysis could be applied to recent U.S. the modus operandi of negotiating and implementing PTAs. It is, however, true that the EU has been at the fore- trade agreements will have to change. One reason is that front in seeking to use PTAs as instruments for promoting many of the poorest countries may not benefit much from development. Finally, the conclusions from the discussion a traditional trade agreement; they already have good are summarized. access to major markets through nonreciprocal preference schemes, and they confront potential welfare losses if they Stylized Facts and Key Policy Challenges pursue preferential liberalization in favor of PTA partners only. Another reason is that the priority needs in many Until the global economic crisis struck in 2008, the world poor developing countries are not related to trade policy had witnessed a rapid expansion in developing-country but revolve around bolstering trade capacity, improving trade. The share of the developing economies in world the investment climate, and maintaining a competitive real trade increased from 20 percent in 2000 to 30 percent in exchange rate. PTAs can help address some of these priori- 2008 (World Bank 2010b), and all regions experienced ties, but only if they are appropriately designed and effec- greater integration with the rest of the world. There are, tively implemented. however, important differences among developing countries 95 96 Bernard Hoekman with respect to both the growth in trade and its pattern and objective at lower social cost, or complementary policies can structure. The poorest countries generally did less well be identified to address the by-product costs of openness. than middle-income economies in the 1990s, partly The persistence of trade policies can be explained by because of the dependence of their foreign exchange earn- political-economy reasons. Even though the aggregate ings on agricultural commodities. This helps explain why income and wealth of a nation may be expected to grow Sub-Saharan Africa’s share of nonoil world trade remains when trade distortions are reduced, not everyone will far below the level that prevailed in the 1970s, despite rela- gain. Owners of previously protected inefficient firms tively good growth and trade performance during most of will lose, as may their workers, especially if there are the first decade of the twenty-first century. doubts regarding the creation of new employment Research suggests that a lack of diversification is associ- opportunities. Social insurance and adjustment assis- ated with lower growth and greater output volatility. Africa tance mechanisms may not exist or may be weak. These is the least diversified region in the world today. Agricul- realities underline the importance of complementary tural trade remains highly important for Sub-Saharan reforms to increase the likelihood of realizing the bene- African countries, accounting for 25 percent or more of fits from trade reforms. total exports. One consequence is that countries are subject The list of beneficial concomitant reforms can be long to greater commodity price (terms-of-trade) volatility than and rather formidable. This, however, does not necessarily are more diversified economies, and the effect of this imply that there are difficult trade-offs to be made; most of volatility is more persistent (World Bank 2010a). Imbs and the policies are essentially additive to trade liberalization in Wacziarg (2003) note that countries at early stages of devel- the sense that they do not give rise to trade-offs (Winters opment experience a positive relationship between export 2004).2 Social costs may be lower if adjustment can be (output) diversification and growth, suggesting that from a spread over a period of time, as long there is confidence development perspective, the policy focus should be on that reform will actually occur. Without credible commit- support for greater diversification (Newfarmer, Shaw, and ment to a clearly defined and commonly known final goal, Walkenhorst 2009; Haddad, Lim, and Saborowski 2010). investments and adjustments may look undesirable, and Existing programs that center on promoting trade of efforts may be diverted to lobbying. Trade agreements can developing countries—especially nonreciprocal preferen- play an important role in this connection by laying out a tial access to Organisation for Economic Co-operation timetable for gradual liberalization that is credible because and Development (OECD) markets—have not been very it is enforceable by trading partners. effective.1 Major reasons for this inefficiency include supply- Failure to maintain a realistic real exchange rate has side weaknesses, civil conflicts, macroeconomic policies been one of the main causes of unsuccessful trade liberal- that resulted in overvalued currencies, governance prob- izations in developing countries (World Bank 2001). lems, corruption, and institutional weaknesses that inhibit Another cause is failure to address the fiscal consequences local entrepreneurs from taking advantage of market of tariff revenue losses. These losses are far from inevitable, opportunities. The question, then, is to identify what PTAs especially if nontariff barriers are converted into tariffs, could do to help achieve progress on such issues. exemptions are reduced, and tariff collections are improved. But they can pose a problem for poorer coun- tries in which trade taxes account for large proportions of Reform Priorities total revenue. Developing alternative sources of revenue Because average tariff barriers in developing countries may take time. Experience suggests, however, that moving remain higher than in industrial ones, much of the poten- toward a more uniform tariff structure and a concomitant tial welfare gain from trade reforms will arise from their elimination of exemptions may increase revenue collec- own liberalization. The reasons for imposing trade barriers tion, providing the space to develop alternative tax bases vary; they include infant-industry protection (import- before undertaking more far-reaching liberalization. substitution industrialization), balance of payments con- A major area in which administrative constraints bind is siderations (concerns that liberalization will increase institutional reforms. Given the importance of building up imports more than exports), and fiscal revenue objectives. the legitimacy and ownership of these reforms among the (Tariffs are easy to collect and can be a significant source of population, not only significant “technical� setup time but government income.) All these reasons are second-best in also a good deal of “political� time is required. Getting most circumstances, in that a lower-cost domestic policy institutions right the first time is very difficult, if not instrument can, in principle, be identified to satisfy the impossible, and continuing monitoring and adjustment North-South Preferential Trade Agreements 97 are needed. There are often advantages to proceeding on a is labor mobility. If labor markets are segmented or dis- broad front in order to maintain some semblance of fair- torted, benefits will be reduced. ness, and extensive institutional reforms are likely to • Establishment of new businesses. Cumbersome regula- require time and considerable administrative skill to carry tions for establishing new firms, constraints on access out. These observations apply, in particular, to regulatory inputs (such as utilities), and restrictions on physical policies and the agencies responsible for implementing expansion or labor recruitment and separation can cur- them, because regulation can give rise to significant operat- tail the willingness of entrepreneurs to start or expand ing costs and entry barriers for firms in a given market operations. once tariffs have been lowered. Tariff reforms can be exe- cuted at the stroke of a pen; regulatory reforms are much All these themes are highlighted in the research that more complex. focuses on the magnitude and determinants of adjustment costs and the factors that affect the size and distribution of the gains from trade reforms (Porto and Hoekman 2010). Designing Successful Reforms Thus, the benefits of trade liberalization depend, in part, on Governments have to build up support for their policies. support from other policies and institutions. Openness can Powerful interests will need to be assuaged—unless help induce improvements in these dimensions by making reforms are implemented in the context of major eco- them more “visible� and by creating incentives to fix the nomic crises. Compensating these interests through the problems, but additional investments and reforms will be careful design of complementary policies is not just a mat- required to address many of the constraints. For landlocked ter of sordid logrolling.3 Although any single efficiency- countries and poor, remote economies, reducing trade costs enhancing reform will hurt someone, if enough of them is often of critical importance. As discussed below, develop- are packaged together, negative effects will be netted out, ment assistance and mechanisms for monitoring impacts and many more people and interests will gain, on balance. can help ensure that reforms are effective and can anchor This is one of the main reasons for proceeding on a broad expectations (i.e., increase credibility). Trade agreements front. It is of great importance to ensure that potential ben- can provide focal points, but only if they address national eficiaries from trade-related reforms have the capacity to priorities—and are seen to do so—and are implemented. actually exploit new trade opportunities. This requires attention to the business environment and transaction Trade Liberalization Strategies: What Role Is costs, measures to enhance the productivity of firms and for Preferential Agreements? farms, improvement of connectivity to markets, steps to ensure access to finance, and the like. Specific areas for For governments that have decided to pursue opening of attention may include the following: the economy, a practical question is how to do so. Starting in the 1960s, a number of economies were highly successful • Infrastructure support. Farmers need to be able to reach in increasing incomes and reducing poverty. Notable major market centers at reasonable cost, firms need examples include Chile; Hong Kong SAR, China; the access to a reliable and efficient power supply, and so Republic of Korea; Singapore; and Taiwan, China. More on. In poor countries, transport (logistics) and transac- recently, China, Malaysia, Mauritius, and Turkey, among tion costs are often a multiple of any tariffs exporters others, have joined the list. All of them dramatically confront. This is one explanation for the more limited increased their ratio of trade to gross domestic product participation of poor countries in the process of inter- (GDP), but they pursued considerably different models of national specialization that was noted previously. trade policy reform and economic integration. The success • Credit markets. Access to finance is a critical input, both stories may be grouped into four broad categories: for new start-ups and for the expansion of existing plants. For example, achieving minimum consignment • Economywide trade liberalization. Some economies have size might entail hiring draft power or seasonal labor, pursued very liberal, most favored nation (MFN) trade but this is not possible without credit. Credit con- regimes, avoiding nontariff barriers and adopting either straints are a major reason for limited adjustment to free trade (as did Singapore and Hong Kong SAR, trade reforms. China) or low, nonnegotiable, uniform tariffs (e.g., Chile • Labor markets and mobility. The primary vehicle for and Estonia, before the latter’s accession to the EU). Since spreading widely the benefits of increased labor demand the creation of the World Trade Organization (WTO) in 98 Bernard Hoekman 1995, a number of countries have pursued MFN liberal- countries simply will not be that interested because the ization in the context of accession to that body, using developing countries’ markets are too small. Second, and a the WTO to anchor and precommit to reform; this was related factor in PTA negotiations, quid pro quo “pay- the case for China and Vietnam. ments� are likely to be requested in nontrade areas such as • Protection with offsetting policies for exporters. Other regulatory regimes, investment policy, and so on (Schiff economies reduced the incentives created by protection and Winters 2003; Limão 2007). That raises the issue of to produce for the domestic market, employing elabo- whether, in the negotiating process, governments lose rate systems that offset the bias against exports, includ- access to potentially useful instruments for promoting ing complex duty-drawback systems. Japan (in the early development. (See, for example, Fink, ch. 18 in this volume.) stages); Korea; and Taiwan, China, are examples. • Protection with export-processing zones. Along with pro- The Primacy of National Circumstances and Priorities tection, exporters located in specific zones are offered tariff-free access to intermediate inputs, with better A challenge for developing-country governments is not only infrastructure and fewer regulatory requirements. The to generate better access to partner markets but also to use limited geographic scope of the zones makes them eas- the PTA as a vehicle for promoting competition, reducing ier to manage, for countries with weak governance, than policy uncertainty, and improving the investment climate the “Asian� economywide model. Few countries have and business environment. Ensuring that deep integration succeeded in stimulating exports substantially through will benefit developing-country PTA members requires that this model. An exception is Mauritius, where the zones the specifics of regulation and cooperation reflect national generated about two-thirds of gross exports and circumstances. Regulatory standards and institutions need employed one-sixth of the workforce. Zones have also to be tailored to national circumstances to be effective and played a significant role in China. attain the desired objective. An increasing body of evidence • PTAs. An increasing number of countries are joining has shown that a “one-size-fits-all� approach—including PTAs to provide a focal point or blueprint for reforms, a international best-practice norms—may not be appropriate. mechanism for increasing market size and enhancing For example, Barth, Caprio, and Levine (2006), in a com- the contestability of markets, and a means of overcom- prehensive cross-country assessment of the impact of the ing political-economy resistance to reforms. Examples Basel Committee’s standards for bank regulation, find no are accession to the EU by many Central and Eastern evidence that any single set of best practices will necessarily European countries, Turkey’s entrance into a customs promote well-functioning banks. They argue that a high union with the EU, and Mexico’s membership in the degree of country specificity may be needed, rather than North American Free Trade Association (NAFTA). In all mere adoption of international norms “off the shelf.� these instances, PTA membership complemented WTO What may be most appropriate from an economic wel- accession or membership. fare (development) perspective is to create a framework for assisting governments in identifying good policies, rather The first three approaches are unilateral. Trade liberal- than a system that is premised on negotiated harmoniza- ization by other countries is clearly desirable, however; tion or convergence. Instead of being (too) prescriptive ex trade negotiations are the time-honored mechanism for ante, there is a case for maintaining flexibility conditional seeking such liberalization. In South-South agreements, on ex post monitoring of outcomes. An important corol- the focus is increasingly on expanding the size of the mar- lary of such an approach must be “restraint� on the part of ket by not only abolishing trade barriers but also easing large industrial partner countries in PTAs and accounta- internal constraints on intraregional trade and investment bility for performance and outcomes. Creating a focal through improvement of infrastructure trade facilitation point for constructive, as opposed to adversarial, interac- and transit and corridor management. Given the large tions between governments on the competitive, market- asymmetries in economic power (as indicated by market segmenting effects of regulation—or lack of regulation— size), the challenge for small and poor countries in North- and on the costs and benefits of specific reforms could do South agreements is to ensure that any negotiated outcome much to mobilize the needed support by constituencies in is in their interest. Such countries have very little scope to developing-country PTA members. This is especially so if use their trade policies as an instrument for inducing other the high costs of adjustment and of subsequent compliance countries to open up their markets. There are two implica- by developing-country members are recognized through tions. First, unilateral reforms cost developing countries increased technical assistance and investment to upgrade little in mercantilist terms—large (potential) PTA partner facilities (Hoekman and Winters 2009). North-South Preferential Trade Agreements 99 The Continuing Evolution of PTAs development perspective, the acid test is whether the pro- posed or negotiated rules in regulatory areas will improve The growth of PTAs has been significant. Recent PTAs tend the business environment and address supply-side priori- to be more open than earlier vintages, many of which were ties. Proponents of deep integration in North-South agree- designed to implement import-substitution strategies at the ments argue that binding disciplines in areas such as regional level. They also increasingly involve North-South competition and investment policy are critical for integrat- cooperation and extend to behind-the-border regulatory ing markets. This point is discussed further below. policies relating to investment, labor, environment, and All this information may help explain why recent PTAs competition. Examples involving the United States include involving countries that object to the inclusion of issues bilateral agreements with Australia, Chile, Central America, such as investment and competition in the WTO—not to Jordan, and Morocco. Investment and competition policies speak of more controversial subjects such as labor and are being discussed as part of the economic partnership environmental standards—may include disciplines in these agreement negotiations between the EU and the African, areas at the regional level. Presumably the net balance is Caribbean, and Pacific (ACP) countries and are on the positive, in part, because the smaller number of negotiating agenda of the EU’s association agreements with non-ACP parties makes it easier to exclude issues that are sensitive countries.4 and to identify quid pro quo deals. But deeper integration Often, the implementation of such policies may entail may not be beneficial to all signatories. From a develop- pecuniary spillovers to other countries, providing a ment perspective, the issue is not whether there are net rationale for cooperation on, for example, tax or other benefits but how to maximize the potential payoffs. There incentive programs to attract foreign direct investment. In may be cause for concern, in particular, about PTAs that many cases, however, the purported rationale is that the offer partial access to large markets for goods in exchange disciplines themselves will promote development. From a for acceptance of regulatory norms that may do little, if development perspective, the extension of PTAs to regula- anything, to increase the flow of investment to developing- tory issues can be beneficial if it improves policy credibil- country partners. The use, however, of PTAs as a frame- ity, thereby reducing risk premiums and helping attract work to reduce the frictional costs of trade by harmonizing investment. There is a prima facie case that regional coop- regulations and standards, increasing the credibility of eration on regulatory issues may be advantageous. Part- reform initiatives, or acting as vehicles for governments to ners, to begin with, may be rather similar and may have test the waters of freer trade may be very beneficial, even if common legal or administrative systems. North-South they are difficult and complex to realize. PTAs also tend to be associated with transfers of finance Summing up, there is a clear trend for PTAs to go and knowledge (technical assistance), potentially helping beyond trade in manufactures. Recent U.S. PTAs include reduce implementation and adjustment costs. In addition, agriculture, and movement on this front is discernable in high-income partners may provide offers of assistance in the EU; the Euro-Mediterranean negotiations include a the form of implicit “insurance,� as in the case of U.S. proposal to pursue reciprocal liberalization of trade in financial intervention to assist Mexico during the “tequila agricultural products. Services, investment, and regulatory crisis.� regimes are areas where, it can be argued, there is much to In part, the expansion of the negotiating agenda is be gained by developing countries from policy reforms and driven by a need to mobilize additional political support liberalization. There are potential concerns about these for abolishing the remaining trade-distorting policies in areas insofar as developing countries perceive proposed or areas such as agriculture. The Uruguay Round was actual disciplines not to be in their interests but still neces- premised on such a grand bargain, with developing coun- sary if these countries are to (continue to) gain preferential tries accepting new disciplines in a variety of areas, includ- access to northern markets. Clearly, much depends on the ing intellectual property rights (IPRs) and services, in coverage of the agreements and, in particular, on whether return for the elimination of the Multifibre Arrangement, the regulatory disciplines for behind-the-border policies the outlawing of voluntary export restraints, and inclusion are appropriate in the sense that the benefits outweigh the of agriculture in the WTO.5 A similar dynamic is driving costs of implementation. Much also depends on the extent PTAs today. The regulatory standards that are written into to which merchandise trade is liberalized, through, for trade agreements generally start from the status quo pre- example, access to agricultural markets and removal of the vailing in OECD countries, so that the lion’s share of asso- threat of contingent protection, and whether services liber- ciated implementation costs—but presumably also of the alization is covered, through all modes of supply, including benefits—lies with developing-country signatories. From a mode 4 (temporary movement of service providers). 100 Bernard Hoekman The tariff equivalents of the trade-restricting effects of The first two objectives are the bread and butter of PTAs. domestic policies are a large multiple of the prevailing bor- Their realization is constrained by political-economy der tariffs today. Further services liberalization would have forces. Small, poor countries have little to offer, in mercan- much greater positive effects on national welfare than would tilist terms, to induce large countries to remove policies that the removal of trade barriers—see, for example, Konan and harm them. Such market access, however, is important for Maskus (2006) on Tunisia, and Jensen, Rutherford, and Tarr mobilizing political support for domestic reforms in the (2007) on the Russian Federation. The “standard� increase in developing-country partner. A problem is that nonrecip- welfare from goods liberalization is 1 percent, but introduc- rocal preference programs may imply that exporters tion of greater competition on services markets raises the already have free access to the high-income market or gains to the 5–10 percent range or more. These large effects markets. The third objective, MFN-based liberalization, of services liberalization—for which there is increasing does not, of course, drive PTAs, which revolve around dis- econometric evidence (Francois and Hoekman 2009)— criminatory access.6 reflect both the importance of services in the economy and Realization of the fourth objective, adoption of comple- the extent to which many sectors continue to be protected. mentary measures, is the key challenge, if North-South There are, indeed, potentially large gains from reducing the PTAs are to be most relevant from a development perspec- prevalence and costs of differences in regulation, as well as tive. It may be impeded by the fact that the rules embedded the incidence of policies that simply prevent access to spe- (or proposed) in North-South PTAs tend to reflect the sta- cific markets. A key question, however, is whether progress tus quo in the high-income countries. From a development on liberalization can be, and is being, facilitated through perspective, the extension of PTAs to regulatory issues can PTAs, specifically. be beneficial if it improves policy quality or credibility, A recent assessment by Roy, Marchetti, and Lim (2006) thereby reducing risk premiums and helping to attract concludes that many of the trade agreements reported to investment. Regional cooperation may be more effective in the WTO since 2000 show a sectoral coverage that greatly this regard than multilateral cooperation because the part- exceeds the commitments the countries involved made in ners may be more similar; for example, they may have the General Agreement on Trade in Services (GATS). In common legal or administrative systems. As discussed areas where there are no WTO disciplines (e.g., on safe- below, North-South PTAs also tend to be associated with guards, subsidies, and procurement), there tend not to be more extensive transfers of finance and knowledge (techni- PTA rules either. The same is true for domestic regula- cal assistance), potentially helping to reduce implementa- tion; only one PTA (the Trans-Pacific Economic Partner- tion and adjustment costs. Proponents of deep integration ship Agreement, between Brunei Darussalam, Chile, in North-South agreements often argue that binding com- New Zealand, and Singapore) has established an across- mitments in areas such as competition law and regulation the-board necessity test—that is, an assessment of are critical for integrating markets and that it is easier to whether the trade-restricting effects of a policy are neces- envisage enforcement among the small number of partner sary to achieve the underlying regulatory objective (Fink countries. Although this can certainly be the case, the spe- and Molinuevo 2007). cific disciplines that are embedded in a PTA may not be a priority for development, implying that even if financial and technical assistance is made available, it could consti- Harnessing Regional Integration for tute diversion because the resources would have had a Development higher return elsewhere. It can be argued that, to be most beneficial to developing- None of the four objectives is straightforward to country signatories, PTAs should achieve, explaining why many PTAs have partial coverage in market access terms and include rules that may not • Remove foreign barriers to trade in products in which be first-best for developing-country members. If PTAs are developing countries have a comparative advantage. to do more to satisfy the four criteria suggested, the modus • Lower domestic barriers that raise the prices of goods operandi of designing and implementing them should give and services that firms and households consume. more weight to development considerations (economic • Promote more general MFN-based liberalization, which efficiency and equity). Concretely, four changes could best serves global development prospects. make North-South PTAs more development-friendly: • Support the adoption of complementary measures, reforms, and investments that allow the potential bene- • Unconditional acceptance by all parties to a PTA of fits of trade opportunities to be realized. MFN liberalization of trade in goods and services by North-South Preferential Trade Agreements 101 developing-country signatories, and preferential removal should be MFN reform as well—there is no reason to dif- of all barriers by OECD partner countries—bound, in ferentiate between developing countries. In effect, the both cases, by an enforceable treaty instrument MFN proposal implies emulation of the type of asymmet- • Construction of mechanisms to pursue priority ric liberalization that has been the norm in the General national regulatory policy objectives in developing- Agreement on Tariffs and Trade (GATT) and WTO, with country partners, as opposed to harmonization with the the difference being that in the North-South PTA con- standards of OECD countries, while maintaining the text, the northern countries “go all the way�—commit to role of PTAs as a commitment device free trade. Insofar as the North is not willing to do this for • Strengthened, grant-based financing mechanisms (aid large developing countries, the WTO can and should be for trade) to improve trade supply capacity and increase used as an instrument for reciprocal liberalization; the fact the benefits of trade reforms for poor households, using that the full preference rule might prevent PTAs from a local analysis of needs, with allocations determined by forming is, of course, not a problem from a global welfare the country’s overall development strategy perspective. • Active engagement by and with the private sector in the The “full Monty� rule for the North is consistent with surveillance and enforcement of the implementation of the thrust of current WTO rules for regional agreements, the various dimensions of the PTA, including the provi- except that it would go beyond the “substantially all trade� sion of financial and technical assistance. requirement to cover all trade. Although a formal rule change to this effect would be desirable, as noted by many—see, for example, Mavroidis (2005)— GATT Article Adjusting the Rules to Promote Market Access XXIV (allowing an exception to the MFN rule to permit PTAs are a steadily increasing source of discrimination in PTAs) and the Committee on Regional Trade Agreements trade today. Lowering the external levels of PTA protection to are basically defunct. Absent effective enforcement of the reduce the extent of discrimination against nonmembers— rules, seeking to change them is a largely irrelevant exer- which will often be developing countries—would promote cise. Thus, in practice, full liberalization is an action that the global public good. A solution offered by “realists� has needs to be taken unilaterally by northern countries. There been to point to MFN liberalization through WTO nego- is no need to change Article XXIV to permit the implemen- tiating rounds, according to the fact that efforts to regu- tation of this aspect of the proposal; what is needed is a late PTAs through the WTO rule-making and enforcement meaningful commitment to take development seriously.8 process have been totally ineffective (Mavroidis 2005). In That cannot be said for the suggestion that developing principle, however, if development were to be taken more country partners commit to MFN liberalization. Both seriously as a goal, changing the WTO rules on regionalism Article XXIV and the Enabling Clause—which allows could help make the PTA process more development- developing countries to liberalize less, especially where friendly. Concretely, high-income countries would be South-South agreements are concerned—would be required to liberalize in all sectors (not just “substantially implicated by an MFN rule. In terms of Article XXIV all�) on a preferential basis, with liberal and simple rules requirements, it may be easier to pursue a waiver for spe- of origin.7 Conversely, developing-country signatories cific PTAs, as MFN-based reform implies that all WTO would reduce their tariffs and apply negotiated trade policy members stand to gain from the PTA. This solution leaves commitments on an MFN basis. This MFN liberalization the inconsistency with the Enabling Clause to be resolved. would not imply a requirement to move to zero tariffs Arguably, the economics here are clear: pursuing a MFN across the board; instead, the goal would be a significant strategy has much less potential for welfare loss and can reduction in applied MFN tariffs by developing-country enhance benefits.9 partners, bound by the WTO. This method would pre- The argument in favor of a MFN approach by developing- vent trade diversion; reduce the administrative burden on country PTA members extends to services. Although actual customs authorities (as there would be no need to enforce additional liberalization in the services area has not been rules of origin on imports); help ensure that the PTA great—with the exception of the EU, most PTAs have not benefits all trading partners, not just members; and allow gone much beyond the GATS—services and investment governments more time to put in place alternative sources policies are very much on the PTA agenda. Multilateral of fiscal revenue. liberalization opens the market to the largest number of Large northern partners will not offer complete duty- competitors and gives consumers maximum choice. It also free and quota-free access to large developing countries leads to a less complex policy regime than a preferential without a quid pro quo. In principle, such a quid pro quo arrangement, implying lower administrative costs for the 102 Bernard Hoekman government and lower transaction costs for the private norms in OECD countries as examples of best practice, sector. cooperation would be geared toward assisting countries in If the market access rule proposals outlined for PTAs attaining their objectives efficiently. The specific content of are formalized in the WTO, then the next question is what regulation should reflect national (or local) circumstances. to do about existing PTAs. In practice, the revealed prefer- Thus, what may be most appropriate from an economic ence of members of these PTAs is clear: full liberalization is welfare (development) perspective is to create a framework often not the objective, and MFN is certainly not the goal. for helping governments identify good policies, not a sys- Seeking to change this status quo is unlikely to be fruitful. tem that aims at harmonization. Existing PTAs will therefore need to be grandfathered, in An important corollary of such an approach must be the unlikely event that it proves possible to change the accountability for performance and outcomes, generating WTO rules. That said, there is no reason existing agree- information (based on analysis) on whether the policies ments could not be reopened by a developing-country gov- employed are effective, what their costs and benefits are, ernment, assuming that development is indeed a major and so on. The fact that trade agreements are binding con- objective of the northern partner. tracts—that commitments are enforceable—gives them their value: traders have greater certainty regarding policy, and governments know what they are “buying� when they Policy Flexibility and Better Economic Governance make commitments. Any approach toward recognizing dif- We next turn to the second element of a strategy for ferential capacities and identifying regulatory options and increasing the development-friendliness of North-South priorities should minimize uncertainty for traders and PTAs. The market access dimension of PTAs arguably investors. Binding, enforceable disciplines on the use of should involve hard law—binding and enforceable com- trade policy are likely to be beneficial for development. The mitments. As mentioned previously, a major element of the case for trade policies designed to deal with specific gov- status quo is an increasing focus on harmonization and ernment and market failures that may prevent a supply hard law for behind-the-border policies concerning response to reforms from emerging is very weak (Pack and services, investment, and regulation. This may well be ben- Saggi 2006). It may not be obvious, however, what types of eficial to signatories, but much depends on country circum- domestic policies might be most appropriate and effective, stances. What are the preconditions for such commitments which suggests that experimentation and learning should to be beneficial? Have they been satisfied? Do the commit- be encouraged (see, for example, Rodríguez-Clare 2004; ments remove access to policy instruments that are desir- Rodrik 2004). able or are the only ones a government can feasibly employ The prevalence of complicated trade policies in many to address a market failure? Given that there is likely to be countries is often driven by industrial policy objectives, uncertainty with respect to these issues—and, often, differ- which have a long history. They span not only trade protec- ences in views between governments—an approach that tion but also subsidies and direct government involvement allows for greater policy flexibility could do much to in industry. Although opinions differ, the weight of the enhance the perceived benefits of engaging in PTA-based evidence suggests that such policies are generally very commitments. costly, often prolong the adjustment period, and distort A precondition for ownership of international agree- competition. That said, subsidies can facilitate learning, ments is that governments and stakeholders perceive the technology acquisition, and dynamic comparative advan- rules as benefiting the economy overall. A more economi- tage in situations where returns to such activities cannot be cally based mode of cooperation—as opposed to a focus on appropriated by private agents. Many commentators have harmonization-cum-approximation of laws—could help argued that policy interventions, including implicit or enhance such ownership. From an economic development explicit subsidies, lay behind the economic “miracles� in perspective, a mechanism for identifying good practices East Asia and were a major factor in the economic devel- makes sense, as these will often differ across countries. opment of European states, the United States, and Japan The focus would be on the provision of information and in the 19th and 20th centuries. Their case is that import learning through regular interactions of relevant policy protection and carefully targeted subsidies allowed govern- makers and constituents (stakeholders), peer review, and ments to stimulate key sectors that became efficient in their multilateral monitoring of the impacts of policies and their own right and provided positive spillovers for the economy effectiveness in attaining stated objectives (see Chayes and as a whole. Chayes 1995; Abbott and Snidal 2000; Helleiner 2000; In considering this infant-industry argument for gov- Sabel and Reddy 2007). Rather than seek to impose existing ernment support, it is important to differentiate between North-South Preferential Trade Agreements 103 sector-specific subsidies and policies aimed at facilitating the reality that PTAs are extending deeper behind the bor- learning and the development of private enterprise. The der. Moreover, with the exception of the WTO, trade policy case for general policy support for certain types of activity, is not a consistent focus of the activities of international including innovation, education, transport infrastructure, organizations. A major advantage of a PTA is that the focus and similar public goods, is uncontroversial. The same is is on trade and trade-related policies. Creation of a focal true for policies aimed at promoting socially beneficial point for a constructive, as opposed to adversarial, interac- activities. Markets can and do fail. There may be good tion between governments could do much to heighten rationales for governments to provide incentives for firms the domestic profile of the trade agenda for developing- and agents to undertake activities that would otherwise be country PTA signatories. It would also increase information undersupplied (Rodrik 2004). Specific interventions, how- on the effects of existing policy instruments—a necessary ever, will often get it wrong, in part as a result of rent seek- condition for adopting better policies—and ensure that ing and in part because of general equilibrium effects (a trade-related policy actions and investments are taken into subsidy for one activity implies a tax on all others).10 Mon- consideration when decision makers allocate resources to itoring and analysis of impacts and of the performance of public expenditures. supported sectors and activities are therefore important, as All these measures can be characterized as an effort to is the establishment of credible exit mechanisms; govern- improve economic governance in partner countries and, ments need to be able to withdraw support for experiments in the process, enhance the ownership of PTAs. Clearly, that fail. Trade agreements offer a potential vehicle for sup- the effectiveness of the implementation of the interactive porting such mechanisms. mechanisms will be critical for their credibility. In the end, PTAs can help by creating institutional mechanisms when governments (the partners jointly) deem that a that can assist in identifying policies that would be effective binding commitment in an area makes sense, entering and efficient in attaining specific goals set by governments into such a commitment will increase the probability of and by increasing the transparency of policies and their enforcement. Insofar as cooperation in specific areas con- outcomes through joint monitoring and analysis. A first tinues to be of the soft-law variety, legal enforcement step is to identify the relevant policies via the equivalent of mechanisms are not available. Here, accountability can what is done at the WTO through the Trade Policy Review only come from transparency, engagement, and publicly Mechanism and then to carry out an economic assessment disseminated analysis of actions and impacts. Oversight by of the rationale for and effectiveness of the relevant poli- parliaments and analytical assessments by institutions in cies. An example of an institution that does the latter is the the North—the U.S. Congressional Budget Office and Productivity Commission in Australia. Assessment of General Accountability Office; similar national institu- whether instruments are achieving development objectives tions in the EU such as the French research institutions and whether less trade-distorting ones can be identified Centre d’Etudes Prospectives et d’Informations Interna- requires judgments regarding appropriate sequencing and tionales and the Centraal Plan Bureau in the Netherlands; the need for complementary reforms and investment. and Australia’s Productivity Commission—could help These judgments must be made by the government con- inform development assistance programs. But strengthen- cerned but would benefit from inputs from other PTA ing the capacity to undertake such analysis in developing- members.11 country partners is crucial. Even if one is not convinced by the upside of pursuing Might such monitoring and interaction be better dele- greater flexibility on regulatory disciplines in PTAs, the gated to the WTO? Although non-PTA members may have downside risk is arguably limited. At worst, the cost is that little immediate interest in the policies pursued by a spe- PTA members conclude after a number of years that the cific developing-country PTA member, the substantive approach is not beneficial. Such an attitude may, however, coverage of PTAs will generally overlap to a great extent be too complacent. A case can certainly be made that if the with the issues that are addressed by WTO agreements or raison d’être of a trade agreement is the negotiation of that may be taken up in the future. There is much to be said binding commitments, policy dialogue discussions may do for considering an expansion of the WTO Trade Policy more harm than good by increasing uncertainty; duplicat- Review Mechanism to enable it to undertake much more ing the efforts of the World Bank, the International Mone- in-depth analyses of the impacts of policies pursued by tary Fund, and others; and incurring transaction costs. An PTA members and by customs unions (Hoekman 2005). alternative is to leave economic policy dialogue to interna- An additional reason for establishing such a mechanism tional development and financial institutions. There is at the WTO is that the consultations and impact assess- much to be said for this counterargument, but it ignores ments associated with any policy flexibility mechanism 104 Bernard Hoekman will entail resource costs. These may be significant for length by Prowse (2002 and 2006), trade-related funding poor countries with a scarcity of skilled personnel. If the should be allocated within the context of an overall coun- required work is undertaken multilaterally, much of it try development program and an agreed macroeconomic could be carried out in the context of mechanisms such policy framework. As a development tool, stand-alone spe- as the Enhanced Integrated Framework for Trade-Related cific funds and associated mechanisms are less likely to be Technical Assistance, reducing the costs. (See Prowse effective than integration of the prioritization and resource [2006] for a discussion.) allocation process into national poverty reduction and More regular interaction on trade policies would pro- development strategies. vide a framework for helping governments assess Trade policy often will not be the most important policy whether instruments are achieving stated objectives. The area from a growth perspective—and it should be borne in publication and dissemination in the countries concerned mind that many low-income countries have taken actions to of the results and findings of reports and discussions move away from nontariff barriers and to reduce dispersion would also increase the public profile of trade-related in tariffs. A fortiori, trade policies and institutions that are policies. More regular cooperative interaction by regula- covered by PTAs may not be among the areas within the trade tors and trade officials concerning trade policies and area where actions and investment are most needed. As constraints on market integration could also improve stressed previously, the primary determinants of success in communications between the development and trade harnessing trade openness to deliver growth are concomitant communities, as the analysis and discussions might assist policies and institutions in the developing countries them- in identifying where development assistance has the selves. Supply capacity is a necessary condition for exploiting greatest potential to help countries benefit more from market access opportunities, which will be determined by the trade agreements and cooperation. prevailing investment climate and the trade and business environment. Much of the associated policy agenda extends beyond trade policies and cannot be addressed through Aid for Trade: Capacity Building for Competitiveness trade agreements that are narrowly conceived. Measures to The third plank of a development-focused approach is facilitate trade—to get goods and services in and out of the expanded development assistance to help address supply country for less cost—are likely to be particularly important capacity constraints in poor countries—the types of in many of the poorest countries. Given that many of these measures briefly discussed in the section on trade liberal- countries are landlocked, cooperation with neighbors to ization strategies. This effort requires identifying and pri- reduce the costs of transit and transport and of access to oritizing needs and providing funds to address them. ports may well generate a particularly high payoff. In general, Especially in small, low-income countries that already have measures aimed at improving the investment climate are relatively free access to major markets, using aid to address likely to dominate trade policy, as are macroeconomic poli- constraints that reduce their competitiveness can have high cies designed to ensure a realistic exchange rate and actions payoffs and, indeed, may be the primary source of benefits. guaranteeing that markets exist and function. A major lesson of experience with projects and programs The implication is that financial and technical resources in the trade area (and in most others) is that country own- made available by high-income countries to developing- ership and leadership at the highest levels are critical fac- country PTA partners should be allocated on the basis of tors in ensuring concrete and sustained follow-up in national priorities and not tied to the narrow ambit of removing constraints to trade expansion. The flexibility whatever is embodied in the PTA. As argued in the next mechanism proposed above could help mobilize this section, a corollary of this is that it would be desirable to follow-up by identifying where specific investments are integrate PTA-based resources into the emerging multilat- likely to be needed, but it will need to be complemented by a eral mechanisms in order to assist poor countries in bol- comprehensive diagnostic analysis of factors that constrain stering their trade capacity. supply responses and reduce competitiveness. Such diag- nostics should feed into the process through which coun- Private Sector Engagement in Implementation tries determine public investment allocations and policy and Enforcement reform priorities. In many low-income countries, this process increasingly centers on poverty reduction strategy To be credible and meaningful, PTA commitments must be papers, which form the basis for the provision of donor enforceable. Signing a PTA is one thing, but implementing assistance at the country level. As discussed at greater it and enforcing its provisions is quite another. Much of the North-South Preferential Trade Agreements 105 literature on PTAs tends to focus on the texts and the cov- the dispute resolution mechanisms of the PTA often erage of disciplines; little attention is given to monitoring choose to resort to the WTO instead. The U.S.–Mexico and assessing implementation. The same is true of enforce- Telmex dispute is an example. Piérola and Horlick (2007) ment, about which not much is known even in the best- provide other instances in which countries went to the documented PTAs. These are areas in which the private WTO because PTA rules were ambiguous or nonexistent; sector can play a major role. Mechanisms to encourage they conclude that case law under NAFTA and similar greater participation by firms, nongovernmental organiza- agreements has entailed “little or no jurisprudential devel- tions (NGOs), and consumer groups need to be designed opment� (Piérola and Horlick 2007, 891). Bown and and incorporated into PTAs. Hoekman (2005 and 2008) discuss at greater length how What matters to firms and consumer groups is PTAs can be complemented by mechanisms through whether market segmentation is being reduced. They which firms can more easily (i.e., at lower cost) obtain have an interest in knowing about the removal of tariff information on potential violations of agreements and on and nontariff barriers and in having real-time informa- the way institutions might be designed so that the behav- tion on what is happening at border crossings and how ior of government agencies can be contested directly by regulatory requirements are being enforced. Regular the private sector. engagement between government and these groups— informed by an annual process through which informa- The Proposed Approach and the tion on the implementation of the PTA is generated Evolving Status Quo (e.g., through a survey of exporters and importers)— would provide a valuable feedback loop and help increase The need for policy flexibility and aid for trade will vary ownership of the PTA. by country: priorities differ, capacity differs, and the poli- Binding commitments (on market access, for example) tics (what is feasible and to what extent there is a need to need to be enforced. Dispute settlement provisions vary use trade agreements to pursue or lock in reforms) dif- widely across PTAs but tend to be weak in many cases. This fers. Thus, differentiation in the agreements is called for. weakness reduces the relevance of the PTA for firms. The In practice, it is already applied in the PTA context. The U.S., NAFTA-type PTAs are by far the most extensive in front-runner is arguably the EU; the European Commis- their dispute settlement provisions—not surprisingly, in sion has stressed that development is an explicit objective the sectors where there are strong lobbies in the United underpinning its pursuit of PTAs with developing coun- States. These are, first and foremost, IPRs and investment tries. The economic partnership agreements (EPAs) have protection and also areas such as product standards and attracted by far the most attention, but to date most EPAs conformity assessment. The latter is actually rather asym- have not addressed behind-the-border policies. (An metric. In the Central American Free Trade Agreement, exception is the agreement with the Caribbean countries, signatories are subject to disciplines to enhance the likeli- which has yet to be implemented.) More informative are hood that U.S. certification of goods will be accepted the approaches the EU is pursuing with neighboring as equivalent, but there is no similar language on U.S. countries that are not accession candidates and with acceptance of the partners’ certification. The seriousness which it already has reciprocal free trade agreements. of U.S. implementation is reflected not only with respect to Cooperation with a number of these countries is now formal, binding dispute settlement but also in terms of under the umbrella of the 2004 European Neighborhood calling for, and setting up, bodies to monitor implementa- Policy (ENP). The ENP has a threefold goal: (a) to sup- tion. For example, the Web site of the Office of the U.S. port the national development strategy of a partner Trade Representative contains documents on compliance country; (b) to integrate partners into some EU eco- by partners, and U.S. PTAs call for and have established nomic and social structures (a stake in the Internal Mar- performance benchmarks and contact points through ket); and (c) to implement existing and future PTAs and which interested parties (citizens) can report perceived association agreements. Technical and financial assistance instances of noncompliance.12 (development cooperation) will focus on the areas that Compared with the active caseload of the WTO, which are identified as priorities under country-specific ENP has adjudicated more than 400 cases since 1995, for most action plans (CEC 2004). North-South PTAs, there is very little evidence of enforce- A premise underlying the ENP is to pursue differenti- ment action, even for U.S.-type PTAs (aside from ated convergence with EU norms—competition policy, NAFTA). Countries that are PTA members and could use regulatory action for services liberalization, and so on. The 106 Bernard Hoekman ENP’s explicit recognition of differences in capacity and progress will be idiosyncratic. Whether an instrument pro- priorities in the context of bilateral economic cooperation motes development (growth, employment creation, and so with its neighbors, and of the need to complement binding on) will depend on what is done. Major questions, then, are treaties (PTAs) with soft law–type cooperation and techni- whether deeper integration would help growth and, more cal and financial assistance, implies that the policy has par- important in the short run, in what areas integration will allels with the proposals set forth in “Harnessing Regional generate the highest payoffs. Table 4.1 simply indicates Integration for Development,� an earlier section. Partner whether the expected sign is positive in terms of realizing countries have to determine whether they want to pursue the objective. From a practical policy-making perspective, integration and, if so, in what areas and how. This presup- it is also necessary to know the rank ordering of policy poses an understanding of the benefits and costs of alterna- instruments, in which areas the goal should be to make tive instruments of cooperation, in particular, hard law binding commitments, and in which areas the focus should (expansion of the coverage of binding treaties) versus soft be on cooperation and aid. law (economic cooperation). Hard law—binding treaty instruments—involves the As table 4.1 illustrates, this is a nontrivial challenge. The extension of association agreements to include services, table maps two objectives—market integration and eco- agriculture, and possibly parts of the EU law (acquis). It is nomic development—against EU instruments. It assumes straightforward to conceptualize hard law with respect to that the EU cares about both the core objective of eco- simple market access—for example, reciprocal agreements nomic integration and economic development, whereas to provide better access for agricultural and services flows, the partner country cares exclusively about national devel- including the right of establishment (investment). There opment. Thus, for the partner country, integration is an is, however, likely to be limited scope for reciprocity when instrument, whereas for the EU it is a goal in itself, as well it comes to the acquis, which is essentially nonnegotiable. as an instrument. As can be seen from the first row, all the Here, the issue is what will be asked of partner countries various EU instruments have the potential to contribute to and the extent to which assistance is offered to achieve the achieving the goal of integration, although, in practice, required minimum standards. In practice, whether an à la much will also depend on the national policies that are carte approach is possible may depend importantly on the pursued by partner countries (e.g., implementation). It is extent to which use is, or can be, made of the recognition not obvious, however, that the instruments the EU has principle, given that regulatory convergence (harmoniza- available will necessarily help achieve development objec- tion) may not be beneficial from a national development tives. With the exception of market access, whether this perspective. It also may or may not be necessary for effec- happens will depend very much on what a specific measure tive access to the EU markets concerned (agriculture and will do in or for the country concerned—the extent to services) or for the abolition of the threat of antidumping which it addresses priority needs. The same point applies and safeguard actions. more generally to the question of whether actions to pur- Indeed, integration (defined by the acquis) may be sue integration will promote development. second-order in terms of payoffs if the associated market There is tension between national development and access benefits are much smaller than gains from purely integration, in that the latter constitutes a unique focal domestic reform. Take the example of services. The case for point which is defined by existing EU members, even if it is action to lower costs and improve quality is well known. one that is constantly evolving. The former has no such There are also incentives for domestic agents to support focal point; both the goal and the metric used to assess services reform, especially in sectors (finance, transport, Table 4.1. European Union (EU) Instruments and EU and National Objectives EU instruments Treaties Soft law Aid Border Internal market Economic Participation in National Objective barriers (acquis) cooperation common programs Grants Loans policies Integration (EU) X X X X X X X Development (EU and partners) X ? ? ? ? ? X Source: Author’s elaboration. Note: X = effective instrument; ? = sign of impact uncertain. North-South Preferential Trade Agreements 107 health, education, and so on) that are inputs into produc- good if they are designed in a way that puts development tion and consumption and that, thus, affect large segments first. Taking development seriously has a number of impli- of the population. Although, in principle, the pursuit of cations, including identification of the most appropriate market-opening reforms through trade agreements can be form of a PTA and its membership. The need to avoid trade motivated on standard political-economy grounds, if there diversion costs and attenuate tariff revenue losses is well is not enough of a domestic constituency to support known, as is the policy recommendation of complement- autonomous reform, it may be difficult to put this in prac- ing reciprocal liberalization with reductions in external tice. Thus, there may not be sufficient (or any) export (MFN) barriers to trade. The challenge is to move in this interests, or alternatively, they may be concentrated in sen- direction, which requires a willingness by the major traders sitive sectors—mode 4, for example—where the scope for to support MFN liberalization by developing-country the EU to make concessions is restricted, given the limited counterparts as an appropriate quid pro quo for preferen- mercantilist value of access to the partner markets. This tial access to their markets. Formally, this process will implies that it will be important to mobilize EU groups, require a renegotiation of GATT Article XXIV (and of such as NGOs, which attach value to the attainment of GATS Article V, which allows for agreements to liberalize development objectives. There are also potential downsides trade in services) or a waiver. Although past experience insofar as partner countries are already pursuing unilateral does not furnish cause for great optimism that this is feasi- services reforms, driven by a desire to improve interna- ble, the effort would provide a signal that development tional competitiveness. Putting the bilateral or regional concerns are being taken seriously. Given, however, that the opening of services on the negotiating table may slow WTO disciplines in this area are effectively redundant, in desirable reforms if governments perceive more open mar- practice this may not be a serious possibility. Insofar as kets as a bargaining chip. developing country partners do undertake MFN reforms, Analysis aimed at prioritizing policy measures and it may be easier to seek a waiver. related actions is therefore critical. Benefiting from North- Market access commitments should be binding to ensure South PTAs requires a coherent national development that they are credible. The acid test for whether regulatory strategy in which trade-related and integration-related disciplines in trade agreements should be binding is whether measures are part of a country’s overall agenda. Hard law, benefits outweigh costs. Often, there will be uncertainty as to soft law (economic cooperation), and financial and techni- whether this is the case. Mechanisms to exchange informa- cal assistance options all need to be clearly mapped to the tion on the effects of policies and the development of rules of pursuit of national priorities. Hard law can be useful and thumb for behind-the-border, trade-related policies could be beneficial for overcoming political-economy resistance to very beneficial for developing countries. That suggests that reform, reducing uncertainty, and locking in market access. greater reliance on a soft-law approach that establishes broad But integration for its own sake, or the adoption of the EU guidelines and relies on transparency and accountability model, is not necessarily going to be beneficial, and even if through regular (multilateral) monitoring of performance it is, it may not be a priority at a given point in time. Mech- may be more effective in promoting development than har- anisms to generate information and analysis of the impacts monization, not least because it will enhance the eventual of different options, ex ante and ex post, are therefore of ownership of any specific norms that are adopted. Although great importance. The lack of such mechanisms is perhaps detailed international harmonization through trade agree- the weakest element of the evolving status quo. ments may not be appropriate in many instances, one cannot generalize here except to note that careful, country-specific analysis and assessments are called for. International disci- Conclusions plines embedded in PTAs, starting with monitoring, analysis A precondition for benefiting from trade agreements is a of impacts of policies, and information exchange, can help clear understanding of the objectives to be achieved—in ensure transparency and promote increased accountability of particular, the type of trade policy that the government governments. wants to pursue. Trade agreements can play a useful role in The policy agenda confronting developing countries at the design and implementation of trade reforms, acting as the regional (PTA) level is similar to that at the multilateral a mechanism for locking in reforms, a focal point for level. A major difference is that in the case of North-South future reforms, and a device to help overcome resistance by PTAs, more significant development assistance commitments vested interests. are associated with the implementation of agreements. The PTAs that the EU and the United States are increas- Trade-related assistance should focus on national, country- ingly negotiating with developing countries can do much specific priorities; after all, there is only one national trade 108 Bernard Hoekman agenda. Technical and financial assistance should be man- 5. The relevant agreements on IPRs and services are the Trade- Related Aspects of Intellectual Property Rights (TRIPS) agreement and aged accordingly. the General Agreement on Trade in Services (GATS). In the past five years, much progress has been made by 6. Recent empirical analysis, however, has begun to reveal evidence the international community toward improving aid effec- that PTAs generate incentives to pursue MFN liberalization once prefer- tiveness and prioritization of aid across sectors and uses. ential market access reforms have been implemented. See Bohara, Gawande, and Sanguietti (2004) as well as Estevadeordal, Freund, and This also spans the trade area, the Enhanced Integrated Ornelas (2008). The driver for such complementary MFN liberalization Framework being a prominent example. Trade will be one may be a desire to reduce the trade diversion costs of the PTA. priority among many and should, therefore, be considered 7. Although preferential liberalization by OECD members of PTAs is, of course, undesirable from a global welfare perspective, the solution is in the context of the overall national development strategy. continued multilateral negotiations to remove trade-distorting policies on Ensuring that trade-related assistance provided in the con- an MFN basis. text of PTAs conforms to this principle would further 8. This commitment should not be difficult, insofar as partners are small, poor countries. Both the United States and the EU already give improve the potential development impact of PTAs, as it many of these countries duty-free access to markets. implies that national priorities come before PTA-defined 9. If a North-South PTA has a South-South PTA as its partner (as requirements. opposed to only one developing country), then arguably the latter PTA Indeed, much of what has been suggested in this chapter also should completely liberalize trade to ensure that an integrated market is indeed being created. Rather than continuing to use trade policies should be pursued multilaterally. There is only one domes- (including nontariff barriers) to shelter local firms from competition, tic trade agenda in developing counties, one set of regula- other policies should be used to address coordination problems, under- tory reform priorities, and one first-best trade policy. The provision of public goods, and other market failures and to achieve equity objectives. Import-substitution policies that rely on trade protection have rules of thumb proposed for PTAs are generalizable. In all proved to be largely ineffective and costly. the areas proposed, the WTO could play a supporting 10. In a comprehensive retrospective on the East Asian development role—by requiring a shift from discriminatory to MFN experience, Noland and Pack (2003) argue that sector-specific policies did not result in high rates of total factor productivity growth for manufac- reforms in a PTA context, by enhancing information on turing. In the case of Korea and Taiwan, China, productivity growth was and analysis of the impacts of PTAs and national trade- not much higher than in OECD economies. The authors argue that the related policies, and by encouraging donors to channel primary reason these countries developed rapidly has more to do with through multilateral aid mechanisms the resources that are economywide policies in areas such as education and infrastructure than with industrial targeting, not least because the government did not prove mobilized to support PTA implementation. very effective in identifying winners. 11. International financial institutions could be brought into this process, in an advisory capacity. Their involvement would be desirable for Notes at least two reasons. First, they have the mandate, experience, local pres- ence, and capacity to provide policy advice. Second, these organizations This chapter draws on joint work with Alan Winters and is a substantially generally take the lead in the development and financing of projects and revised version of a paper presented at the International Trade Round- programs in developing countries. table “The WTO at 10 Years: The Regional Challenge to Multilateralism,� 12. A noteworthy feature of recent U.S. PTAs is that dispute settlement Brussels, June 27–28, 2005. Parts of that paper were subsequently pub- makes provision for compensatory payments in lieu of implementation lished as Hoekman (2006) and Hoekman (2007). (or retaliation). In the U.S.-Chile agreement, in the case of nonimplemen- 1. There is extensive literature on the effect of preferences. See Hoek- tation of a panel finding, the losing party may offer to pay 50 percent of man and Özden (2007) for a survey and summary of both the old and the damage caused. It appears that this penalty is open-ended, as the text more recent literature. speaks of annual payments. This PTA also provides for monetary fines 2. Of course, this is not always true. For example, a trade-off between of up to US$15 million per year in case of violations of labor or environ- trade liberalization and other objectives can arise in the short run if too mental provisions. Proceeds go into a fund earmarked for labor initiatives large a shock would lead to the complete collapse of a market. Local labor or green initiatives. markets may seize up in the face of large-scale redundancies because workers cease to leave their jobs speculatively for fear of not finding others. This is essentially a matter of timing—of staggering the trade lib- eralization (as is very common in practice) and ensuring that it is not References accompanied by a negative macro shock. 3. In many countries, import-competing firms were compensated for Abbott, Kenneth, and Duncan Snidal. 2000. “Hard and Soft Law in Inter- trade reform by depreciation and other reforms that reduced their input national Governance.� International Organization 53 (3): 421–56. costs—for example, measures to improve the efficiency of service providers. Barth, James R., Gerard Caprio, Jr., and Ross Levine. 2006. Rethinking 4. Bilateral investment treaties (BITs) have been the primary vehicle Bank Regulation: Till Angels Govern. Cambridge, U.K., and New York: for international cooperation in this area. BITs are, in part, designed to Cambridge University Press. spur investment flows by providing recourse to international dispute res- Bohara, Alok, Kishore Gawande, and Pablo Sanguietti. 2004. “Trade olution in the event of conflict with governments. Although some PTAs Diversion and Declining Tariffs: Evidence from Mercosur.� Journal of explicitly do not include investment policy disciplines because of preexist- International Economics 64: 65–88. ing BITs—an example is the 2000 Canada–Costa Rica PTA (Gestrin Bown, Chad P., and Bernard Hoekman. 2005. “WTO Dispute Settlement 2002)—most recent U.S. PTAs have more ample rights and coverage and and the Missing Developing Country Cases: Engaging the Private Sec- generally subsume prior BITs. tor.� Journal of International Economic Law 8 (4): 861–90. North-South Preferential Trade Agreements 109 ———. 2008. “Developing Countries and Enforcement of Trade Agree- Konan, Denise, and Keith Maskus. 2006. “Quantifying the Impact of Serv- ments: Why Dispute Settlement Is Not Enough.� Journal of World ices Liberalization in a Developing Country.� Journal of Development Trade 42 (1): 177–203. Economics 81 (1, October): 142–62. CEC (Commission of the European Communities). 2004. “European Limão, Nuno. 2007. “Are Preferential Trade Agreements with Non-Trade Neighbourhood Policy: Strategy Paper.� COM(2004) 373 final, CEC, Objectives a Stumbling Block for Multilateral Liberalization?� Review Brussels. http:// ec.europa.eu/world/enp/pdf/strategy/strategy_paper_ of Economic Studies 74 (3): 821–55. en.pdf. Mavroidis, Petros. 2005. “Testing the Compliance of PTAs with the Multi- Chayes, Abram, and Antonia Handler Chayes. 1995. The New Sovereignty: lateral Rules.� Presented at the International Trade Roundtable 2005, Compliance with International Regulatory Agreements. Cambridge, “The WTO at 10 Years: The Regional Challenge to Multilateralism,� MA: Harvard University Press. Brussels, June 27–28. Estevadeordal, Antoni, Caroline Freund, and Emmanuel Ornelas. 2008. Newfarmer, Richard, William Shaw, and Peter Walkenhorst, eds. 2009. “Does Regionalism Affect Trade Liberalization Towards Non- Breaking into New Markets: Emerging Lessons for Export Diversifica- Members?� Quarterly Journal of Economics 123 (4): 1531–75. tion. Washington, DC: World Bank. Fink, Carsten, and Martín Molinuevo. 2007. “East Asian Free Trade Noland, Marcus, and Howard Pack. 2003. Industrial Policy in an Era of Agreements in Services: Roaring Tigers or Timid Pandas?� World Globalization: Lessons from Asia. Washington, DC: Institute of Inter- Bank, Washington, DC. http://siteresources.worldbank.org/INTEAP national Economics. SUMEASPR/Resources/2576847-1163691185244/East_Asian_FTAs_ Pack, Howard and Kamal Saggi, 2006. “Is There a Case for Industrial Pol- in_Services.pdf. icy? A Critical Survey.� World Bank Research Observer 21 (2): 267–97. Francois, Joseph, and Bernard Hoekman. 2009. “Services Trade and Piérola, Fernando, and Gary Horlick. 2007. “WTO Dispute Settlement Policy,� CEPR Discussion Paper 7616, Centre for Economic Policy and Dispute Settlement in the ‘North-South’ Agreements of the Research, London. Americas: Considerations for Choice of Forum.� Journal of World Gestrin, Michael. 2002. “The Relationship between Regional Trade Agree- Trade 41 (5): 885–908. ments and the Multilateral Trading System: Investment.� TD/TC/ Porto, Guido, and Bernard Hoekman, eds. 2010. Trade Adjustment Costs in WP(2002)18/FINAL, Organisation for Economic Co-operation and Developing Countries: Impacts, Determinants and Policy Responses. Development, Paris. Washington, D.C.: World Bank and Centre for Economic Policy Haddad, Mona, Jamus Jerome Lim, and Christian Saborowski. 2010. “Trade Research (CEPR). Openness Reduces Growth Volatility When Countries Are Well Diversi- Prowse, Susan. 2002. “The Role of International and National Agencies in fied.� Policy Research Working Paper 5222, World Bank, Washington, Trade–Related Capacity Building.� World Economy 25 (9, September): DC. http://econ.worldbank.org/external/default/main?pagePK=641 1235–61. 65259&theSitePK=469382&piPK=64165421&menuPK=64166093& ———. 2006 “Aid for Trade: Increasing Support for Trade Adjustment entityID=000158349_20100226112948. and Integration—A Proposal.� In Economic Development and Multi- Helleiner, Gerald Karl. 2000. “Markets, Politics, and Globalization: Can lateral Trade Cooperation, ed. Simon Evenett and Bernard Hoekman. the Global Economy be Civilized?� 10th Raúl Prebisch Lecture, United Washington, D.C.: World Bank; Houndmills, Basingstoke, Hampshire, Nations Conference on Trade and Development, Geneva, December U.K: Palgrave Macmillan. 11. http://www.unctad.org/en/docs/prebisch10th.en.pdf. Rodríguez-Clare, Andres. 2004. “Clusters and Comparative Advantage: Hoekman, Bernard. 2005. “Operationalizing the Concept of Policy Space Implications for Industrial Policy.� Inter-American Development in the WTO: Beyond Special and Differential Treatment.� Journal of Bank, Washington, DC. International Economic Law 8 (2): 405–24. Rodrik, Dani, 2004. “Industrial Policy for the Twenty-First Century.� ———. 2006. “Trade Liberalization, Trade Agreements and Economic CEPR Discussion Paper 4767, Centre for Economic Policy Research, Development.� In The WTO Trade Remedy System: East Asian Per- London. spectives, ed. Mitsuo Matsushita, Dukgeun Ahn, and Tain-Jy Chen. Roy, Martin, Juan Marchetti, and Hoe Lim. 2006. “Services Liberalisation London: Cameron May. in the New Generation of Preferential Trade Agreements: How Much ———. 2007. “Regionalism and Development: The European Neighbor- Further Than the GATS?� WTO Staff Working Paper ERSD-2006-07, hood Policy and Integration à la Carte.� Journal of International Trade World Trade Organization, Geneva. and Diplomacy 1 (1): 1–55. Sabel, Charles F., and Sanjay G. Reddy. 2007. “Learning to Learn: Undoing Hoekman, Bernard, and Çaglar Özden, eds. 2007. Trade Preferences and the Gordian Knot of Development Today.� Challenge 50 (5, Differential Treatment of Developing Countries. Cheltenham, U.K.: Sept.–Oct.): 73–92. Edward Elgar. Schiff, Maurice, and L. Alan Winters. 2003. Regionalism and Development. Hoekman, Bernard and L. Alan Winters. 2009. “Multilateralizing Preferential Oxford, U.K.: Oxford University Press. Trade Agreements: A Developing Country Perspective.� In Multilateraliz- Winters, L. Alan. 2004. “Trade Liberalization and Economic Performance: ing Regionalism: Challenges for the Global Trading System, ed., Richard An Overview,� Economic Journal 114 (February), F4–F21. Baldwin and Patrick Low. Cambridge, U.K.: Cambridge University Press. World Bank. 2001. Global Economic Prospects and the Developing Coun- Imbs, Jean, and Romain Wacziarg. 2003. “Stages of Diversification.� tries 2002: Making Trade Work for the World’s Poor. Washington, DC: American Economic Review 93: 63–86. World Bank. Jensen, Jesper, Thomas Rutherford, and David Tarr. 2007. “The Impact of ———. 2010a. Global Monitoring Report 2010: The MDGs after the Crisis. Liberalizing Barriers to Foreign Direct Investment in Services: The Washington, DC: World Bank. Case of Russian Accession to the WTO.� Review of Development Eco- ———. 2010b. World Trade Indicators 2009/2010: Trade under Crisis. nomics 11 (3): 482–506. Washington, DC: World Bank. 5 Customs Unions Soamiely Andriamananjara A customs union (CU) is a form of trade agreement A Special Case of PTA under which certain countries preferentially grant tariff- CUs have been around for a long time and were once more free market access to each other’s imports and agree to prevalent than FTAs. Early efforts toward economic inte- apply a common set of external tariffs to imports from gration were generally driven by the desire to establish a the rest of the world. That is, they enter into a free trade political union, and the members were willing to relin- agreement (FTA) and apply a common external tariff quish some political autonomy.2 Early examples include (CET) schedule to imports from nonmembers.1 A CU can the Zollverein, formed in 1834 by several German princi- be thought of as a deeper form of integration than an palities, which turned out to be a step toward political uni- FTA, generally requiring more coordination and a greater fication, and the 1847 customs union between Moldavia loss of autonomy. and Walachia, a precursor to the creation of Romania. The aims of this chapter are to provide, from an eco- More recently, CUs appear to have become less popular, nomics perspective, an overview of the key features of at least with respect to the number of arrangements. As is CUs and to examine some design issues that may be of thoroughly documented by Acharya et al. (ch. 2 in this interest to policy makers and (nonspecialist) analysts. The volume), almost 85 percent of the regional integration discussions are meant to be relatively conceptual and arrangements notified to the World Trade Organization nontechnical, but real-world illustrations are provided (WTO) through 2009 consisted of FTAs. This trend when available. reflects the nature of the current wave of regionalism, To begin with, the main economic costs and benefits of which has been broadly characterized by smaller cross- opting for a CU, relative to those for an FTA, are discussed, regional deals, flexibility, selectivity, and, most important, and selected issues regarding the design and determinants speed.3 Recent FTAs are inclined to be pragmatic and to of a common external tariff are examined. Although a focus more on strategic commercial market access and number of arguments seem to suggest that CUs may be less on geographic considerations or political ambitions.4 subject to more protectionist pressures than FTAs, the They generally involve a small number of partners (fre- existing literature does not provide an unequivocal answer. quently, just two), which are often geographically distant There then follows a conceptual discussion of the implica- from each other. They tend to achieve significant prefer- tions of the various administrative options related to the ential and reciprocal trade liberalization within a short collection and sharing of customs duties; this, it is shown, time while simultaneously preserving a member’s sover- is not only a technical issue, but also (and perhaps more eignty over its trade policy vis-à-vis the rest of the world, important) a question of trust among member countries. including its option of joining other preferential trade The chapter concludes with an overview of a number of agreements (PTAs). systemic aspects of CUs. Readers interested in the more By contrast, CUs usually involve a relatively large num- operational and detailed aspects of CUs are referred to the ber of geographically contiguous countries (see table 5.1 for excellent surveys in Development Network Africa (2007) a selected listing). They generally take longer to negotiate and Keick and Maur (2011). 111 112 Soamiely Andriamananjara Table 5.1. Selected Customs Unions, in Force and Planned Agreement Date In force Southern African Customs Union (SACU) 1910 Switzerland–Liechtenstein 1924 European Union (EU) January 1, 1958 Central American Common Market (CACM) October 12, 1961 Caribbean Community (CARICOM) August 1, 1973 Andean Community (CAN) May 25, 1988 EU–Andorra July 1, 1991 Southern Cone Common Market (Mercosur, Mercado Común del Sur) November 29, 1991 Israel–Palestinian Authority 1994 EU–Turkey January 1, 1996 Eurasian Economic Community (EAEC) October 8, 1997 Economic and Monetary Community of Central Africa (CEMAC, Communauté Économique et Monétaire de l’Afrique Centrale) June 24, 1999 West African Economic and Monetary Union/Union Économique et Monétaire Ouest-Africaine (WAEMU/UEMOA) January 1, 2000 East African Community (EAC) July 7, 2000 EU–San Marino April 1, 2002 Gulf Cooperation Council (GCC) January 1, 2003 Customs Union of Belarus, Kazakhstan, and Russia July 1, 2010 Planned Arab Customs Union (ACU) 2010 Southern African Development Community (SADC) 2010 Economic Community of Central African States (ECCAS) 2011 Economic Community of West African States (ECOWAS) 2015 African Economic Community (AEC) 2019 Arab Common Market (ACM) 2020 Australia–New Zealand Closer Economic Relations Trade Agreement (ANZCERTA) 2020 Source: Author’s compilation. and implement than do PTAs, and they entail a certain loss of Central African States (ECCAS)—take in virtually every of policy-making autonomy. By 2009, they accounted for country in the region. Many countries in the Middle East and less than 10 percent of the regional integration arrangements North Africa regions are members of the Gulf Cooperation notified to the WTO. CUs are less numerous than PTAs, but Council (GCC), the Arab Customs Union (ACU), or both. they generally have much larger memberships. They also tend Although a CU is no longer the most popular option, to cover much larger geographic areas. The four main CUs in it remains a central component of regional integration Latin America—the Central American Common Market strategy in many developing regions. For many developing (CACM), the Andean Community (CAN), the Caribbean countries, the design and implementation of a CET, along Community (CARICOM), and the Southern Cone Common with the elimination of intraregional trade barriers, con- Market (Mercosur, Mercado Común del Sur)—include tinue to be key drivers of trade policy reform and to almost all the region’s economies. The existing and planned occupy an important place in policy debates. Given that CUs in Sub-Saharan Africa—the Economic and Monetary customs duties constitute a significant source of govern- Community of Central Africa (Communauté Économique et ment revenues in most of those countries, choosing the Monétaire de l’Afrique Centrale), the East African Commu- appropriate mechanism for collecting and allocating cus- nity (EAC), the Southern African Customs Union (SACU), toms revenues is an important challenge for officials. the West African Economic and Monetary Union (WAEMU), the Economic Community of West African Economic Implications of Customs Unions States (ECOWAS), the Common Market for Eastern and Southern Africa (COMESA), the Southern African Develop- Chauffour and Maur (ch. 1 in this volume) discuss in detail ment Community (SADC), and the Economic Community the economic, societal, and political-economy motives for Customs Unions 113 signing preferential trade agreements. A central issue is form of an FTA, member countries grant free trade to each whether to opt for an FTA or a CU. By definition, both are other but effectively maintain sovereignty over the con- preferential in nature and discriminate against third-party duct of trade policy vis-à-vis the rest of the world. Thus, (nonmember) suppliers. The primary effect of a customs the tariffs charged to nonmember suppliers will vary union, as with an FTA, is the expansion of trade flows among across members. This could lead to opportunities for trade member countries, often at the expense of trade with non- deflection—a situation in which goods from outside the members. This expansion, a consequence of the removal of FTA are shipped to a low-tariff country and then trans- the intraunion tariff barriers, can be decomposed into trade shipped tariff-free to the high-tariff country. Such round- creation (more efficient suppliers in CU partners replace about shipping patterns, which have the sole purpose of domestic suppliers of a given good) and trade diversion exploiting the existing tariff differential, are inherently (more efficient third-party suppliers are displaced by less inefficient and can create friction among members.8 efficient suppliers located in partner countries, as a result of One way to avoid such wasteful trade deflection is for the discriminatory liberalization).5 As is well established in the members of the FTA to adopt a rules-of-origin system. the literature, when trade diversion dominates trade cre- Rules of origin can take various forms, but generally they ation, CUs and FTAs tend to be welfare reducing (Viner require that goods (or value added) qualifying for tariff- 1950).6 The likelihood of significant trade diversion is free trade be produced within the FTA and that imports closely related to the degree of discrimination associated from outside the FTA pay the tariff of the final destina- with the agreement (see Baldwin, ch. 3 in this volume). tion country, even if they pass through another member country (see Brenton, ch. 8 in this volume). In practice, rules of origin are particularly complex, and their imple- Reasons for Choosing a Customs Union mentation costs can be high.9 They necessitate significant There are many possible rationales for choosing a CU over internal border controls to ensure compliance and to col- an FTA, including political and economic ones. Some lect the relevant customs duties. regional groupings consider the establishment of a CU a Another way to prevent trade deflection is to establish prerequisite for the future establishment of a political a customs union, which would require all members to union, or at least some deeper form of economic integra- apply the same external tariff to imports coming from tion, such as a common market.7 The African Economic outside the union. Because of the common external tariff Community provides an illustration. The Abuja Treaty of (i.e., the absence of tariff differentials across members), 1991 envisaged gradual implementation in the following the potential for trade deflection and the need for intra- stages: (a) creation of regional blocs, by 1999; (b) strength- union border inspections are, theoretically, minimized. In a ening of intrabloc integration and interbloc harmonization, fully implemented customs union, it is no longer neces- by 2007; (c) establishment of an FTA and then a CU in each sary to maintain internal border controls for customs regional bloc, by 2017; (d) establishment of a continentwide duty purposes or to design and implement the cumber- customs union, by 2019; (e) realization of a continentwide some and costly rules of origin that are necessary in a African Common Market (ACM), by 2023; and (f) creation free trade area in which members have different external of a continentwide economic and monetary union (and tariff structures.10 The simplification offered by a CU can thus also a currency union) and a parliament, by 2028. greatly facilitate cross-border trade, which is especially Some groups, such as CARICOM, consider a CU to be relevant because existing CUs generally involve geograph- a useful way of pooling market power, coordinating trade ically contiguous countries, reflecting the traditional policies, and combining efforts to negotiate with the rest objective of regional integration.11 In this regard, a CU can of the world. The more intense degrees of coordination approximate a larger single market (as compared with a and interaction associated with a CU can foster trust and number of separate markets in an FTA), which can gener- familiarity among the parties and may even decrease the ate greater economies of scale, as well as procompetitive risk of conflicts, as has been the case with the European pressures. These, in turn, can greatly benefit consumers Union (EU). The fact that the external tariff is agreed with and can translate into lower business costs and enhanced other parties through a legal agreement may help reform- competitiveness for member countries. minded governments lock in their trade policies and can shelter them from domestic lobbies. Economic Implications of the Common External Tariff On a lower level, a customs union can simply be a prac- tical device for avoiding trade deflection while facilitating As was mentioned earlier, the key difference between a CU more fluid trade flows among member states. In the simplest and an FTA is the need to adjust the tariff structure applied 114 Soamiely Andriamananjara to third-party suppliers, at least for some members. Coun- discriminatory aspect of the FTA is, in a sense, diluted in tries that join an FTA are not required to change the tariffs this case. Starting from an FTA situation, a decrease in B’s they apply to imports from the rest of the world.12 What external tariff would have two effects, both working in the will differentiate the effects of a CU from those of an FTA same direction. First, the liberalization will directly will be the extent to which the external tariff is increased or increase B’s imports from the rest of the world (trade cre- decreased by a given member with respect to a given good. ation). Second, since it effectively dilutes the existing The net economic effect of a CU crucially depends on how preference margin, it will reduce the attractiveness of the adjustment of the external tariff affects the degree of sourcing from country A relative to sourcing from the rest discrimination vis-à-vis nonmember countries. of the world (less trade diversion). By reversing the trade In order to isolate the impact of a CU, it is useful to diversion caused by the FTA (i.e., by inducing consumers start with a case in which an FTA is already in place (i.e., to switch from less efficient suppliers in A to more effi- trade is already liberalized among the partners) and mem- cient ones in the rest of the world), country B’s tariff ber countries are considering establishing a customs reduction will benefit its consumers. It could also help union by harmonizing their external tariff duties.13 For increase government revenues. as dutiable imports from ease of presentation, the following discussion assumes the rest of the world expand (albeit at lower tariffs) and that two countries, A and B, are members of an FTA and as country B shifts to dutiable imports and away from have decided to form a customs union. Without loss of duty-free imports from A. Domestic producers will face generality, it will be further assumed that, for a particular more competition from nonmembers, but this will be good, A has a low tariff and B has a high one. Two possi- offset by consumer gains resulting from lower prices and, ble cases are relevant and are examined here. potentially, by higher tariff revenues. Although adopting One possibility is that the agreed common external a lower CET may not lead to tariff-jumping investment, the tariff (CET) leads to a higher tariff rate for a given CU higher returns associated with the more liberal economic member (say, country A). The bloc’s degree of discrimi- environment could attract efficiency-seeking investments. nation is thus enhanced, and the negative impact of trade Krueger (1995) has argued that if the CET level is chosen as diversion caused by the FTA is exacerbated. This usually the union’s average tariff for a given commodity, an FTA happens when a less advanced member has to implement will not lead to more net trade creation than a CU. Fur- a CET aimed at protecting the industries of a more devel- thermore, as long as the CET is set below the tariff level of oped member.14 In this case, consumers in country A will the high-cost country, an FTA will not be more welfare lose because they have to pay higher prices for imports enhancing than a CU. from the rest of the world or switch to less efficient In the actual implementation of a CET, an individual suppliers from country B. Despite the higher external CU member will generally have to increase its external tariff, the government in country A could collect less tariff tariffs on certain products while decreasing them on oth- revenues if the higher degree of discrimination leads to a ers. The overall impact will depend on the balance. Kemp greater propensity for switching the sourcing of imports and Wan (1976) demonstrate the existence of conditions to duty-free, country B suppliers. Producers of the good in that suffice to ensure that a CU is welfare enhancing. A face less competition from the rest of the world but In particular, they show that if the CET is chosen so that more competition from B. In fact, the adoption of a high trade with the rest of the world is kept unchanged, then tariff by A effectively extends the protection received by following the establishment of a CU, welfare could poten- country B producers to country A’s markets. These pro- tially increase for all parties, including nonmembers, con- ducers may be the only ones to gain from the CU in this tingent on compensatory transfers. This increase occurs scenario. In some cases, and in a more dynamic setting, because any additional trade between CU members would the expansion of the protected market may lead to some be welfare-enhancing trade creation. Although this is an tariff-jumping types of investment in the customs union, important result, it is a “possibility� and does not guarantee motivated by the prospect of taking advantage of the that the existing political-economy equilibrium will be a larger, more protected market. welfare-enhancing one.15 Joining a CU may offer a second possibility to consoli- date the existing tariff schedule and adopt a more liberal Factors Driving the Design and Level of the trade regime. If the establishment of the CU yields a CET Common External Tariff that is lower than the pre-CU tariff (say, in country B), the potential for trade diversion is reduced, or even reversed, A well-designed and generally accepted CET is crucial for because there is less potential for switching suppliers. The the sustainability of a customs union.16 National tariffs Customs Unions 115 must be harmonized at some agreed level, taking into favored nation (MFN) tariffs by making them more uni- account not only the often-conflicting positions of each form and lower. (This is referred to as the “uniform tax member but also the various special interests within each rate principle.�) In a related fashion, if trade diversion member. Setting the level of the CET in a consensual becomes apparent (i.e., if a country sees itself importing manner could be a complex undertaking, entailing long a good from a partner country at a higher cost than the and involved negotiations between member country cost of similar goods from nonmembers), an FTA mem- governments, which are themselves subject to lobbying ber has the flexibility to cut tariffs on these third-party by different interest groups. For instance, it took the EU imports.19 Similarly, the potential for trade deflection may 11 years (1957 to 1968) to complete its CET, and Mercosur lead high-tariff countries to cut tariffs to just below the level members took four years just to agree on their nonagricul- of their partners’ rates to prevent imports from going tural CET. through low-tariff countries that would otherwise capture In many developing-country customs unions, the diffi- the tariff revenue. In addition, lowering import tariffs on culties of agreeing on a common external tariff and on the inputs used in producing exports to other FTA members distribution of revenues have proved to be so great that can render exporters more competitive.20 Do these argu- the resulting tariff schedules tend to include numerous ments apply to CUs? country or sector-specific exceptions and sensitive lists. An often-stated objective of most customs unions Although the CARICOM CET is largely in place, it allows among developing countries is to promote a harmonized broad scope for tariff reductions and suspensions, as well reduction in internal and external trade barriers in order as for national derogations. The CET in Mercosur does to better integrate the region into the multilateral trading not cover all sectors, and it includes special regimes for the system. There are, however, arguments that seem to sug- automotive and sugar sectors.17 In some CUs, temporary gest that CUs create pressures for more protectionism. national exemptions are allowed—for example, EAC Like other integration initiatives, CUs permit member members Kenya and Tanzania were allowed to unilaterally countries to combine their market size and thereby increase reduce tariffs on selected grain imports. Derogations and their market power. Since trade policy is set jointly, this safeguards are widely used in most CUs. Not only can measure could strengthen their incentive to adopt high these exceptions reduce the transparency and effectiveness CETs in order to improve their terms of trade.21 That is, of the CU, but they also can complicate trade negotiations they can reduce global demand for an imported product, and increase transaction costs. Furthermore, they reintro- and thereby decrease the import price, by charging higher duce the potential for trade deflection—the very phenom- tariffs. The larger the size of the union, the stronger this enon that the CU is designed to prevent. proprotectionist effect will be. Like most other forms of regional integration agree- Furthermore, if CU members negotiate effectively as a ments, a CU is inherently preferential and is, thus, discrimi- bloc, they can pool their negotiating power and enhance it natory against third parties. As argued in the previous sec- against the rest of the world, thus affecting the outcome of tion, the economic impact of a CU will be closely related to negotiations. Given the mercantilist nature of trade negoti- the degree of discrimination, which depends on the CET ations, increased negotiating power is likely to lead to a level that is selected. The higher the CET, the more trade more protectionist outcome (in exchange for better market diversionary will the union be. An important question is access). It could also be argued that nonmembers will act in thus whether a CU provides incentives for selecting higher a more conciliatory way when negotiating with a (single, or lower external tariffs than those in, say, an FTA.18 The large) customs union than with separate FTA members, and existing theoretical and empirical literature does not provide the result will be smaller requests for concessions. an unequivocal answer to this question. The result seems to The internal process of decision making within the CU depend on the way preferences (or objective functions) are could also place upward pressure on tariffs. The joint, con- aggregated across members and within each member. sensual determination of the external tariffs may provide A number of analysts have demonstrated that an FTA incentives to agree on higher CETs, since these imply may create downward pressure on external tariffs. (See, higher preference margins and benefit partners’ firms. for example, Richardson 1993; Bagwell and Staiger 1999; (Protection is afforded to all producers in all CU member Freund 2000.) For instance, Baldwin and Freund (ch. 6 in countries.) CU members will internalize this fact and will this volume) argue that preferential trade liberalization choose a higher external tariff (Freund and Ornelas 2010). in an FTA tends to make tariffs against nonmembers’ Accordingly, one can think of a situation in which each third-nation tariffs more distortionary and that it creates CU member feels strongly about protecting a particular an incentive for FTA members to reoptimize their most sector but would like lower tariffs on the other sectors. As 116 Soamiely Andriamananjara Winters (1996) argues, this may create a prisoner’s-dilemma Overall, whether opting for a CU leads to higher external outcome under which the CET would provide high protec- tariffs remains an open question. A number of arguments tion in all sectors, even though each country would be better seem to suggest that CUs may provide more protectionist off with low protection in all sectors.22 pressures than FTAs. This, however, remains an empirical The establishment of a CU also changes the power of question to which the existing literature has not been able to lobbies, but it is not clear whether the result will be provide an unequivocal answer. stronger or weaker demand for protection. It is possible that lobbying pressure within a CU may be diluted, com- Allocation and Collection of CET Revenue pared with national lobbying for protection within an FTA. As Winters (1996) suggests, it is more costly to lobby For most CUs among developing economies, the potential for a tariff increase in a CU than in an individual FTA for losses of tariff revenues constitutes an important nego- member country because there may be more opposition to tiation issue. These losses may result from the liberalization overcome or more representatives to influence. Moreover, of intraunion tariffs, from the adoption of the common the returns to lobbying activities are less under a CU, given tariff schedule, or from changes in trade patterns. Given that an extra 1 percent tariff protection becomes available the significance of tariff revenue for most developing to all members. Panagariya and Findlay (1996) provide a countries, at least two issues need to be addressed when formal treatment of the argument that a customs union is establishing a customs union: (a) Who has a claim on the a more effective instrument for diluting the power of collected customs duties? (b) Where and how should those interest groups than is an FTA. The high cost and low duties be collected? returns of lobbying under a CU could lead to a free-rider problem in lobbying, and all lobbying could end up taking Ownership of the Collected Duties place in one country. The author finds that such a process would yield a lower (common) external tariff under a CU The use or allocation of the collected duties is an impor- than under an FTA. The larger the size of the customs tant consideration. Should the customs revenues collected union, the lower the resulting (lobbied) level of common be treated as community property, or as income accruing external tariff would be. to each member state? Generally, it is necessary to establish The argument could, of course, cut in the other direc- a regional, supranational institution or a secretariat to tion. In some sectors, lobbyists in different member ensure smooth operations of the union. Although such an countries may be able to overcome the free-rider prob- institution could be funded through direct contributions lem, pool their resources, and cooperate. This is likely to from members—for example, WAEMU provides for an happen in sectors in which they produce relatively similar additional tax of 1 percent on imports— treating customs goods (say, in agriculture) and where there is little intra- revenues as the collective property of the union may be a bloc trade flow. In this case, the national lobbies would be more useful financing mechanism. In some cases, such as able to organize themselves into a regional lobby, and the the EU, the union may decide to allocate (a fraction of) resulting common external tariff would be higher in a CU these revenues to a joint fund to finance regional develop- relative to what would prevail in the individual markets ment initiatives or to provide support to poorer CU mem- under an FTA. bers.23 Of course, pooling customs revenues necessitates a The degree of “permanence� of the policy outcomes will high level of coordinating capacity and a certain degree of also affect the incentives for, and the amount of, lobbying. trust among members. This arrangement seems more An FTA does not require member countries to immediately likely to be sustainable when tariff revenues do not consti- adjust their external tariffs, and it preserves discretion for a tute an important part of government revenue for individ- country to adjust its trade policy in the future. By contrast, ual members—as is the case in the EU, but usually not for a CU requires both tariff adjustments and a relatively developing economies. longer-term commitment to the trade policy jointly agreed In other cases, CUs treat customs revenues as the on by the CU members—the CET. It is therefore likely that property of individual members. Collected duties are lobbying for protection would be stronger during the allocated either according to the final destination or in negotiation and establishment of a CU than in the case of line with an agreed sharing formula. Such a formula an FTA. Also, the difficulties in renegotiating or readjusting could provide for a simple reallocation based on negoti- the CET could lead to the emergence of less transparent ated and fixed shares, or it could involve a more complex nontariff barriers that would be implemented at the range of economic and demographic variables. SACU, national (instead of the regional) level. for instance, has a fairly complicated revenue-sharing Customs Unions 117 system in which the share accruing to each member is would not need to monitor the passage across their borders calculated from three basic components: a customs pool, of goods originating outside the CU for duty collection an excise pool, and a development component. The cus- purposes.26 This method would greatly enhance efficiency toms pool is allocated according to each country’s share by reducing transaction costs at internal border posts, but of total intra-SACU trade, including reexports. The excise it requires the existence of the appropriate institutional component is allocated on the basis of gross domestic capacity to administer the revenues and, most important, product (GDP). The development component (fixed at a high level of trust among members. Both of these meas- 15 percent of the total excise pool) is distributed to all ures tend to be harder to achieve as the number and SACU members according to each country’s per capita diversity of member states increase. The mechanism is GDP; that is, countries with lower per capita income will more likely to be sustainable if customs duties are deemed receive more (WTO 2003). to be community property of the CU and are used for col- Most existing customs unions allocate revenues lectively determined community purposes, or if members according to the final-destination principle. This method, can devise a mechanism for identifying imports accord- although apparently simple in theory, requires a mecha- ing to their final destinations.27 A potential issue is the nism for identifying the final destination of each shipment possibility of some diversion in revenue collections (and entering the union; the destination country would then even economic activity) away from landlocked and less claim the appropriate duty amount. One way to handle developed countries and toward the more developed this procedure is to keep the imported shipment in trading hubs in the region. bonded facilities until it reaches the country of ultimate Alternatively, customs duties could be collected at consumption. This may work for whole shipments of final the final destination or the final consumption point.28 goods that are entirely consumed in the destination coun- Although conceptually straightforward, this type of agree- try, but it may not be the appropriate mechanism for an ment can be complex to implement and can be very costly. imported shipment that undergoes transformation in an In fact, to be workable, it requires that significant border intermediate country before reaching its final destination. controls remain or that goods be shipped in some sort of Indeed, incentives could emerge for some members to transit and bonding facility all the way to the final destina- collect revenues on imports that are then wholly trans- tion, where duties would be collected.29 Not only would shipped or minimally “transformed� or “repackaged� the logistical costs of running such facilities be substantial, before being exported duty-free elsewhere in the CU. In but they would also tend to diminish some of the expected such cases, burdensome internal border controls, guaran- gains from establishing a CU. For instance, they could dis- tee mechanisms, or even some rules of origin are needed courage the establishment of regional value chains or pro- within the CU to determine what fraction of the collected cessing chains (using imported inputs) or the generation duties should go to which member. This could be an of retail and wholesale services in intermediate locations important issue for small landlocked economies that rely between the initial port of entry and the final destination. on their larger coastal neighbors for transit and that could It is clear from the foregoing discussion that the collection lose revenues as a result of leakages or fraud. When trade and allocation of customs revenue in a customs union setting flows are sufficiently symmetrical, a member’s losses could are not only technical issues, but also (perhaps more impor- be offset by the gains it realizes when goods imported into tant) a question of trust. Good technical coordination and its territory (for which it collects the tariff) are consumed enforcement generally promote trust among CU members. in a neighbor’s.24 Conversely, lack of trust would require more stringent and cumbersome controls on intrabloc transit and stricter appli- cation of the agreed disposition of revenue. This is clearly an Collection of Duties area in which harmonization of border management (cus- At what point should customs duties be collected—at the toms procedures), cooperation, modernization, and capacity initial port of entry into the CU or at the final import building could be very useful (see Keick and Maur 2011; destination? Collecting import duties at the first port of Dawar and Holmes, ch. 16 in this volume). entry into the CU (say, in the coastal member with the more developed port and transit facilities) could be one Conclusions way of ensuring freer movement of goods within the CU and minimizing intra-CU border controls.25 Indeed, if all The conceptual discussions in this chapter demonstrate trade taxes were collected at the point of first entry and that CUs generally require a much greater degree of policy administered or distributed centrally, member countries coordination among members than do FTAs. This is 118 Soamiely Andriamananjara because they require member countries to agree to a are valued, membership in a CU, if played by the rules, common external tariff and to set up institutional mecha- could constitute a straitjacket for some countries. nisms to collect and distribute the tariff revenue. When a In reality, of course, there are numerous cases in which country joins a customs union, it agrees to relinquish a CU member alone negotiates an FTA with a third party. some of its national sovereignty with respect to the formu- Examples of such a situation include the FTAs between the lation and implementation of trade policy. The fact that a EU and South Africa (a member of SACU) and between country is willing to surrender such autonomy over trade the United States and Bahrain (a member of the GCC). policy suggests that it considers this loss to be more than Similarly, Bolivia, Colombia, Ecuador, Peru, and the offset by the economic benefits of securing access to a República Bolivariana de Venezuela form the Andean Pact larger and more harmonized regional market and of (a CU), while Colombia and the República Bolivariana de enhancing the depth and effectiveness of the ongoing Venezuela have joined with Mexico to make up the Group regional integration. The loss of autonomy may also be of Three, an FTA. In some instances, one CU may overlap acceptable to members because in most cases, CUs are another. For example, Lesotho, Namibia, and Swaziland driven by objectives that go beyond trade, such as eco- belong to COMESA while also belonging to SACU, and nomic and monetary unions or even political integration, Tanzania is a member of both the SADC and the EAC. and that require supranational institutions. Multiple and overlapping memberships in regional trade Ceding the control of some aspects of national trade agreements can create difficulties because different groups policy may yield economic benefits, to the extent that it can have conflicting operational or liberalization modali- shelters the trade policy–making process from the influ- ties, and so member countries will have to make different, ence of special interests, at least at the national level. Com- incompatible commitments. This not only could render mitting to a regionally agreed trade policy regime can CUs less effective but could also confuse traders (and even serve as an effective lock-in mechanism for trade reform customs officers) as to which commitments or tariff efforts and can send a strong signal to investors regarding schedules apply to a particular shipment. Unnecessary the predictability of the policy environment.30 In some transaction costs will be created because traders are cases, it may be more practical for a country to delegate obliged to find their way around a number of trade the conduct of trade policy to another (larger) CU mem- regimes with different tariff schedules, different rules of ber or to a supranational agency. In others, a larger mem- origin, and different procedures. ber country can impose its own trade policy or tariff structure on that of the union. In this case, there may be grounds for establishing a regional mechanism, such as a Notes development fund, to compensate other member coun- The author thanks Jean-Christophe Maur and anonymous reviewers for tries for adopting a tariff structure that is not inherently in their helpful comments and suggestions. 1. GATT Article XXIV:8(a) defines a customs union as a single cus- their own economic interests. toms territory substituting for two or more customs territories and having It should be noted that the loss of autonomy implied two characteristics: (a) duties are eliminated on substantially all trade by a CU also covers certain aspects of national trade poli- between the constituent territories, and (b) substantially the same duties are applied by each member to trade with nonmembers. cies, going beyond external tariffs. For instance, whereas 2. Historically, the early academic literature on regional integration FTA members retain full flexibility with regard to future tended to focus on customs unions; classic works are Viner’s The Customs PTA partners, CU members may be limited in their indi- Union Issue (1950) and Meade’s The Theory of Customs Unions (1955). vidual choices for future partners Indeed, membership in 3. Fiorentino, Verdeja, and Toqueboeuf (2007) suggest that CUs are “out of tune with today’s trading climate.� a CU, at least in principle, prevents an individual member 4. In some cases, PTAs are driven by narrow foreign policy or diplo- from acting individually, since any agreement with a third matic considerations. party or any change to the CET needs to be decided by the 5. In practice, tariff liberalization could be accompanied or comple- mented by a rationalization of border management policies. See Keick CU as a whole. It can be argued that CUs could help pre- and Maur (2010) for a thorough survey of the border management issues vent the emergence of a hub-and-spokes trading pattern facing CUs. (see the discussion in Baldwin and Freund, ch. 6 in this 6. Interestingly, although the seminal work by Viner (1950) focused on CUs, the analysis studied the removal of the intrabloc tariff while volume). assuming that other policies (e.g., external tariffs) remained constant. In a world of criss-crossing and overlapping trade Thus, it was more a study of FTAs than of CUs. agreements, the issue of the loss of autonomy can severely 7. In the traditional literature on regionalism, the degrees of eco- constrain members of CUs in using trade agreements as an nomic integration correspond to different stages: preferential trading area, free trade area, customs union, single market, economic and mone- effective commercial instrument—at least in theory. In the tary union, and complete economic integration. These stages were gener- current wave of regionalism, in which flexibility and speed ally thought to be a progression, with political union as the ultimate Customs Unions 119 objective. Nowadays, this progression seems a bit outdated, as many exist- 20. Winters (1996) suggests that Canada’s decision to reduce 1,500 ing FTAs have no ambition to move to the next stage. tariffs on inputs shortly after the initiation of the North American Free 8. This situation effectively reduces the tariff of every FTA member to Trade Agreement was driven by this consideration. that of the lowest (plus the transport cost involved in roundabout 21. This possibility is demonstrated, for instance, by Kennan and importing). Riezman (1990) and by Krugman (1990). 9. Given that rules of origin have to take into account tariffs on 22. This situation is illustrated by the disproportionate influence imported intermediate goods used in the products manufactured within of EU members on policy in sectors in which they claim “vital interests.� the FTA, as well as the extent of value addition, they tend to be complex to Other members strategically and willingly accommodate, in anticipa- administer. Rules of origins also raise other issues; they can, for instance, tion that the favor will be returned when their own “vital interests� create incentives for regional buyers to purchase high-cost inputs from come up. partner countries, thus exacerbating trade diversion, and they can be cap- 23. It should be noted that in the case of the EU, tariff revenues con- tured by interests that favor protection in the form of stringent require- stitute a small portion of government revenue, which is why EU members ments. Robson (1998) argues that even with rules of origin, the problem of are more willing to forgo them and cede them to the commission. trade deflection is not entirely solved, as the low-tariff partner may meet its 24. On a related note, Dawar and Holmes (ch. 16 in this volume) own requirements for a product from the rest of the world and can then argue that CUs provide incentives for stronger regional cooperation in export a corresponding amount of its own production to its partners. the area of trade facilitation. Indeed, there is a need for harmonization of 10. The implementation of a customs union does not necessarily the quality of border enforcement across CU members to ensure that the mean the abolition of border posts. These facilities serve many other weaknesses of one member do not undermine the tax collections of other (trade-related or nontrade-related) purposes, including collecting other members (since imports may be diverted to the border where enforce- taxes, controlling the movement of persons, and enforcing compliance ment is weakest). with domestic standards and health and safety requirements. 25. In SACU, for instance, most customs duties are collected by South 11. The contiguous pattern of CU formation contrasts with the pro- Africa, the dominant economy and trading hub of the bloc. liferation of cross-regional PTAs that characterizes the current wave of 26. In this context, it should be noted, again, that in addition to regionalism. import duties, other taxes may be collected by customs at the border and 12. Although PTA members are not required to change their external that the establishment of a CU does not mean abolition of all border con- tariffs, there may be strong political-economy pressures or incentives for trols. Keick and Maur (2010) discuss the issues related to border manage- adjusting them after the removal of intrabloc tariffs. See Winters (1996) ment in the context of CUs. and Freund and Ornelas (2010) for useful reviews of the literature on this 27. Of course, no existing customs union allocates tariff revenues topic. according to where imports first enter the union. The EU allows member 13. This method is of course a simplification. In many CUs, many states to keep 10 percent of the tariffs they collect. transitional tariffs and not-so-transitional nontariff barriers on intrabloc 28. Conceptually, setting up a CU among countries that have already trade remain. Even the EU only removed the last physical, technical, and established an FTA among themselves does not necessarily require any tax-related obstacles to intraunion trade in 1993, 36 years after the com- changes in the procedure for collecting customs duties. In most cases, munity was founded. The free trade area component of the East African duties are collected on a destination basis before and after the establish- Community (EAC) completely abolished tariffs on trade between ment of a customs union. Tanzania and Uganda and on exports from these two countries to 29. Some customs unions allow for members to establish customs Kenya. Many goods exported from Kenya to Tanzania or Uganda, how- facilities in the territories of other member states in order to collect the ever, continue to be subject to tariffs (Development Network Africa customs that are due to them at the first point of entry. In practice, duties 2007). In Mercosur, there is not yet free internal trade in sugar, and most paid on imports from the rest of the world could still be collected by the favored nation tariffs are applied with a 20 percent preference for intra- country of final destination. zone trade. 30. This argument is generally true for international trade agree- 14. In Mercosur, for instance, the smaller economies (Paraguay and ments, whether bilateral, regional, or global. The need for agreeing on the Uruguay) preferred a CET structure with low tariffs, whereas Argentina CET makes it stronger in the case of customs unions. and Brazil successfully pressed for higher rates. 15. As Hoekman argues in chapter 4 of this volume, real-world PTAs do not select their CETs in a Kemp-Wan manner, nor do they have access References to large compensatory or lump-sum transfers. 16. A poorly designed CET can create divisive tensions among CU Adams, Charles. 1993. For Good and Evil: The Impact of Taxes on the members. Adams (1993) provides an interesting example of how external Course of Civilization. Lanham, MD: Madison Books. tariffs triggered conflicts between the North and the South in the antebel- Bagwell, Kyle, and Robert Staiger. 1999. “Regionalism and Multilateral lum United States. Tariff Cooperation.� In International Trade Policy and the Pacific Rim, 17. Mercosur allows members to temporarily deviate from CET rates ed. John Piggott and Alan Woodland. London: Macmillan. under specific conditions and with some limitations. Following the peso Bohara, Alok K., Kishore Gawande, and Pablo Sanguinetti. 2004. “Trade crisis, Argentina raised tariffs to 35 percent on numerous consumer goods Diversion and Declining Tariffs: Evidence from Mercosur.� Journal of and cut the tariff on capital goods to zero. International Economics 64 (1, October): 65–88. 18. Note that CETs are, in principle, subject to multilateral rules. Development Network Africa. 2007. “Evaluation of an Appropriate GATT Article XXIV:5(a) stipulates that the duties (in this case, the CET) Model for a SADC Customs Union.� Report commissioned by the and other trade regulations imposed on trade of nonmembers shall not, SADC Secretariat, Southern African Development Community, on the whole, be higher or more restrictive than the general incidence of Gaborone. duties and other trade regulations applicable in the participants prior to Fiorentino, Robert, Luis Verdeja, and Christelle Toqueboeuf. 2007. “The the formation of the CU. Under the usual interpretation, this implies Changing Landscape of Regional Trade Agreements: 2006 Update.� that the common tariff of a CU should not be higher than the preunion WTO Discussion Paper 12, World Trade Organization, Geneva. average. Freund, Caroline. 2000. “Different Paths to Free Trade: The Gains from 19. Bohara, Gawande, and Sanguinetti (2004) provide some empiri- Regionalism.� Quarterly Journal of Economics 115 (4): 1317–41. cal evidence for this argument in the context of Argentina’s external tariffs Freund, Caroline, and Emanuel Ornelas. 2010. “Regional Trade Agree- under Mercosur. ments.� Policy Research Working Paper 5314, World Bank, Washington, 120 Soamiely Andriamananjara DC. http://www-wds.worldbank.org/external/default/WDSContent Panagariya, Arvind, and Ronald Findlay. 1996. “A Political Economy Server/IW3P/IB/2010/05/19/000158349_20100519101328/Rendered/ Analysis of Free Trade Areas and Customs Unions.� In The Political PDF/WPS5314.pdf. Economy of Trade Reform: Essays in Honor of Jagdish Bhagwati, ed. Keick, Erich, and Jean-Christophe Maur. 2010. “Regional Integration and Robert C. Feenstra, Gene M. Grossman, and Douglas A. Irwin. Customs Unions.� World Bank, Washington, DC. Cambridge, MA: MIT Press. Kemp, Murray C., and Henry Y. Wan, Jr. 1976. “An Elementary Proposi- Richardson, Martin. 1993. “Endogenous Protection and Trade Diversion.� tion Concerning the Formation of Customs Unions.� Journal of Inter- Journal of International Economics 34 (3–4, May): 309–24. national Economics 6 (1, February): 95–98. Robson, Peter. 1998. The Economics of International Integration. 4th ed. Kennan, John, and Raymond Riezman. 1990. “Optimal Tariff London: Routledge. Equilibria with Customs Unions.� Canadian Journal of Economics Viner, Jacob. 1950. The Customs Union Issue. New York: Carnegie Endow- 23: 70–83. ment for International Peace. Krueger, Anne O. 1995. “Free Trade Agreements versus Customs Unions.� Winters, L. Alan. 1996. “Regionalism versus Multilateralism.� Policy NBER Working Paper 5084, National Bureau of Economic Research, Research Working Paper 1687, World Bank, Washington, DC. Cambridge, MA. http://www-wds.worldbank.org/external/default/WDSContentServer/ Krugman, Paul. 1990. “Is Bilateralism Bad?� In International Trade and IW3P/IB/1996/11/01/000009265_3970311114958/Rendered/PDF/mu Policy, ed. Elhanan Helpman and Assaf Razin. Cambridge, MA: MIT lti0page.pdf. Press. WTO (World Trade Organization). 2003. “Trade Policy Review: Southern Meade, James. 1955. The Theory of Customs Unions. Amsterdam: North- African Customs Union.� Press release, April 25. WTO, Geneva. Holland. http://www.wto.org/english/tratop_e/tpr_e/tp213_e.htm. 6 Preferential trade agreements AND MULTILATERAL LIBERALIZATION Richard Baldwin and Caroline Freund With the Doha Round of trade negotiations ailing, the outcome from high-tariff regional blocs and for large future of multilateral liberalization in the near term looks amounts of trade diversion, the theoretical literature shows bleak. By contrast, preferential trade agreements (PTAs) that incentives to reduce external trade barriers so as to continue to multiply (see Acharya et al., ch. 2 in this vol- limit costly diversion are likely to be present.2 The nascent ume), making regionalism the most active mode of trade empirical literature is tackling the question of how trade liberalization. The regionalization of trade is of serious liberalization has been affected by the formation of PTAs. concern to many international economists who view mul- Although the verdict is not yet in, the evidence indicates tilateralism as far superior to regionalism for improving that regionalism is broadly liberalizing. welfare. At issue is the preferential nature of regional agree- This chapter summarizes the available theoretical and ments, which could divert trade and reduce the potential empirical evidence on the relationship between regionalism for future multilateral liberalization. and multilateralism, with the aim of discerning whether the The multilateralists argue that, although regionalism may spread of regionalism is likely to be a threat to, or an oppor- increase trade, its effects on welfare and on the world trade tunity for, broader trade liberalization. The next section system are likely to be harmful. There are two main con- identifies the distortions that generate a need for regional cerns.1 The first is trade diversion: preferential trade agree- and multilateral trade agreements. There follows an ments, by diverting trade away from the most efficient global overview of the available theoretical work on whether producers in favor of regional partners, may prove welfare regionalism constitutes a stumbling block or a building reducing. The second concern, which is of greater impor- block on the path to trade liberalization. The effect of tance, is that regionalism may hinder multilateralism, lead- regionalism on world welfare is then examined, the empiri- ing to a bad equilibrium in which several regional trade cal literature is surveyed, and conclusions are drawn. blocs maintain high external trade barriers. Regionalism can also undermine multilateralism simply by diverting limited Reciprocal Trade Agreements government resources from multilateral negotiations. These two concerns are related: in a highly regionalized If there is one thing economists agree on, it is that free world, there is likely to be significant trade diversion and trade is best. Why, then, do we need trade agreements to hence lower welfare. Still, this feature of the bad equilibrium lower tariffs? In fact, although global free trade may be makes it less likely in practice. It is precisely because the good for world welfare, countries nevertheless have reasons trade diversion is costly to bloc members that there is an to maintain tariffs. There ensues a prisoner’s dilemma: incentive to reduce external tariffs. As tariffs fall, trade each country may be unilaterally better off with a tariff, but diversion disappears, and regionalism becomes a force for jointly they are worse off. Cooperation through a trade general liberalization. Thus, despite the potential for a grim agreement is necessary to liberalize trade. 121 122 Richard Baldwin and Caroline Freund Countries have tariffs for three main reasons. First, reciprocal tariff cutting. Moreover, once the agreement is terms-of-trade considerations may induce a country to use in place, the reciprocity can help lock in the reform—can tariffs to drive down the price of its imports relative to the tie the government’s hands, so to speak—in a way that price of its exports, thus raising welfare for the nation as a helps governments resist special-interest pressures for whole. For example, suppose a large country exports wine protection. Even if domestic consumers do not object to and imports cheese. By placing a tariff on imported cheese, new protection, the export interests of the foreign trade the country increases domestic demand for wine relative to partner will. cheese, and if the country is large enough, this practice The third reason for tariffs, revenue generation, is raises the world price of wine relative to cheese. If both the slightly different, as the government faces a loss of income importing and the exporting country impose tariffs, how- with no offsetting added source of revenues if tariffs are ever, the relative world prices of wine and cheese remain reduced. Trade agreements with the World Trade Organi- roughly unchanged, leaving consumers in both countries zation (WTO), or bilateral agreements with industrial worse off because domestic prices are distorted. Con- countries, therefore often include provisions to help coun- sumers in both countries are consuming more of the tries adjust to revenue loss. domestic good than is ideal (given world prices), terms-of- Both the WTO and PTAs allow countries to cooperate trade gains do not materialize, and world welfare is and commit to reducing trade barriers, but there is an reduced. important distinction between the two types of agreements. Second, there may be political constraints. Tariffs can be In particular, PTAs go against the principle of nondiscrimi- used as an internal redistribution tool by governments— nation that is at the heart of the multilateral system. The typically to shift income to some favored industry, region, first article of the General Agreement on Tariffs and Trade or group of voters or political contributors. For example, (GATT) concerns nondiscrimination, or most favored many economies tend to protect steel because production is nation (MFN) treatment. MFN means that every time a commonly concentrated regionally and the workers and country lowers a tariff, it has to offer the same treatment to firm owners tend to be politically organized. This combina- all its trading partners. The purpose of this clause is to pre- tion of political organization and geographic concentration vent trade diversion and the cumbersome tariff structure often means that the sector’s concerns may receive much that would likely prevail in the absence of MFN. greater weight in political decisions than would be sug- By offering preferences to specific countries, PTAs gested by its share of employment or output. Similarly, cer- definitively violate MFN treatment. Yet they are allowed, in tain industries may receive special treatment if the owners part, because the MFN clause creates a free-rider problem. of capital are connected to the government. Protection MFN means that countries in the WTO must offer the reduces consumer welfare, since consumers are forced to same tariff to all members. But if only a subset of members pay more than the world price, but governments cannot agrees on significant tariff reduction, other members can credibly commit to withholding protection of the special “free ride�—they get expanded market access without new sectors.3 In addition, protection may be in place because of commitments. If all members participated equally in MFN historical reasons and can be politically difficult to remove tariff reduction, PTAs might not be needed. PTAs enable because industry has adapted to it. countries that want to pursue deeper trade liberalization to Third, countries, especially developing countries, may evade the free-rider problem. Snape (1993) discusses the rely on tariffs for a share of the tax revenue they need to history of GATT Article XXIV, which allows for an excep- fund general government expenditures. Border transac- tion to nondiscrimination, and argues that this article is tions are an easy way to collect revenue when income tax vital for maintaining the multilateral club because some systems are weak. This motive tends to be more important members might opt out if it were not included. (In partic- in developing countries, where other forms of revenue col- ular, the United Kingdom, which was the largest importer lection are difficult. at the time, insisted on keeping its Commonwealth prefer- Reciprocal liberalization helps neutralize the first two ences.) Summers (1991) asserts that all types of liberaliza- forces described above: terms-of-trade motivations are tion, whether unilateral, bilateral, or multilateral, are very neutralized, and joint welfare is higher because, with lower likely to be good and that regionalism gives governments a tariffs, consumption patterns are less distorted. Reciprocity way to maintain progress on liberalization. in tariff cutting makes it easier to form a proliberalization This section has shown that reciprocal liberalization is coalition because the political strength of the domestic usually necessary for removal of barriers to international losses from tariff cutting can be balanced against the polit- trade. Although multilateral liberalization is preferable, ical strength of the exporters that gain from foreigners’ regional liberalization is ubiquitous and cannot be Preferential Trade Agreements and Multilateral Liberalization 123 ignored. The key question is whether this advancement of countries, most notably members of the Southern Cone regionalism is broadly beneficial or detrimental to global Common Market (Mercosur, Mercado Común del Sur).6 trade and welfare (see box 6.1). In the next section, we dis- The rise of North American regionalism coincided with cuss the recent rise of regional liberalization and how, in two other major developments in the world trade system. theory, it may affect the multilateral trade system.4 First, GATT negotiations lurched from crisis to crisis in the late 1980s and then seemed to die with the acrimonious collapse of the Uruguay Round’s “final� summit in Decem- Stumbling Blocks and Building Blocks ber 1990. Second, European regionalism was reignited by From 1960 to the late 1980s, regionalism was a simple the Single European Act and the collapse of the Soviet matter. It was represented by the European Economic Union. Community (EEC), which encompassed a third of world Many respected thinkers looked at this temporal corre- trade in a highly effective customs union, and by a slew of lation and saw causality. They feared that the spread of PTAs among developing nations that covered a trivial regionalism might kill the world trade system. These fears fraction of world trade and in any case never operated are easy to understand. Two-thirds of world imports went effectively. The systemic implications of regionalism were to North America and Europe; 40 percent of this total was simply not an issue. intrabloc trade that was soon to be covered by discrimina- Regionalism became complicated in the late 1980s tory liberalization schemes. Still more worrisome, North when Canada and Mexico changed their minds about American and European countries were the stalwarts of the regionalism (Krugman 1991b, 7).5 The United States had GATT system. If regionalism weakened their support of long been interested in regional preferential trade, but multilateralism, the GATT was indeed in deep trouble. Canada and Mexico had resisted, fearing domination by Spreading regionalism had become much more than a their giant neighbor. In 1985, Canada proposed a free small-think “should I join?� question. trade agreement (FTA) with the United States that These fears promoted regionalism to a prominence on entered into force in 1989. In 1990, Mexico, too, proposed the world’s policy agenda that it had not enjoyed since the an FTA with the United States. This initiative evolved into 1950s. The shift naturally attracted paradigm-setting efforts the North American Free Trade Agreement (NAFTA) at by the profession’s leading international economists. the insistence of Canada, which wished to safeguard its Krugman (1991b) is clearest in rejecting the relevance Auto Pact arrangement with the United States for tariff- of the 1950s small-think approach, which focused on static free trade in automobiles and parts. The U.S.–Mexico ini- welfare issues and delineating the outlines of a new line of tiative triggered a wave of Latin American requests for inquiry—what we call big-think regionalism: bilateral trade agreements with the United States and gave In a fundamental sense, the issue of the desirability of free greater urgency to arrangements among Latin American trade areas is a question of political economy rather than of economics proper. While one could argue against the formation of free trade areas purely on the grounds that Box 6.1. Is Bilateralism Bad? they might produce trade diversion . . . the real objection is a political judgment: fear that regional deals will under- Paul Krugman, in a series of papers published in 1991, mine the delicate balance of interests that supports the reframed the 1950s national welfare question as a global- GATT. (Krugman 1991b, 14–15) level question. Krugman (1991a) introduced a new approach by asking whether the spread of regionalism raises or lowers world welfare. This discussion spawned a Krugman’s framing of what he identified as the key decade-long literature and continues to influence research issue—the impact of regionalism on support for the GATT even today. The “is bilateralism bad?� literature—also system—did not catch on, however.7 known as multilateralism versus regionalism literature— looks distinctly odd from today’s perspective in that it tries The focus of this part of the discussion is on what we to use simple theory to answer what is intrinsically a consider to be the central theoretical question: does complicated empirical question. At the time, however, it regionalism help or hinder multilateralism? Ultimately, was the best that economists could do; they had limited access to the necessary data and lacked the panel this question is also empirical, but given the relative econometric techniques to exploit them. Moreover, paucity of experience with the regionalism-multilateralism spreading regionalism was at the time more of a threat interface (only one multilateral trade negotiation has been than a reality, so there was little experience to support empirical tests. Baldwin (2009) offers a full discussion of completed since 1991), convincing empirics is at an early the issue. stage, with some tantalizing results just beginning to emerge (see “Welfare Consequences of PTAs,� later in this 124 Richard Baldwin and Caroline Freund chapter). Moreover, given the complexity of the interlink- are put forward in the help-or-hinder literature (for exam- ages, a clear theoretical understanding is a necessary condi- ple, PTAs may slow the achievement of global free trade), tion for well-structured empirical work. but the desire for clarity has led the profession to focus on the strong form. In our opinion, only three forms of stumbling block Are Regionalism and Multilateralism Friends or Foes? logic are currently relevant to real-world policy analysis: Bhagwati, in The World Trading System at Risk (1991), does the preference-erosion (exploitation), goodies-bag, and not focus on regionalism. In the first part of the book, “The cherry-picking variants. (Many more will undoubtedly be GATT Architecture under Threat,� he lists regionalism as illuminated in coming years.) For simplicity’s sake, these one of four main threats. Nevertheless, his writing has possibilities are demonstrated under the naïve but trans- helped establish big-think regionalism as the new para- parent assumption that national governments choose tar- digm. In the first paragraph of his chapter on regionalism, iffs to maximize national welfare. he writes, “These regional alignments have led to fears of Preference erosion, or exploitation. Starting from a world fragmentation of the world economy into trading blocs in in which all nations have MFN tariffs, the question is antithesis to GATT-wide multilateral free trade. Does such whether some group of nations can raise its collective wel- regionalism truly constitute a threat to multilateralism?� fare above the free trade level by forming a trade bloc and (Bhagwati 1991, 58). Although he does not set out an ana- thus exploiting other nations. If the answer is “yes,� that lytical framework for answering the question, his writing bloc is a stumbling block on the road to multilateral free influenced the intellectual paradigm for more than a trade because the members would veto global free trade as decade. undermining their exploitation of third nations. Theory requires explicit questions. Asking whether In trade models, the answer is almost always “yes,� but regionalism and multilateralism are friends or foes is not the answer may depend on the level of MFN tariffs when sufficient. Pure logic identifies three mutually compatible the bloc is formed.8 Given Smith’s certitude (briefly, that all ways that regionalism and multilateralism could interact: parties to a preferential trade arrangement benefit) and Haberler’s spillover (postulating that third nations must • Regionalism could affect multilateralism. lose from such an arrangement), some combination of • Multilateralism could affect regionalism. nations is bound to be better able to exploit third nations • Both multilateralism and regionalism could be driven by acting as a bloc. This is almost trivially true if the bloc by third factors. can violate its WTO tariff bindings by raising external tar- iffs. After all, the bloc as a whole has more buying power The literature has looked at all three ways of framing than its constituents do individually, so it can better exploit the issue, but the first has dominated since Krugman foreigners. Less obvious, but equally true, is that stumbling (1991b, 1996) presented a simple analytical framework for blocks can be found even when external tariffs are main- posing the question. His explicit question was, How does tained (as has been the case for all the major postwar an exogenous variation in regionalism (specifically, the PTAs). formation of a new PTA) affect nations’ incentives to cut Consider the simple model presented in chapter 3 of tariffs multilaterally? Although most of the literature has this volume, with three symmetrical countries (Home, followed Krugman’s lead in asking how exogenous varia- Partner, and Rest of the World, or RoW) and three goods. tions in regionalism affect multilateralism, authors have Home imports Good 1 and exports Good 2 to Partner; also discussed the specific effect of PTAs on multilateral Partner imports Good 2 and exports Good 1 to Home; and trade negotiations (MTNs); the deeper forces driving Rest of the World exports Goods 1 and 2 to Home and PTAs and MTNs (Summers’ notion that all the “isms� are Partner, respectively, and imports Good 3 from both (see good); and the effect of multilateralism on regionalism. figure 3.1, in ch. 3). The first welfare theorem tells us that These are discussed below. global free trade is efficiency enhancing (a move to the first best), and symmetry ensures that each nation receives an equal slice of the gains. Three Kinds of Stumbling Block Logic This conclusion, however, can be reversed when we start In its cleanest form, the stumbling block logic asserts that if from the situation in which Home and Partner have the stumbling block PTAs were forbidden, global free trade formed a PTA by eliminating tariffs among them. Taking would be achieved, but that since they are permitted, global the PTA as the base case, a move to global free trade elimi- free trade becomes impossible. Weaker forms of this thesis nates the preference margin (P' – PFT) that Home exporters Preferential Trade Agreements and Multilateral Liberalization 125 Figure 6.1. Net Welfare Effects, Preferential Trade Figure 6.2. Relationship between MFN Tariffs and Home Agreement to Global Free Trade Welfare border price internal price home welfare XS P global free trade C2' A' 0.60 P' C1' C1' PFT D' 0.59 B' P–T P' – T MD 0.58 0.57 stumbling block tariffs MFN exports XR' imports 0.56 PTA Source: Author’s elaboration. 0 0.05 0.10 0.15 0.20 0.25 MFN tariff tariff (T) (MFN T) enjoyed in Partner, thereby leading to a terms-of-trade loss Source: Author’s elaboration. Note: Home welfare is plotted against the most favored nation (MFN) tariff of area C1' and a trade-volume loss of C2' (figure 6.1, left- for three trade regimes: MFN (all nations impose the same MFN tariff T on hand panel). all goods); PTA (Home and Partner form a preferential trade agreement, On the import side, because of lower internal prices, PTA); and global free trade, which is the MFN regime with T = 0. The parameters chosen are a = 1, b = 1/2; the qualitative results are unaltered from P' to PFT, global free trade would win Home an addi- for other choices of parameters. Note that T = 1/4 is the prohibitive tariff tional trade volume gain of area A', a terms-of-trade with a = 1, b = 1/2. In ad valorem terms, the Viner crossing occurs at a tariff of about 27 percent of the free trade price, and the ad valorem tariff improvement with respect to Partner exporters of area C1', threshold for a stumbling block occurs at about 9 percent; the prohibitive and a terms-of-trade loss on imports from RoW, shown as tariff is about 42 percent for these parameter values. area B' (right-hand panel). Global free trade would also improve Home exporters’ market access to RoW, and this would boost Home welfare by area D'. Overall, the net wel- forms of exploitation, but in exchange, it provides better fare change of moving from the PTA to global free trade access to the RoW market and more liberalization in is – C2' – B' + D' + A'; the sign of this is ambiguous. Home’s import market. It is straightforward to show that, in a simple model When the initial tariff T is low, the market access and with linear demand, a low MFN tariff and a PTA with zero home liberalization gains are modest, and so the net effect internal tariffs generate gains for the trade partners that is negative. In other words, the basic logic of the stumbling exceed those from free trade. (This is shown in figure 3.2, block result turns on the way that a PTA allows the PTA in ch. 3; see Baldwin 2009, box 4, for further details of the partners to exploit excluded nations. model.) Figure 6.2 presents the Home welfare levels for dif- The model presented here is very special, but the heart ferent values of the MFN tariff. and soul of the stumbling block effect—the exploitation of For initial MFN tariff levels that are sufficiently low, we excluded nations—is a general result, and one that is surely see that Home’s welfare is higher with the PTA than it is an important consideration in the real world. with global free trade, even though Home would have The opposition of small developing nations—especially agreed to global free trade starting from the initial situa- those that benefit from European Union (EU) unilateral tion without the PTA. (The line marked MFN is com- preferences—to agricultural liberalization in the WTO pletely below the line marked “global free trade,� but the Doha negotiations is a classic example of the preference- PTA line is above the global free trade line for sufficiently erosion stumbling block. Had the EU not unilaterally low tariffs.) granted preferences to these nations, they would probably Intuitively, the PTA allows Home to exploit RoW both have been pushing for opening of the EU market in sugar on the import side (by pushing down the price it pays RoW and other goods. exporters of Good 1) and on the export side (by raising Goodies bag. The goodies-bag version follows closely the the price in Partner at the expense of RoW exporters of fundamental economic logic of the preference-erosion Good 2). The move to global free trade undoes these two stumbling block.9 Briefly, the rents corresponding to 126 Richard Baldwin and Caroline Freund Smith’s certitude can be thought of as a set of “goodies� and only a modest amount of additional variety or com- that can be used by one or both PTA parties to buy parative advantage gains. Depending on parameters, espe- noneconomic benefits from its partners. Because the size cially the political power of the sufferers, the gains may of noneconomic benefits that can be “purchased� is linked not be sufficient to make global free trade attractive to the to the richness of the goodies bag—that is, the margin of bloc members. preferences—PTA members have an extra incentive to maintain high margins of preference by avoiding multilat- Building Block Logic eral liberalization. The goodies-bag logic, however, extends to a far greater range of issues than the tariffs that Whereas many trade policy scholars, including Krugman are the focus of the preference-erosion stumbling block. In (1991b) and Bhagwati (1991), worried that regionalism the case of a PTA between very large and very small was a stumbling block to global free trade, others, such as nations—a case that is extremely common in the new cen- Summers (1991) and Bergsten (1991), viewed regionalism tury (e.g., the United States and Costa Rica; Japan and as a largely benign or even constructive force in the world Singapore)—the large country’s interest in the PTA can trade system. hardly be thought to be preferential market access. Here, we consider the economic logic of the assertion The EU, for example, grants extensive preferences to its that PTAs can foster multilateral liberalization. There are members’ former colonies by using the justification of four main arguments in the literature. We begin with the international solidarity. In other words, the economic gains one that permeates the rationales used by countries that to the EU’s partners count as a plus inside the EU because simultaneously pursue regional and multilateral liberaliza- they advance one of the EU’s noneconomic objectives— tion: the notion that preferential liberalization creates a fostering development. Similarly, but more explicitly, the political-economy momentum that makes multilateral lib- United States justifies many of its PTAs with small, poor eralization easier (and vice versa). nations on the basis of noneconomic objectives, typically, Juggernaut. According to the juggernaut building block antidrug or antiterrorism policies. logic, liberalization begets liberalization. The logic comes The earlier discussion illustrated how the desire to safe- in two parts that are most easily explained in the context guard rents created by a PTA could make a nation reject of multilateral liberalization. When the GATT began in global free trade when it would have embraced MFN free 1947, import duties were high worldwide, since they had trade without the PTA. The goodies-bag stumbling block been set without international coordination during the logic amplifies this mechanism by making both nations tariff wars named for the U.S. Smoot-Hawley Act. The tar- interested in each other’s export rents—the area correspon- iffs balanced the supply of and demand for protection in ding to C1' in figure 6.1, with the link operating through the the “political market� of each nation separately. The main pursuit of noneconomic (in the narrow sense) objectives. demandeurs of import protection were import-competing Cherry picking. An entirely distinct mechanism is at firms and the workers they employed. Government was work in the cherry-picking stumbling block.10 Assume the supplier of protection, but concern for the country’s that the trading environment is marked by both intrain- general economic well-being meant that the government dustry trade in differentiated products and interindustry would set the protection level below what was lobbied for trade. In this world, trade liberalization will produce gains by special interests (the supply of protection being not from trade because of the variety effect in the differenti- perfectly elastic). ated product sectors (as in Helpman-Krugman models) Starting from this situation of uncoordinated tariff and because of comparative advantage effects. The com- setting, announcement of an MTN based on the princi- parative advantage gains, however, come bundled with ple of reciprocity alters the array of political forces inside politically difficult effects on domestic factor prices, which each participating nation. The central point is reciproc- will be lower (e.g., lower wages).11 Now suppose that two ity, which converts each nation’s exporters from large nations have similar factor endowments. If they form bystanders in the tariff debate to antiprotectionists. For a trade bloc, they will win a large share of the variety gains exporters, lobbying against domestic tariffs becomes a that would come with global free trade, and because of means of lowering foreign tariffs. Because the MTN their similarity, they will experience little pain from lower rearranges the political-economy forces inside each factor prices. nation, all governments find it politically optimal to Taking the trade bloc as the base case, the bloc mem- choose tariff levels that are lower than the unilaterally bers may find a move to global free trade unattractive. It optimal tariffs.12 This is the first part of the juggernaut would entail a good deal of pain in terms of factor prices theory.13 The logic is not new. Preferential Trade Agreements and Multilateral Liberalization 127 Informed observers have long known that the GATT’s Figure 6.3. Juggernaut Logic reciprocal MTNs mostly had to do with helping nations tariff (T ) internalize a political-economy externality inside their own polities, making it easier for national politicians to put FE together a national coalition in support of freer trade. Eo GFOCunil Writers such as Robert Baldwin (1970, 1985) and Destler To (1986) are explicit on this point, but historical accounts of the Cobden-Chevalier Treaties show that the use of exter- nal trade deals to realign domestic political forces was very GFOCMFN much in the minds of nineteenth-century thinkers (Irwin 1996, 96). Even Krugman writes that “the process of multi- lateral negotiation . . . sets each country’s exporting inter- ests as a counterweight to import-competing interests; as trade negotiators bargain for access to each others’ mar- Efinal size of import- kets, they move toward free trade despite their disregard for competing the gains from trade as economists understand them� sector, n (Krugman 1991b, 17). Source: Author’s elaboration of figure 4 from Baldwin 2006. The second part of the juggernaut logic concerns the Note: FE, free entry; GFOC, government first-order condition; MFN, most favored nation. effects of the tariff cuts on openness. The cuts make all nations more open: export sectors expand with the foreign tariff cuts, and import-competing sectors contract with number of firms), via politics. The politically optimal domestic tariff cuts. Assuming that political influence is tariff choice, which takes as given the size of the import- linked to industry size, this economic relandscaping competing sector, is plotted as GFOCunil, the solution to strengthens proliberalization forces and weakens antiliber- the government’s first-order condition without MTN. The alization forces in all nations—although, of course, such politically optimal tariff rises with n because the larger the industrial restructuring takes years. In other words, the ini- import-competing sector, the higher the political benefit tial reciprocal tariff cuts start a liberalization juggernaut from a marginal increase in the tariff. The free entry curve, rolling. Because of the economic relandscaping that occurs FE, relates the equilibrium number of firms to the tariff. As during the phase-in of the initial tariff cuts, all govern- the tariff rises, more firms find it optimal to enter the mar- ments find that their politically optimal tariff in the next ket. These two relationships assume that the government MTN is below the levels that they found politically optimal and firms are shortsighted and that the government chooses during the previous MTN. These fresh tariff cuts continue T, taking n as given, while firms choose n, taking T as given. the relandscaping, and the juggernaut continues to roll for- Note that the figure only captures the size of anti-trade ward. Once the liberalization juggernaut starts rolling, it forces; the size of the pro-trade export sector is suppressed crushes all tariffs in its path.14 to avoid the need for a three-dimensional diagram. To the extent that regionalism can start the juggernaut To see the two steps of the juggernaut effect, note that rolling, PTAs can serve as building blocks.15 The precise announcement of the MTN shifts the GFOC curve to mechanism is a simple extension of the juggernaut logic. GFOCMTN. This curve is lower because the government finds PTAs reconfigure members’ economies, making export it politically optimal to set a lower tariff for any given level of sectors larger and import-competing sectors smaller. Thus, n when domestic protection is linked to foreign protection the PTA can alter the member governments’ stance toward via reciprocity. As drawn, the new long-run equilibrium MFNs, making it politically optimal to cut MFN tariffs to Efinal entails free trade, but since entry and exit occur slowly, levels that would not have been politically optimal without the tariff and the state of the import-competing industry do the PTA. Of course, if a PTA results in higher external tar- not jump to Efinal. The figure illustrates one possible adjust- iffs (as in the case of the EU’s agriculture tariffs), then it ment path. Each MTN results in an instantaneous drop in can start the juggernaut rolling backward. the tariff, but slow entry and exit mean sluggish movement The basic idea is presented in figure 6.3. The two curves, of the state variable, as indicated by the horizontal arrows. FE (free entry) and GFOC (government first-order condi- The juggernaut effect acts as a building block if the PTA tion), show how the size of the import-competing sector reduces the importance of import-competing industry in depends on the tariff (free entry) and how that tariff governments’ objective functions. In many models, a PTA depends on the size of the import-competing sector (n, for between Home and Partner does reduce the size of the 128 Richard Baldwin and Caroline Freund Home import-competing sector, and then PTAs would be bundles. The nature of the proposed trade deal can affect a building blocks on the road to global free trade. nation’s ranking of choices (unilateral versus reciprocal, for The FE curve in figure 6.3 was drawn for symmetric example). These rankings, unlike the rankings of a standard MFN tariffs. When the PTA is signed, the FE curve rotates consumer, are path dependent because historical liberaliza- inward, as shown in figure 6.4, because the additional com- tion can affect the current political strength of various pro- petition from Partner producers lowers the Home price trade and anti-trade special interest groups. A number of facing import-competing firms, and some of them exit. building block arguments assume, however, that nations, like Consequently, the GFOC under reciprocal trade will yield a consumers, have exogenous preferences about outcomes. We lower MFN tariff after a PTA (point E2) than before (point next consider the easiest of these arguments, which assumes a E1). Of course, if the PTA somehow increases protection of representative consumer and a government that acts to max- the Home import-competing sector, the effect is reversed, imize the individual’s well-being. and the PTA acts as a stumbling block.16 Kemp-Wan theorem. The assertion that trade blocs may Frankel and Wei momentum. Frankel and Wei (1998) be building blocks in a static world is as easy and as general illustrate another juggernaut-like mechanism. In their as the assertion that they may be stumbling blocks. Starting model, imperfect information makes workers uncertain as from a world in which all nations have MFN tariffs, the to whether they will win or lose from global free trade. Since question is: can some group of nations always raise its col- a PTA is an intermediate form of liberalization, the authors lective welfare by forming a trade bloc? If the answer is show that a PTA could be politically feasible even when “yes,� then a piecemeal enlargement of the bloc will raise global free trade would not be. After the PTA is signed, the bloc members’ well-being monotonically. Bloc members nation’s true comparative advantage is revealed, and workers attain the highest welfare when all nations are part of the now know whether they will win or lose from free trade. If bloc. In this world, the formation of a single bloc should the parameters are chosen carefully, the certainty resolution trigger a domino effect that leads to worldwide free trade. may mean that global free trade is politically feasible only As seen in chapter 3, the Kemp-Wan theorem tells us after the PTA. Thus, the PTA is a building block, and since it that the answer to the above question is always “yes� when operates by altering the political-economy landscape, it nations have access to international lump-sum transfers can be thought of as a momentum-generating mechanism. (Kemp-Wan 1976) or to a complete set of commodity taxes The juggernaut logic exploits the fact that nations do have and subsidies (Dixit-Norman 1980). Kemp and Wan preferences about trade arrangements, in the way that indi- (1976), in probably the first formal contribution to the vidual consumers have preferences about consumption building block–stumbling block discussion, make exactly this point. (See Aghion, Antràs, and Helpman 2007 for an elaboration of the Kemp-Wan argument that uses modern Figure 6.4. Juggernaut Building Block Logic cooperative game theory concepts.)17 tariff (T ) Although the Kemp-Wan building block logic is flawless, FE (with PTA) it falls down in the face of the real-world problem that FE nations do not have access to massive lump-sum transfers. Indeed, the assumption that such international transfers are Eo GFOCunil a realistic possibility basically assumes away most of the To core difficulties facing the international trade system (and international relations more broadly). Without interna- GFOCMFN tional transfers, the logic of preference-erosion stumbling E1 blocks and cherry-picking stumbling blocks suggests that in many blocs, some members would eventually veto some E2 enlargements. Veto avoidance. The preference-erosion stumbling block logic discussed above rests on the fact that bloc members can veto the move to global free trade. The veto-avoidance size of import- building block logic points out that, although bloc mem- competing sector, n bers can veto multilateral trade liberalization, they cannot veto further PTAs that may eventually eliminate all tariffs Source: Author’s elaboration of figure 4 from Baldwin 2006. Note: FE, free entry; GFOC, government first-order condition; PTA, globally. The explosion in the number of PTAs among preferential trade agreement. small nations witnessed in the new century may very well Preferential Trade Agreements and Multilateral Liberalization 129 be attributable to a combination of the juggernaut effect world. Customs unions require supranational decision- and veto-avoidance logic.18 making capacity to keep all external tariffs in line in the We start by considering the development of an extremely face of changes in antidumping duties; special unilateral common form of regionalism, hub-and-spoke PTAs. Here, preferences to third nations, such as the generalized system one partner (e.g., the United States, the EU, or India) has a of preferences; and tariff changes in multilateral trade network of radial bilateral PTAs with some of its trading talks. The groups of nations that manage such coordina- partners, but these trading partners do not have PTAs with tion are of just two types: the EU, and nations involved each other. Simplicity dictates our continued use of the in superhegemon relations (e.g., France and Monaco; symmetric framework shown in figure 3.2 in chapter 3, so Switzerland and Liechtenstein; and the South African Cus- we arbitrarily bestow hub status on Home. toms Union). Given that the real world is covered with Roughly speaking, Home found the bilateral PTA with hub-and-spoke trade arrangements, we assume henceforth Partner attractive because the improved market access for that Partner has no veto over Home’s PTA policy, and the Home exporters in Partner’s market more than out- hub-and-spoke system is indeed set up. weighed the potential welfare losses from trade diversion in The story, however, is not finished. As it turns out, the Home’s import market. This suggests that Home might two spokes may find a spoke-spoke PTA to be advanta- also find a second bilateral PTA with RoW to be welfare geous, and this would achieve global duty-free trade. enhancing.19 (Trade, however, would not necessarily be free because of As it turns out, in the PTA diagram framework, Home the exclusion of various sensitive sectors, rules of origin, always gains from signing a second PTA with RoW. Intu- and cumulation.) itively, the point is that Home gains the same preferential The hub-and-spoke PTA puts Home in an enviable market access as it did from the first PTA, and it undoes the position, giving it the benefits of free trade for its imports potentially harmful trade diversion by fully liberalizing its and preferential market access for all of its exporters. In import market. To see this in more detail, we reinterpret this sense, hub-and-spoke bilateralism might be thought of figure 6.1. On the export side, Home’s second PTA wins its as another example of the preference-erosion stumbling preferential access to RoW’s market without giving up its block logic: Home would veto WTO talks aimed at achiev- preferences in Partner; this has a net welfare value of areas ing global free trade. This simple world, however, can D' + C1' + C2' in the left-hand panel.20 On the import side, attain global duty-free trade without multilateral talks—a the second PTA brings the price in Home’s market for PTA between Partner and RoW would do the job. So, Good 1 to the global free trade level, PFT. The welfare would Partner and RoW be interested in a PTA? impact of this is the positive trade volume effect area A' Taking hub-and-spoke bilateralism as the point of plus the conflicting terms-of-trade effects, areas –B' and + departure, the spokes (in this case, Partner and RoW) C1'. (See Baldwin 2009, box 5, for a mathematical proof.) clearly have a very different view of global free trade than Would RoW accept Home’s offer of a second PTA? As it does the hub. A move to global free trade would do nothing turns out, RoW gains from such a PTA as long as T is not to erode Partner’s preferences in Home, since those were too high (see Baldwin 2009 for details). That is, the hub- already eroded by Home’s second PTA. For Partner, the and-spoke situation is better for RoW than the initially dis- shift to the global free trade regime would involve a stan- advantaged position, when it was excluded from the dard exchange of market access with RoW; Partner would Home-Partner PTA. On the export side, a PTA with Home see its export price to RoW rise from P' – T to PFT for would improve RoW’s market access a great deal (its Good 3, and RoW would see a symmetric border price rise export price would rise from the depressed level of P'– T to for its exports of Good 2 to Partner (see figure 3.2) The the free trade price, PFT). The liberalization on the import attendant liberalization of the two nations’ import mar- side would have the usual positive trade volume effect and kets would have the usual conflicting trade volume and conflicting terms-of-trade effects, identical to those experi- terms-of-trade effects, but overall, the two nations could enced by Home in its first PTA. find the exchange to be welfare enhancing. In fact, Partner Plainly, Partner will be harmed by the formation of the and RoW would always prefer global free trade to the hub- hub-and-spoke system around Home; its preferences in and-spoke situation. Home are eroded, and it receives nothing in compensation. This is certainly not to be taken as a general result. It Partner would thus like to veto Home’s second PTA, but does, however, illustrate how regionalism could be a build- except in extraordinary circumstances, third nations can- ing block in a world in which overall free trade would be in not veto PTAs. The main exception is customs unions, but the interest of all nations, but achievement of the goal is functioning customs unions are quite rare in the modern blocked by nations that fear erosion of their preferences. 130 Richard Baldwin and Caroline Freund Related Logics: Induced Liberalization and Protection slopes of the third-nation import supply curves, Tod is likely to fall because PTAs typically reduce PTA members’ Before ending this review of the helps-or-hinders literature, trade with third nations (Harberler’s spillover); that is, Mod it should be useful to cover two economic mechanisms that is likely to fall. For example, in the simple PTA diagram link PTAs and MFN tariffs without formally making the model presented in figure 3.2 in ch. 3, the import supply connection with multilateral trade talks. Both consider the curves are linear, and so dp*/dM does not change with impact of PTAs on a nation’s MFN stance in the absence of preferential liberalization. However, Haberler’s spillover a new MTN. The first mechanism links PTAs to unilateral effect lowers third-nation trade (Mod), and so Home’s opti- MFN liberalization; the second looks at how a PTA can mal tariff on RoW exports falls. lower or raise a nation’s effective MFN tariff rate. Another mechanism that yields complementarity turns Unilateral liberalization. The building block logics on the general principle that taxes become more distor- examined earlier directly address the issue of whether PTAs tionary when the variance of rates across products help or hinder the attainment of global free trade. Here, we increases. This is termed the uniform tax rate principle, look at a related but logically distinct question: what is the and it is a feature of many economic models, especially impact of a PTA on the tariffs a nation would find unilater- when administrative and enforcement considerations are ally optimal to impose on third nations? Intuitively, the taken into account. It explains why most nations impose question is whether preferential tariffs are complements to fairly even indirect tax rates across products. Since the PTA or substitutes for MFN tariffs.21 The easiest way to organize automatically makes the import tax structure more the various mechanisms is to start from the Meade (1955) uneven, there is some presumption that the PTA makes formula for the welfare impact of any trade policy change the third-nation tariffs more distortionary. In models in a Walrasian economy: where this is true, nations are likely to lower third-nation Net home welfare effect = (p – p*)dM – (M)dp* (6.1) tariffs when they reoptimize their trade tax structures. That where p and p* are the vectors of internal and border prices is, PTAs encourage nations to reduce applied MFN tariffs. (and p – p* = T, the tariff), M is the vector of bilateral Preferential and MFN tariffs as substitutes. The most obvi- imports (exports are negative imports), and dM and dp* ous mechanism that suggests the substitutes result (i.e., are the vectors of changes in bilateral trade volumes and nations find it optimal to raise third-nation tariffs after border prices, respectively. having signed a PTA) concerns the market power of the A nation choosing its bilateral tariffs optimally would new bloc. If the PTA allows PTA members to better coordi- view this formula as a first-order condition and set it nate their third-country tariffs, the members are likely to to zero to find its optimal tariff. The optimal bilateral raise external tariffs because they will have more purchasing tariffs are: power than before. This effect is only likely to be relevant in customs unions, where countries set tariffs jointly. Even so, ⎛ dp* ⎞ two fairly unrealistic assumptions are required: that the gov- Tod = M od ⎜ ⎟ (6.2) � dM ⎠od ernments share sufficiently similar objective functions, and that their external tariffs are not subject to WTO bindings (or where o indicates the origin nation and d the destination that they are willing to violate their WTO commitments). nation (the nation choosing the tariffs). In general, any- Since most of the effective PTAs are among developed coun- thing can happen to Tod when the nation signs a free trade tries whose tariffs are almost universally bound at near-zero agreement, because, according to the Slutsky equation, the levels (apart from a few low-volume items) and since such direct and cross-good income and substitution effects of the countries rarely violate their WTO bindings, this mechanism PTA-induced price changes could raise or lower the right- is probably of little real-world relevance except for a few com- hand side of the equation (6.2). Real income is expanded modities (agriculture before the Uruguay Round) and a few because the price of the preferential good fell, raising low-trade-volume PTAs among developing countries. demand and hence the optimal external tariff. However, Imported MFN liberalization and protection. A closely there is also substitution toward the low-cost preferential related line of reasoning considers the automatic impact of good, reducing demand for other foreign goods and push- PTAs on the external protection of PTA members, when ing the optimal external tariff down. Attempts to resolve the members impose different tariffs on third countries. Under inherent ambiguity have led to the emphasis on several eco- some circumstances, the PTA effectively lowers the higher nomic mechanisms in the literature. MFN tariff (imported MFN liberalization); in other cir- Preferential and MFN tariffs as complements. If the PTA- cumstances, the PTA effectively raises the lower MFN tariff induced price changes have little impact on the equilibrium (imported MFN protection).22 Preferential Trade Agreements and Multilateral Liberalization 131 A good example of imported MFN liberalization can be The feasibility of this outcome is established by noting found in North-South PTAs. The concept can be explained that the U.S. supply at Po is more than sufficient to cover intuitively with reference to Mexico’s experience. In 1994, the entire Mexican import demand (point 2 is to the right Mexico, along with Canada and the United States, formed of point 1). Note that there is a secondary effect on world NAFTA, which phased in tariff cuts over 10 to 15 years. prices as the United States expands its imports. The new Mexican MFN tariffs were (and still are) much higher than price is at the intersection of the dotted MD curve and the U.S. and Canadian MFN tariffs, but as NAFTA brought XSRoW curve. For simplicity’s sake, this second-order Mexican prices down to the U.S. internal level, domestic impact is not shown in the two leftmost panels. prices in Mexico came to resemble those that Mexico would If one combines this imported MFN liberalization with have observed if it had lowered its MFN tariffs to U.S. levels. the juggernaut logic, the PTA can eliminate all the firms in To put it differently, the high Mexican MFN tariffs became Mexico that would otherwise have opposed MFN liberal- irrelevant because the same goods could be purchased from ization. That is, Mexican industry has no interest in lobby- the United States duty-free, and the U.S. internal price was ing for the maintenance of high Mexican MFN tariffs, linked to the world price via its low MFN tariff (leaving which provide no protection to Mexican industry. In the aside the small sectors still protected by high U.S. MFN tar- case at hand, the Mexican government signed a vast array iffs, such as clothing, textiles, and footwear). In this sense, of PTAs to exchange its now politically useless MFN tariffs Mexico ended up “importing� the low U.S. MFN tariffs.23 against preferential access for its exporters. The argument can be illustrated more precisely by using Since developed countries (the North) tend to have figure 6.5, which shows import demand by the United much lower MFN tariffs on most manufactured goods States (left-hand panel), Mexico (middle panel), and the than developing countries (the South), the mechanism world market for the good under consideration (right-hand suggests that an important implication of North-South panel). The U.S. total supply curve is shown in Mexico’s PTAs for the world trading system is their potential to panel, for reasons that will become clear. The United States lower the southern country’s resistance to further liberal- initially imposes a zero tariff on imports from the rest of ization. Given that most of the South does not participate the world, whereas Mexico imposes a tariff of TMX on in MTN tariff-cutting exercises on the basis of reciprocity, imports from both the United States and the rest of the the North-South PTAs are one of the few ways of triggering world. When Mexico eliminates duties on U.S. imports, juggernaut effects in developing countries. U.S.-made goods can enter Mexico duty-free. Since the Rules of origin and imported MFN protection. The oppo- Mexican internal price is initially above the U.S. internal site result, in which a PTA imports MFN protection to a price, U.S. firms sell to the Mexican market and, in doing country with low MFN tariffs, can occur when highly so, drive down Mexico’s internal price to the U.S. internal restrictive rules of origin are imposed. The argument can price—which, of course, is just the world price. U.S. pro- be illustrated with reference to NAFTA. Since the first U.S. duction entirely displaces Mexican imports from the rest of foray into regionalism, the 1965 U.S.–Canada Auto Pact, the world, and the Mexican MFN tariff becomes irrelevant. U.S. and Canadian rules of origin on autos have been Figure 6.5. Imported MFN Liberalization Partner Home U.S. price Mexican price RoW border price XSRoW Po + TMX SUS 1 Po 2 Po MDUS+MX MDUS MDMX imports imports imports MoMX Source: Authors’ elaboration. 132 Richard Baldwin and Caroline Freund highly restrictive. One of Canada’s motives in pushing for Figure 6.6. An Economic Theory of the GATT the trilateralization of the U.S.–Mexico free trade agree- ment was to extend its restrictive rules of origin to Mexico nation 2’s tariff, t2 and thereby avoid the undermining of the Auto Pact. The rules of origin forced Mexican-based car producers to Nash tariffs import parts and components from the United States or * t2 Canada instead of from third countries. As before, NAFTA equalized U.S. and Mexican internal prices, but this meant that the Mexican prices were linked to the world prices via the higher MFN tariffs in Mexico’s partners. In this way, “imported MFN protection� occurred. NAFTA, with its GATT tariffs rules of origin, had effects that mimicked a rise in the efficient bargaining curve tariffs Mexican MFN tariffs to U.S. and Canadian levels. Although the distortionary impact of rules of origin is * t1 nation 1’s tariff, t1 limited by the level of the MFN tariff, the all-or-nothing Source: Krugman 1991a. feature of rules of origin for final goods can lead to large effective rates of protection. For example, if a US$20,000 NAFTA-origin automobile pays zero tariff and the same non-NAFTA auto would pay 5 percent, a rule of origin of the cooperative outcome� (Krugman 1996, 72).25 The stipulating that a particular component must be made crux of his analysis is to examine the impact that an within NAFTA could make it economical to pay up to exogenously formed PTA has on the costs and benefits of US$1,000 more for the local versus the imported compo- cheating. nent. Although the distortion in the final good market is Much of Krugman’s reasoning is informal, so it is worth limited to 5 percent, the distortion in the component mar- spelling it out explicitly. The whole analysis turns on three ket can be much larger. (This distortion represents the tra- equilibrium welfare levels: ditional logic of effective rate of protection.) • WGATT is the level of a nation’s welfare with global coop- eration (GATT tariffs) Bargaining-Model Stumbling Block–Building Block Logic • WNash is national welfare under noncooperative tariffs The stumbling block–building block mechanisms discussed • Wcheat reflects the nation’s welfare when its government earlier resonate strongly with real-world considerations “cheats�—that is, chooses a tariff to maximize its own because they take advantage of the simple institutional fea- welfare when the foreign government embraces its tures of real-world tariff cutting in PTAs and MTNs. This, GATT tariffs. however, is not how big-think regionalism started reasoning about the issue (Krugman 1991b, 1996).24 There are two logical steps in the approach. The first step When Krugman wondered how regionalism would consists of the obvious point that, from a global perspec- affect the GATT, the tool he grabbed for was simple tive, tariffs are worse than a zero-sum game. Therefore, bargaining-game theory, with two countries that are con- some form of cooperation could be Pareto improving, yet sidering setting tariffs cooperatively (under GATT), or nations have an incentive to cheat. Formally, this is a pris- noncooperatively (Nash tariffs). As figure 6.6 shows, both oners’ dilemma, and it arises when Wcheat > WGATT > WNash. countries prefer the cooperative outcome. Krugman The second step involves a dynamic game that models the observes, “Trade bargaining . . . is characterised by a Prison- circumstances under which cooperation is sustained. As ers’ Dilemma. This Dilemma arises in part from the terms Krugman notes, cooperation is self-enforcing when of trade effect of conventional optimal tariff analysis, but the gains from cheating are more than offset by the also (and presumably in practice mostly) from the effect of losses from the (infinite) punishment. Taking d as the dis- each country’s tariff on the other country’s producer inter- count factor, the present value of cooperating forever is ests� (Krugman 1996, 72). He goes on to invoke all the WGATT/(1 – d ). If cooperation is to be sustained, its value usual theorems of repeated games to think about the build- must exceed the one-period gain from cheating, Wcheat, ing block–stumbling block issue and concludes, “Trade lib- plus the present value of the infinite sequence of the Nash eralisation must be supported by the belief of countries that outcomes that are felt in the next period, after the foreign- if they cheat they will lose from the subsequent collapse ers realize that cheating has occurred, d WNash/(1 – d ). Preferential Trade Agreements and Multilateral Liberalization 133 Clearing the (1 – d ) terms, the condition for self-sustain- cooperation, as depending on the strength of various ing global free trade is: domestic special interest groups. The question here, in the big-think framework, is whether a PTA weakens or WGATT > (1 – d )Wcheat + d WNash strengthens pro-trade and anti-trade interest groups. At In words, each nation compares the value of welfare one end, Winters (1993) argues that regionalism (e.g., in under cooperation with a weighted average of the cheating the EU, on agriculture) strengthened the hand of protec- outcome and the Nash outcome. tionists, since it worsened Olsen’s asymmetry (which The contribution of this approach comes in considering holds that winners from protection are few in number and how a PTA changes the three levels, Wcheat, WGATT, and are easy to organize, whereas losers are dispersed, numer- WNash. Krugman (1996) asks whether the formation of a ous, and difficult to organize politically). Winters terms trade bloc among nations makes them more or less able or this the “restaurant bill� problem. Just as diners at a table willing to cooperate. His answer is that it can cut either way. where the bill will be split equally tend to order too much, Krugman’s core insight—one that has been followed up the EU tended to grant too much protection to farmers. At in a dozen subsequent articles—is that PTAs typically act the other end, Richardson (1994) and Panagariya and to reduce their members’ trade with the rest of the world Findlay (1994) argue that a PTA tends to dilute the influ- and so reduce both the cost and the benefit of cheating. ence of special interest groups via various mechanisms. Since these work in opposite directions, some bargain- Another important line of thinking asserts that the for- approach papers find that PTAs are building blocks (i.e., mation of PTAs creates forces that induce nations to begin they make cooperation more likely), whereas others find or complete multilateral trade talks. For example, Lawrence they are stumbling blocks. (1991), WTO (1995), and Sapir (1996) all argue that the This approach works well for explaining strategic inter- threat of regionalism was a critical element in inducing actions among private agents, but it may be less useful for GATT members to initiate the Uruguay Round and to explaining trade negotiations among large countries. In accept the final Uruguay Round agreement. Bergsten (1996) particular, it requires two assumptions: that the cheating dubs this effect “competitive liberalization� and asserts that period is long enough to make it tempting, and that pun- regionalism fosters multilateralism, and vice versa. Formal- ishment can be usefully modeled as consisting only of tariff izing regional agreements as a force for competitive liber- changes. Since tariff changes are immediately observable, alization, Saggi and Yildiz (2008) study the role of PTAs in the first assumption is problematic unless partner coun- multilateral liberalization in the presence of a nondiscrim- tries are very slow to respond, which may well reflect policy ination constraint, such as the MFN rule in the WTO. in small developing countries. Punishment is likely to They show that if PTAs are not permitted, a country with include more than just tariffs; it may affect external devel- special interests aligned against free trade may oppose a opment aid, treatment of expatriates, migration, military multilateral free trade agreement because it can ride for aid, political support in the international arena, participa- free on the liberalization efforts of others. Nondiscrimina- tion in the North Atlantic Treaty Organization, and other tion means that when any small group of countries liber- concerns. With a shorter period of deviation and more alizes, all WTO members benefit from lower tariffs on tools of punishment available, as is the case for large devel- their exports. The threat of PTAs can reverse that situation oped countries, it is therefore likely that all welfare- by offering the liberalizing countries a way to stop the out- improving agreements are enforceable. Thus, the effect of sider from free riding. In this model, competition to the PTAs on the cost and benefit of deviating from cooperation multilateral system in the form of PTAs can make multi- is unlikely to have a first-order impact on multilateral lateral free trade feasible when it would not be feasible in cooperation when the major countries are involved. This the absence of that competition. type of model may help explain why some agreements A somewhat related idea, which has not been formal- among developing countries are never implemented and ized, is that PTAs are a testing ground for the GATT/WTO why some do not survive when external conditions change. (Bergsten 1996; Lawrence 1996). The prime example here is the EU, which dealt with deeper-than-tariff-cutting lib- eralization for decades before the issues arrived on the Other Links from PTAs to MTNs GATT agenda in the Tokyo and Uruguay Rounds. (See A number of points made in the literature do not fit neatly Ludema 1996 for a partial formalization of this idea.) into the stumbling block–building block framework as we Another line of thinking suggests that PTAs can provide have delineated it. One line of reasoning views a nation’s commitments that boost the credibility of a country’s policy MFN tariffs, or its stance favoring or against multilateral reforms (Fernandez and Portes 1998). This was explicitly 134 Richard Baldwin and Caroline Freund mentioned by Mexico in its request to the United States for have adjusted their external tariffs. She finds that both inte- an FTA and has been highlighted as an important reason that grating and nonintegrating countries have reduced their Eastern European countries have been keen to join the EU. trade barriers, suggesting that regionalism is benign. How- Finally, studying how regionalism affects multilateralism ever, Foroutan does not control for other factors that may is not the only way to look at the relationship between the have induced countries to behave as they did, making it two. A few authors have examined it from the other direc- impossible to disentangle the effects of trade agreements tion: how multilateralism affects regionalism. Both Ethier from those of other global, regional, or sectoral trends. (1998) and Freund (2000a) view regional initiatives as a con- Baldwin and Seghezza (2010) study the relationship sequence of the success of multilateralism. Ethier asserts that between preferential and MFN tariffs in 23 developed and it is a benign consequence, since PTAs intensify world invest- developing countries and find a positive relationship, sug- ment and create incentives for economic reforms in less gestive of the tariff complementarity discussed earlier. developed countries. Freund studies the incentives for and Because their dataset is a cross-section (for 2005), the the sustainability of preferential liberalization when multi- authors note that they cannot assess whether their result lateral tariffs are lower and finds that deeper multilateralism arises from a causal relationship or is just a consequence of, provides greater incentives to form PTAs. The intuition say, fixed effects that affect both preferential and multilateral draws from the complementarity effect between internal and tariff levels. Still, their findings indicate that regionalism has external tariffs. When external tariffs are low, the loss from not led to significantly higher external tariffs. trade diversion is small, but the gains to producers from Using a detailed cross-industry dataset on Argentina for preferential access and to consumers from lower prices 1992, 1993, and 1996, Bohara, Gawande, and Sanguinetti remain. This reasoning could help explain, at least in part, (2004) examine the influence of imports from Mercosur’s the large increase in PTAs since the conclusion of the partner Brazil on Argentina’s external tariffs. They find that Uruguay Round and the formation of the WTO in 1995. increased preferential imports vis-à-vis the value added of the domestic industry led to lower external tariffs in Argentina, especially in industries that experienced trade Regionalism Versus Multilateralism: diversion. This, again, is consistent with the complemen- Empirical Evidence tarity of tariffs. Trade diversion is costly, and to minimize The empirical literature on regionalism as a building it, tariffs were lowered in precisely those industries with the block or stumbling block with respect to free trade is most diversion. One issue is that the study concentrates on small, compared with the theoretical literature, but inter- the effects of increases in preferential imports and does not esting results are beginning to emerge. To date, there is no address the direct effect of preferential tariffs. evidence that regionalism has been a major stumbling Estevadeordal, Freund, and Ornelas (2008) offer the block to free trade and some evidence that it has promoted first attempt to evaluate empirically the effect of preferen- broad liberalization. On the question of whether regional tial tariffs on external trade liberalization in a large group agreements alter countries’ willingness to move to free of developing countries. They examine changes in prefer- trade, there is only anecdotal evidence. PTA members have ential tariffs and MFN tariffs in 10 Latin American coun- continued to participate in the WTO, with little evidence tries and 100 industries over 12 years. When countries of a distinct change in priorities. Finally, on the welfare form a PTA, they lower the tariffs they apply to each other, impact of regionalism, evidence of trade creation exceeds but the duties on imports from outside countries can evidence of diversion. increase, decrease, or remain unchanged. The authors look History points to complementarity between PTAs and at how countries in Latin America, where regionalist forces external liberalization. Irwin (1996) shows that bilateral have been particularly strong since the early 1990s, altered agreements during the nineteenth century induced their trade policies vis-à-vis bloc outsiders after forming broader liberalization. The Anglo-French treaty of 1860 PTAs and, specifically, whether sectors with relatively large led to a host of bilateral agreements that were ultimately preferences have been liberalized or protected to the same linked by the inclusion of an unconditional MFN clause. extent as other sectors. If countries raise their external tar- Precisely because trade diversion associated with high tar- iffs (or reduce them by less) as a result of regional liberal- iffs was costly, the French negotiated numerous such ization, such preferential arrangements should indeed raise MFN-style agreements. concerns about the recent trend. If, instead, preferences Using data on trade and trade policy in 50 countries lead to relatively lower external tariffs, regional agreements between 1965 and 1995, Foroutan (1998) provides a gen- should be accorded a more benign reputation than they eral account of how countries forming regional trade blocs currently have. Preferential Trade Agreements and Multilateral Liberalization 135 The results imply that regionalism is a building block The findings of Foroutan (1998), Bohara, Gawande, and for free trade. There is no clear evidence that trade prefer- Sanguinetti (2004), Estevadeordal, Freund, and Ornelas ences lead to higher tariffs or smaller tariff cuts, and there (2008), and Calvo-Pardo, Freund, and Ornelas (2009)—all is strong evidence that preferences induce a more rapid of which imply that regionalism is a building block for decline in external tariffs in free trade areas. For example, if external liberalization in developing countries—contrast a country that follows a strict policy of nondiscrimination sharply with those of Limão (2006) and of Karacaovali and offers free access to another country in a sector in which it Limão (2008), who find that the United States and the EU applies a 15 percent multilateral tariff, the country would liberalized less during the Uruguay Round in sectors in tend to subsequently reduce that external tariff by more which preferences were granted. One reason for differing than 3 percentage points. As in the study by Bohara and results is that that the countries analyzed are very different. colleagues, Estevadeordal, Freund, and Ornelas (2008) find Since the multilateral system has not enforced much tariff that the complementarity effect is stronger in sectors where reduction on developing countries, tariffs are relatively trade bloc partners are more important suppliers, which is high among that group, creating a large potential for trade precisely where trade discrimination would be more dis- diversion. Lower external tariffs moderate that loss. The ruptive. Using a similar methodology for the Association of results of the first group of researchers named above sug- Southeast Asian Nations, Calvo-Pardo, Freund, and gest that this force is important in explaining changes in Ornelas (2009) also find evidence that regionalism is asso- MFN tariffs of developing countries involved in free trade ciated with unilateral tariff reduction. areas. In contrast, Limão’s work focuses on industrial Recent studies by Limão (2006) and Karacaovali and countries. Tariffs were already quite low in the United Limão (2008) address a related question: whether preferen- States and the EU at the onset of the Uruguay Round, thus tial liberalization by the United States and the EU hindered reducing the importance of this channel. In addition, the multilateral trade liberalization at the Uruguay Round. In theoretical underpinnings of Limão’s analysis, which is the context of the theoretical literature described earlier, used to justify the importance of preferences in North- they examine the goodies-bag stumbling block and, specif- South agreements, rely on the formation of PTAs for ically, whether commitments to liberalize were signifi- noneconomic reasons; preferential treatment is extended cantly different in goods that offered preferences and in in return for noneconomic benefits, such as cooperation goods that did not. These papers, however, do not take into on migration, drug trafficking, or a global political agenda. account the size of the preferences or the importance of This is not the case in South-South PTAs, where the goal is trade in the products that received preferential treatment. to exchange access to markets and improve regional eco- Both papers find that liberalization was relatively smaller nomic cooperation. in products where preferences were used. They argue that, intuitively, because the United States and the EU offer pref- Welfare Consequences of PTAs erences on a unilateral basis to extract concessions from recipients in nontrade areas, they tend to resist liberaliza- A large portion of the empirical literature on trade diver- tion to prevent erosion of preferences. sion versus trade creation has attempted to provide answers The evidence in Limão (2006) is widely misrepresented as to the question of whether bilateralism is bad (see box 6.1, showing that the United States raised tariffs in the Uruguay above). If regionalism is moving world trade away from Round for items on which it granted PTA preferences. Of natural trade patterns, thus reducing world welfare, more course, this cannot be correct, since MTN market access talks diversion will be observed; if regionalism is pushing trade in only involve tariff bindings, and the United States did not the right direction, we should observe little diversion. The violate any of its bindings in the Uruguay Round. Indeed, the analyses also offer an indirect check on the effect of regional data show U.S. tariffs decreasing for all but 12 of the agreements on trade liberalization. If regional members thousands of tariff lines, defined at the Harmonized tend to raise barriers to nonmembers, there should be System–8 product level in the WTO’s database. Formally, strong evidence of trade diversion—increased trade with Limão estimates an econometric model of U.S. tariff cuts members at the expense of nonmembers. By contrast, if during the Uruguay Round. His famous stumbling block regional members tend to lower barriers to nonmembers in finding is that the United States cut tariffs by less than his concert with PTAs, diversion should be limited. econometric model predicted they should have on items for Unfortunately, estimating trade diversion is no easy which the United States had granted PTA preferences before task. It requires knowledge of the counterfactual: what the Uruguay Round. In short, he shows that the U.S. prefer- would have happened to trade if there were no trade agree- ences acted as a “slowing block,� not as a stumbling block. ment? Since this is unknown, assumptions must be made. 136 Richard Baldwin and Caroline Freund Most studies use a gravity equation (which predicts A different perspective is taken by Chang and Winters bilateral trade on the basis of income and other character- (2002), who study the effects of Mercosur—a trading bloc istics) and focus on variables that capture the extent to formed by Argentina, Brazil, Paraguay, and Uruguay in which PTA partners trade more or less than would other- 1991—on export prices to Brazil. They find that wise be expected.26 The key trade creation variable is a Argentina’s export prices increased, whereas the export dummy that is equal to 1 if both countries are members of prices of countries outside Mercosur fell. These price a common PTA; the key trade diversion variable is a effects indicate that Mercosur has hurt outsiders while dummy that is 1 if one country belongs to a PTA and the helping Brazil, a Mercosur partner.27 other does not. (See Frankel, Stein, and Wei 1995; Carrère As Krugman (1991b) shows, whether an agreement is wel- 2006; Lee and Shin 2006.) A positive coefficient on the for- fare improving also depends on trade costs. To determine mer offers evidence of trade creation; a negative coefficient whether nature plays a role in PTA formation, Frankel, Stein, on the latter offers evidence of diversion. Overall, the mes- and Wei (1995) examine whether regional trade is sage from such studies is that trade creation predominates. greater than could be explained by natural determinants— In fact, a concern is that the estimates of the creation effect proximity, size, per capita gross national product, common may be implausibly large, as well as too dependent on the border, common language, and so on. They find in favor of sample of countries and variables included (Haveman and the formation of “natural� trade blocs. Hummels 1998). Krishna (2003) tackles the natural trade bloc question Magee (2008) expands on the traditional approach with by using detailed U.S. trade data to estimate the welfare insights from the literature on the proper estimation of effects from 24 hypothetical bilateral trade agreements in a gravity models. He uses panel data for 133 countries general equilibrium framework and then correlating the between 1980 and 1998 and includes country-pair fixed estimated welfare changes with geographic variables and effects, exporter-year fixed effects, and importer-year fixed trade volumes. Neither geography nor trade volume is effects to capture the counterfactual more accurately than found to be significantly correlated with welfare gains, standard gravity specifications would. The dyad effects indicating that they are not good indicators of the gains pick up what is natural about the trade partners, and the from trade, as the natural trade bloc approach would sug- exporter-year and importer-year effects pick up country- gest. Still, Krishna finds that 80 percent of the potential specific dynamics. Magee finds that the average impact of agreements he examines are welfare improving. Given the agreements on trade flows is small, only 3 percent. More- predominance of trade creation, it is not clear that a cor- over, on average, trade creation dominates trade diversion relation between distance or trade volume and welfare is by about one order of magnitude. necessary to indicate that blocs are formed naturally. To Another strand of the literature uses more disaggre- determine which agreements are most natural, the costs of gated data to examine specific agreements. Clausing (2001) forming an agreement should also be included, and such develops an analysis at the product level of the 1988 costs are plausibly lower with a neighbor or with a large Canada–U.S. Free Trade Agreement (CUSFTA). Using vari- trade partner, as the relationship between the two coun- ations in liberalization across industries to identify trade tries is likely to be well developed. creation and diversion, she finds that in most sectors trade Baier and Bergstrand (2004) develop a general equilib- creation tends to be the rule and trade diversion the excep- rium model to determine which country pairs would gain tion. Taking a similar approach, Trefler (2004) finds both the most from forming PTAs. They then examine whether trade creation and trade diversion in CUSFTA but calcu- these dyads were actually linked by a PTA in a sample of lates positive welfare effects for the average Canadian. 53 countries in 1996. They find that the likelihood of a PTA Romalis (2007) finds that the expansion of CUSFTA to is larger the closer the two countries are to each other, the Mexico, through NAFTA, has been trade diverting. Romalis’s more remote they are from RoW, the larger their gross exercise is similar to Clausing’s and Trefler’s, but he uses domestic products (GDPs), the smaller the difference changes in EU trade over the period to capture what would between their GDPs, the larger their relative factor endow- have happened in the absence of the agreement. Although ment difference, and the wider the (absolute) difference this might create a better counterfactual if the NAFTA coun- between their and RoW’s capital-labor ratios. These vari- tries were very similar to the EU, it could lead to overesti- ables predict the formation of 85 percent of the bilateral mates of trade diversion in NAFTA if the EU’s trade with its PTAs in their sample. Their results thus offer support for the own new and existing trade agreement partners were to natural trade bloc view.28 In subsequent work, Baier and expand more rapidly. Even so, Romalis’s results suggest that Bergstrand (2007) use the same approach to estimate the the welfare costs of the agreement are tiny. impact of PTAs on trade flows. Their key finding is that, Preferential Trade Agreements and Multilateral Liberalization 137 once one takes into account the endogeneity of the agree- prevents countries from backtracking on previous liber- ments, the positive impact of PTAs on bilateral trade alization. Although this would be difficult to accom- becomes more robust and much larger—in fact, five times plish, it is a worthwhile goal. Tariffs in most developing larger—than in estimates that take agreements as exoge- countries are set well below their bindings (by 20 to 30 nous. Thus, countries seem to form PTAs when there is percentage points!), making the term “binding� mean- much to be gained from liberalizing bilateral trade. ingless. A move toward more restrictive bindings would Proving that agreements are natural or unnatural is make regionalism less dangerous and would give mem- daunting, as it requires an assessment of many potential ber countries a lever against a potential increase in the agreements and their welfare consequences—and calculat- political forces favoring protection. ing trade diversion and creation in even one agreement is 2. Agree to lower multilateral tariffs partway, through the already difficult. Nevertheless, there is solid empirical sup- use of preferential tariffs. Even though empirical work port for the more general premise of the natural trade bloc has shown that regionalism tends to be a force for gen- view: that trade blocs are formed by countries that have eral liberalization, irrespective of restrictions, a commit- much to gain from freer trade. ment in this direction would ensure that regionalism The theoretical literature on static effects of trade agree- serves as a building block for free trade. ments highlights the potential costs of preferential liberal- 3. Redefine North-South agreements to incorporate ization and the possibility that trade-diverting agreements preferences from the North in response to MFN liber- may be more viable politically. The empirical literature is alization on the part of the South. In the developing not entirely conclusive, but it does suggest that trade diver- countries, where tariffs are generally higher, this sion is not a major concern, although in some agreements would prevent sizable diversion. In addition, it would and sectors it may matter. Trade diversion may be less rele- be far easier to implement than a range of tariffs across vant than was initially thought because countries form various agreements. trade agreements with “natural trading partners,� where 4. Keep regional agreements open, extending eligibility to trade creation is the norm, or because governments may all countries willing to follow the rules.29 This helps respond to trade diversion by reducing external tariffs. ward against a global outcome dominated by a handful Overall, the empirical literature shows that countries in of protective trade blocs. regional agreements tend to liberalize trade broadly. There is evidence that regional agreement members tend to Trade negotiators should have these guidelines in mind, reduce external tariffs, and that this is especially true of but the WTO should also do its part to ensure a positive members of free trade agreements (as opposed to customs spread of regionalism. This can be accomplished by moni- unions). In addition, trade diversion tends to be small or toring regional agreements among members, publishing nonexistent, which is consistent with endogenous tariff reports on regional agreements that call attention to bad changes that reduce costly diversion. There is, however, behavior, and securing the authority to impose more some evidence that regional agreements may limit trade restrictions on regional agreements. liberalization in the multilateral setting. Given the concerns about regionalism, it is important to highlight where the long-run potential benefits can be found. The best way to ensure that regionalism is welfare Conclusions and Rules for PTAs improving is for countries to pursue serious deep integra- In this chapter, we have examined the effect of regionalism tion agreements. Real resource gains are obtainable if on the multilateral trade system. The theoretical literature countries integrate labor markets, combine regulatory underscores the diverse mechanisms by which regionalism institutions, harmonize standards, and cooperate exten- can be helpful or harmful to that system. So far, there is lit- sively on trade facilitation. Removal of behind-the-border tle evidence that regionalism is overwhelmingly bad for the barriers will enhance trade and welfare without the tradi- multilateral trade system, as some had feared, and there is tional costs of PTAs in tariff revenue and trade diversion. some evidence that regionalism is associated with general The benefits of deep integration include real resource liberalization. gains that will accrue to nonmembers as well as members. To ensure that regionalism is a positive force in the The focus should be on the quality, not the number, of future, four ideal guidelines can be kept in mind: agreements. There is a danger, however, that the present wave of PTAs is being generated by minor agreements that 1. Bind tariff rates at applied rates, leaving no room for tar- will not produce significant benefits, especially given their iff increases following a trade agreement. This measure cost to the world trading system. How to achieve deeper 138 Richard Baldwin and Caroline Freund integration in services, agriculture, transport, and other Nordström (1995), and Bond and Syropoulos (1996a, sec. 3) for examples of preference-erosion stumbling blocks. sectors will be the focus of most of the remaining chapters 9. The theoretical notion was formalized by Limão (2007). of this volume. 10. Levy (1997) illustrates a cherry-picking stumbling block in a highly stylized setup, but his main result is surely more general than his model. 11. This is called the Stolper-Samuelson effect in the economic Notes literature. 12. More formally, without MTNs, governments maximize a polit- The authors are grateful for comments from Jean-Pierre Chauffour, Jaime ically weighted objective function that includes matters affected by the de Melo, and Jean-Christophe Maur. nation’s own tariffs: profits of import-competing sectors, consumer 1. Bhagwati (1993, 2008) and Panagariya and Bhagwati (1999) pro- surplus, and tariff revenue. During the negotiations, a nation’s tariff vide in-depth discussions of these concerns. affects all those things, but foreign tariff levels are linked to domestic 2. Grossman and Helpman (1995), Levy (1997), and Krishna (1998) tariffs via reciprocity, and the objective function now includes the show that regionalism can lead to a reduction in support for multilateral effect of foreign tariffs on exporter’s profits. Since this new impact is liberalization. However, Bagwell and Staiger (1999), Freund (2000b), and negative (higher domestic tariffs reduce exporter’s profits via reciproc- Ornelas (2005b) find that the formation of a PTA can cause its members to ity), announcement of the MTN leads the government to find it politi- lower tariffs. Panagariya (2000) offers an excellent survey of the literature. cally optimal to choose a tariff that is lower than the politically optimal 3. A tariff can theoretically be used optimally to protect an infant tariff before the MTN. industry until it is strong enough to compete. It is difficult, however, to 13. The word “juggernaut,� defined as “any massive inexorable force identify the infants that have potential to grow, and it is often difficult to that advances crushing whatever is in the path,� stems from a British drop protection after it has served its purpose. Moreover, it is typically mispronunciation of the name of the Hindu deity of the Puri shrine, politically connected sectors, rather than growing sectors, that obtain Jagannath. A festival held in Puri involves the “chariot of Jagannath,� an protection. enormous and unwieldy construction that requires thousands of people 4. The next two sections draw heavily on Baldwin (2009). to get it rolling. Once started, however, it rolls over anything in its path. 5. Bhagwati (1991, 71) ascribes the shift to the United States’ conver- 14. The juggernaut logic is from Baldwin (1994, ch. 2.5); it is elab- sion to regionalism, but this contradicts the judgments of trade policy orated in Baldwin and Robert-Nicoud (2005) and Baldwin (2006) and scholars who were engaged in the details of policy at the time (Wonnacott formalized in part by Freund (2000b). The first part of the juggernaut 1987, 17; Schott 1988, 29; Whalley 1996; Hufbauer, Schott, and Clark mechanism—realignment of domestic special interests through 1994, 100). It also contradicts the facts. Bearing witness to the long- reciprocity—has long been recognized in histories of trade liberalization— standing U.S. interest in regionalism is a string of deals that were struck, for example, by Bergsten (1996) and Destler (2005), under the name “bicy- or almost struck, in 1854, 1874, and 1911. In March 1948 the United States cle theory and export politics.� The point was also made by many others, and Canada concluded a secret draft protocol eliminating most tariffs and including Robert Baldwin (1985). The basic idea dates much further quotas bilaterally, but this was ultimately rejected by the Canadians. In back, as Irwin (1996) points out. More recently, the first half of the jugger- 1958, U.S. government procurement was preferentially liberalized in naut logic has been studied formally by Grossman and Helpman (2001) Canada’s favor, and in 1965, the U.S.–Canada Auto Pact came into force. and by Bagwell and Staiger (2002). Juggernaut-like mechanisms were dis- The 1974 Trade Act authorized the U.S. president to negotiate an FTA cussed independently by Hufbauer, Schott, and Clark (1994, 164) and by with Canada, and the 1979 Trade Agreements Act required the president Richardson (1993). Bergsten (1998) mentions an alternative source of to study an FTA in North America. political-economy momentum (“modest liberalization begets broader lib- 6. See Baldwin (1993, 1997) and Serra et al. (1997) for an account of eralization by demonstrating its payoff and familiarizing domestic politics this domino effect. with the issue�). Staiger (1995) uses a repeated game setting, with workers 7. Krugman argued that the multilateral process had run aground moving slowly out of the import-competing sector to generate gradual- with the December 1990 failure of the Uruguay Round and was unlikely ism, but MTNs and GATT reciprocity play no role. Milner (1997) and to get afloat any time soon because the system was plagued by profound Oye (1992), working independently in the international political-economy problems. “While some kind of face-saving document will probably be context, discuss mechanisms by which PTAs can create a proliberaliza- produced, in reality the [Uruguay Round] has clearly failed either to sig- tion political-economy momentum. More recently, Hathaway (1998) nificantly liberalize trade or to generate goodwill that would help sustain presents a similar logic in her positive-feedback model. further rounds of negotiation� (Krugman 1991b, 5). Regionalism, how- 15. But multilateral tariff cutting may also lower tariffs to a level at ever, was not one of those fundamental problems. “But while the move to which PTAs become feasible, when previously they were not; this may free trade areas has surely done the multilateral process some harm, it is have been the case with the Canada–U.S. FTA. almost surely more a symptom than a cause of the decline of the GATT. . . . 16. Agriculture in Europe is a good example. Formation of the cus- The problems of the GATT are so deep-seated that it is unlikely that a toms union realigned special interests in the European Economic Com- world without regional free trade agreements would do much better� munity (EEC) in a way that fostered higher agricultural tariffs. EEC tariffs (Krugman 1991b, 20). He closed his essay with a prediction that history on agricultural goods were not bound until the 1990s, and so the commu- falsified and with a faute de mieux view of regionalism: nity was free to raise its agricultural tariffs without appealing to the Arti- cle 24 exception. Since EEC tariffs were bound for manufacturing goods, The world may well be breaking up into three trading blocs; trade EEC members needed the Article 24 exception to establish the common within those blocs will be quite free, while trade between the blocs external tariff (CET), and this led them to respect the article’s require- will at best be no freer than it is now and may well be considerably ment that the CET not be higher, on average. Roughly speaking, France less free. This is not what we might have hoped for. But the situa- lowered its tariffs; Germany raised its tariffs; and Belgium, Luxembourg, tion would not be better, and could easily have been worse, had the and the Netherlands did little, as their tariffs were initially between the great free trade agreements of recent years never happened. French and German levels. In this way, the formation of the EEC customs (Krugman 1991b, 20–21) union probably reduced the overall size of the EEC import-competing 8. The logic of the preference-erosion or exploitation stumbling sector in manufacturing but raised it in agriculture. block was demonstrated in a Walrasian setting by Riezman (1985) and 17. The key concepts are “coalition externalities� (Haberler’s spillover) Kennan and Riezman (1990) and in a Brander-Krugman setting by and “grand-coalition superadditivity� (global free trade is first-best). The Krishna (1998) and Freund (2000a). See also Goto and Hamada (1995), authors assume that one nation is the undisputed agenda setter and that Preferential Trade Agreements and Multilateral Liberalization 139 unlimited international transfers (transferable utility, in game theory ———. 2002. The Economics of the World Trading System. Cambridge, parlance) are possible. MA: MIT Press. 18. See Lloyd (2002) for a clear development of the veto-avoidance Baier, Scott L., and Jeffrey H. Bergstrand. 2004. “Economic Determinants logic. This line of thinking is one strand in the widely discussed competi- of Free Trade Agreements.� Journal of International Economics 64, no.1 tive liberalization logic advanced by Bergsten (1996). (October): 29–63. 19. This is especially true given all the separability that rules Meade’s ———. 2007. “Do Free Trade Agreements Actually Increase Members’ secondary and tertiary effects (see Meade 1955). International Trade?� Journal of International Economics 71, no.1 20. Given the separability of the markets, the second FTA with RoW (March): 72–95. would yield a price for Good 3 equal to P'. Baldwin, Richard. 1993. “A Domino Theory of Regionalism.� NBER Work- 21. Contributions to the literature that have looked at the comple- ing Paper 4465, National Bureau of Economic Research, Cambridge, mentarity versus substitutes effects include Riezman (1985), Kennan and MA. http://www.nber.org/papers/w4465. Riezman (1990), Krugman (1991a, 1991b, 1996), Richardson (1993), Bond ———. 1994. Towards an Integrated Europe. London: Centre for Eco- and Syropoulos (1996a), Freund (2000a), and Ornelas (2005b, 2007). nomic Policy Research. 22. Shibata (1967), Vousden (1990), Krueger (1993), Richardson ———. 1997. “The Causes of Regionalism.� World Economy 20 (7): (1993, 1994, 1995), and Grossman and Helpman (1995) are all important 865–88. contributors to or users of this line of analysis. ———. 2006. “Multilateralising Regionalism: Spaghetti Bowls as Building 23. Extension of this analysis led to the proposition of the unsustain- Blocs on the Path to Global Free Trade.� World Economy 29 (11): ablity of FTAs. Vousden (1990, 234) argues that Home would be tempted 1451–1518. to lower its MFN tariff to just under that of Partner in order to recapture ———. 2009. “Big-Think Regionalism: A Critical Survey.� In Regional Rules the tariff revenue and that Partner would have an incentive to reply, with in the Global Trading System, ed. Antoni Estevadeordal, Kati Suominen, the resulting race-to-the-bottom tariff making FTAs “unsustainable.� and Robert Teh. Cambridge, U.K.: Cambridge University Press. Richardson (1995) extended and popularized this result. The main results Baldwin, Richard, and Frédéric Robert-Nicoud. 2005. “Juggernaut in Shibata (1967), Vousden (1990), and Richardson (1995)—the irrele- Model—The Lego Version.� Graduate Institute of International Stud- vance of rules of origin and the unsustainability of FTAs—are of little rel- ies, Geneva. evance to real-world policy concerns. (Rules of origin are at the heart of Baldwin, Richard, and Elena Seghezza. 2010. “Are Trade Blocs Building or many current policy debates, and FTAs, not customs unions, are by far the Stumbling Blocks?� Journal of Economic Integration 25, no.2 (June): most prevalent form of PTA.) 276–97. 24. This approach came to be known as the terms-of-trade approach Baldwin, Robert E. 1970. Non-Tariff Distortions of International Trade. after Kyle Bagwell, Robert Staiger, and their students formally modeled Washington, DC: Brookings Institution. the issues, starting with Bagwell and Staiger (1993). For examples of this ———. 1985. The Political Economy of U.S. Import Policy. Cambridge, sort of application to regionalism questions, see Bond and Syropoulos MA: MIT Press. (1996b); Bond, Syropoulos, and Winters (1996); Campa and Sorenson Bergsten, C. Fred. 1991. “Commentary: The Move toward Free Trade (1996); Yi (1996); Bagwell and Staiger (1999); Conconi (2000); Conconi Zones.� Economic Review, Federal Reserve Bank of Kansas City, 76 and Perroni (2000); Freund (2000b); and Ornelas (2005a, 2007). (November–December): 27–35. 25. This is not a new point. It is very clear in the discussion of Johnson ———. 1996. “Globalizing Free Trade: The Ascent of Regionalism.� For- (1953) but probably dates much further back. Indeed, the notion that a eign Affairs 75, no.3 (May–June): 105–20. quid pro quo would be mutually advantageous was probably well under- ———. 1998. “Fifty Years of the GATT/WTO: Lessons from the Past for stood by trade diplomats as far back as Roman times. Strategies for the Future.� Working Paper 98-3, Peterson Institute 26. An alternative is to use computable general equilibrium models to for International Economics, Washington, DC. http://www.iie.com/ identify counterfactuals, but the results are highly dependent on the param- publications/wp/wp.cfm?ResearchID=144. eters assumed. See, for example, Brown, Deardorff, and Stern (1992). Bhagwati, Jagdish N. 1991. The World Trading System at Risk. Princeton, 27. Schiff and Chang (2003) find that the threat of duty-free exports NJ: Princeton University Press. from Argentina to Brazil, measured by Argentina’s exports of the same ———. 1993. “Regionalism and Multilateralism: An Overview.� In New good to another country, also lowers prices of exports by nonmembers of Dimensions in Regional Integration, ed. Jaime de Melo and Arvind Mercosur to Brazil. Panagariya. Oxford, U.K.: Oxford University Press. 28. Egger and Larch (2008) confirm those findings in a larger sample, ———. 2008. Termites in the Trading System. Oxford, U.K.: Oxford Uni- finding also that preexisting nearby PTAs increase the probability that a versity Press. country pair will form a PTA. Bohara, Alok K., Kishore Gawande, and Pablo Sanguinetti. 2004. “Trade 29. Baldwin (2006) argues that the WTO can be involved, as was done Diversion and Declining Tariffs: Evidence from Mercosur.� Journal of in the Information Technology (IT) Agreement, which bound IT tariffs at International Economics 64 (1): 65–88. zero for all countries willing to join. This is a way of “multilateralizing Bond, Eric, and Constantinos Syropoulos. 1996a. “The Size of Trading regionalism.� Blocs: Market Power and World Welfare Effects.� Journal of Interna- tional Economics 40 (3–4): 412–37. ———. 1996b. “Trading Blocs and the Sustainability of Interregional References Cooperation.� The New Transatlantic Economy, eds. Matthew Canzoneri, Wilfred Ethier, and Vittorio Grillli. Cambridge, U.K.: Aghion, Philippe, Pol Antràs, and Elhanan Helpman. 2007. “Negotiating Cambridge University Press. Free Trade.� Journal of International Economics 73, no.1 (September): Bond, Eric W., Constantinos Syropoulos, and L. Alan Winters. 1996. 1–30. “Deepening of Regional Integration and External Trade Relations.� Bagwell, Kyle, and Robert W. Staiger. 1993. “Multilateral Tariff Coopera- CEPR Discussion Paper 1317, Centre for Economic Policy Research, tion during the Formation of Regional Free Trade Areas.� NBER London. Working Paper 4364, National Bureau of Economic Research, Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stern. 1992. “North Cambridge, MA. http://www.nber.org/papers/w4364. American Integration.� Economic Journal 102, no.415 (November): ———. 1999. “Regionalism and Multilateral Tariff Cooperation.� In 1507–18. International Trade Policy and the Pacific Rim, ed. John Piggott and Calvo-Pardo, Hector, Caroline Freund, and Emanuel Ornelas. 2009. “The Alan Woodland. London: Macmillan. ASEAN Free Trade Agreement: Impact on Trade Flows and External 140 Richard Baldwin and Caroline Freund Trade Barriers.� Policy Research Working Paper 4960, World Bank, Management—Center for International Business Education and Washington, DC. http://www-wds.worldbank.org/servlet/WDSCon Research (CIBER), Purdue University, West Lafayette, IN. tentServer/WDSP/IB/2009/06/10/000158349_20090610085729/Ren Hufbauer, Gary C., Jeffrey Schott, and Diana Clark. 1994. Western Hemi- dered/PDF/WPS4960.pdf. Forthcoming in Quantifying the Costs and sphere Economic Integration. Washington, D C: Institute for Interna- Benefits of Regional Economic Integration in Asia, ed. Robert Barro tional Economics. and Jong-Wha Lee. Manila: Asian Development Bank. Irwin, Douglas. 1996. “Multilateral and Bilateral Trade Policies in the Campa, José Manuel, and Timothy L. Sorenson. 1996. “Are Trade Blocs Con- World Trading System: An Historical Perspective.� In New Dimensions ducive to Free Trade?� Scandinavian Journal of Economics 98 (2): 263–73. in Regional Integration, ed. Jaime de Melo and Arvind Panagariya. Carrère, Cèline. 2006. “Revisiting the Effects of Regional Trade Agree- Cambridge, U.K.: Cambridge University Press. ments on Trade Flows with Proper Specification of the Gravity Johnson, H. 1953. “Optimum Tariffs and Retaliation.� Review of Economic Model.� European Economic Review 50: 223–47. Studies 21 (2): 142–53. Chang, Won, and L. Alan Winters. 2002. “How Regional Blocs Affect Karacaovali, Baybars, and Nuno Limão. 2008. “The Clash of Liberaliza- Excluded Countries: The Price Effects of Mercosur.� American Eco- tions: Preferential vs. Multilateral Trade Liberalization in the Euro- nomic Review 92 (4): 889–904. pean Union.� Journal of International Economics 74: 299–327. Clausing, Kimberly A. 2001. “Trade Creation and Trade Diversion in the Kemp, Murray C., and Henry Y. Wan, Jr. 1976. “An Elementary Proposi- Canada–United States Free Trade Agreement.� Canadian Journal of tion Concerning the Formation of Customs Unions.� Journal of Inter- Economics 34: 677–96. national Economics 6 (1): 95–97. Conconi, Paola. 2000. “Trade Bloc Formation under Imperfect Competi- Kennan, John, and Raymond Riezman. 1990. “Optimal Tariff Equilibria tion.� The Warwick Economics Research Paper Series (TWERPS) 571, with Customs Unions.� Canadian Journal of Economics 23 (1): Department of Economics, University of Warwick, Coventry, U.K. 70–83. Conconi, Paoli, and Carlo Perroni. 2000. “Issue Linkage and Issue Tie-In Krishna, Pravin. 1998. “Regionalism and Multilateralism: A Political in International Negotiations.� The Warwick Economics Research Economy Approach.� Quarterly Journal of Economics 113, no.1 (Febru- Paper Series (TWERPS) 558, Department of Economics, University of ary): 227–50. Warwick, Coventry, U.K. ———. 2003. “Are Regional Trading Partners ‘Natural’?� Journal of Politi- Destler, I. M. 1986. American Trade Politics. Washington, DC: Peterson cal Economy 111 (1): 202–26. Institute for International Economics. Krueger, Anne O. 1993. “Free Trade Agreements Versus Customs Unions.� ———. 2005. American Trade Politics. 4th ed. Washington, DC: Peterson Journal of Development Economics 54: 169–87. Institute for International Economics. Krugman, Paul. 1991a. “Is Bilateralism Bad?� In International Trade and Dixit, Avinash, and Victor Norman. 1980. Theory of International Trade. Trade Policy, ed. Elhanan Helpman and Assaf Razin. Cambridge, MA: Cambridge, U.K.: Cambridge University Press. MIT Press. Egger, Peter, and Mario Larch. 2008. “Interdependent Preferential Trade ———. 1991b. “The Move toward Free Trade Zones.� Economic Review, Agreement Memberships: An Empirical Analysis.� Journal of Interna- Federal Reserve Bank of Kansas City, 76, no.6 (November–December): tional Economics 76 (2): 384–99. 5–25. Estevadeordal, Antoni, Caroline Freund, and Emanuel Ornelas. 2008. ———. 1996. “Regionalism versus Multilateralism: Analytic Notes.� In “Does Regionalism Affect Trade Liberalization toward Nonmem- New Dimensions in Regional Integration, ed. Jaime de Melo and Arvind bers?� Quarterly Journal of Economics 123, no.4 (November): 1531–75. Panagariya. Cambridge, U.K.: Cambridge University Press. Ethier, Wilfred J. 1998. “Regionalism in a Multilateral World.� Journal of Lawrence, Robert Z. 1991. “Emerging Regional Arrangements: Building Political Economy 109 (6): 1214–45. Blocs or Stumbling Blocks?� In Finance and the International Economy Fernandez, Raquel, and Jonathan Portes. 1998. “Returns to Regionalism: 5: The AMEX Bank Review Prize Essays, ed. Richard O’Brien, 23–35. An Analysis of Nontraditional Gains from Regional Trade Agree- New York: Oxford University Press. ments.� World Bank Economic Review 12, no.2 (May): 197–220. ———. 1996. Regionalism, Multilateralism, and Deeper Integration. Foroutan, Faezeh. 1998. “Does Membership in a Regional Preferential Washington, DC: Brookings Institution Press. Trade Arrangement Make a Country More or Less Protectionist?� Leamer, Edward E. 1990. “The Structure and Effect of Tariff and Non- World Economy 21: 305–35. Tariff Barriers in 1983.� In The Political Economy of International Frankel, Jeffrry A., and Shang-Jin Wei. 1998. “Regionalization of World Trade: Essays in Honor of Robert E. Baldwin, ed. Ronald W. Jones and Trade and Currencies: Economics and Politics.� In The Regionalization Anne O. Krueger. Cambridge, MA: Blackwell. of the World Economy, ed. J. A. Frankel, ch. 7. Chicago, IL: Chicago Lee, Jong-Wha, and Kwanho Shin. 2006. “Does Regionalism Lead to More University Press. Global Trade Integration in East Asia?� North American Journal of Eco- Frankel, Jeffrey A., Ernesto Stein, and Shang-Jin Wei. 1995. “Trading Blocs nomics and Finance 17: 283–301. and the Americas: The Natural, the Unnatural, and the Super-Natural.� Levy, Philip I. 1997. “A Political-Economic Analysis of Free-Trade Agree- Journal of Development Economics 47: 61–95. ments.� American Economic Review 87 (4): 506–19. Freund, Caroline. 2000a. “Different Paths to Free Trade: The Gains from Limão, Nuno. 2006. “Preferential Trade Agreements as Stumbling Blocks Regionalism.� Quarterly Journal of Economics 115 (4): 1317–41. for Multilateral Trade Liberalization: Evidence for the United States.� ———. 2000b. “Multilateralism and the Endogenous Formation of Prefer- American Economic Review 96 (3): 896–914. ential Trade Agreements.� Journal of International Economics 52: 359–76. ———. 2007. “Are Preferential Trade Agreements with Non-trade Objec- Goto, Junichi, and Koichi Hamada. 1995. “EU, NAFTA, and Asian Responses: tives a Stumbling Block for Multilateral Liberalization?� Review of A Perspective from the Calculus of Participation.� NBER Working Paper Economic Studies 74 (3): 821–55. 5325, National Bureau of Economic Research, Cambridge, MA. Lloyd, Peter John. 2002. “New Bilateralism in the Asia–Pacific.� World Grossman, Gene, and Elhanan Helpman. 1995. “The Politics of Free Trade Economy 25 (9): 1279–96. Agreements.� American Economic Review 85: 667–90. Ludema, Rodney D. 1996. “Increasing Returns, Multinationals and Geog- ———. 2001. Special Interest Politics. Cambridge, MA: MIT Press. raphy of Preferential Trade Agreements.� Journal of International Eco- Hathaway, Oona. 1998. “Positive Feedback: The Impact of Trade Liberal- nomics 56 (2): 329–58. ization on Industry Demands for Protection.� International Organiza- Magee, Christopher S. P. 2008. “New Measures of Trade Creation and tion 52, no.3 (Summer): 575–612. Trade Diversion.� Journal of International Economics 75 (2): 340–62. Haveman, Jon D., and David L. Hummels. 1998. “What Can We Learn Meade, James E. 1955. The Theory of Customs Unions. Amsterdam. North- from Bilateral Trade? Gravity and Beyond.� Krannert School of Holland. Preferential Trade Agreements and Multilateral Liberalization 141 Milner, Helen V. 1997. Interests, Institutions and Information: Domestic Schiff, Maurice, and Won Chang. 2003. “Market Presence, Contestability, Politics and International Relations. Princeton, NJ: Princeton Univer- and the Terms-of-Trade Effects of Regional Integration,� Journal of sity Press. International Economics 60, no.1 (May): 161–75. Nordström, Håkan. 1995. “Customs Unions, Regional Trading Blocs and Schott, Jeffrey. 1988. “The Free Trade Agreement: A US Assessment.� In Welfare.� In Expanding Membership of the European Union, ed. The Canada–United States Free Trade Agreement: The Global Impact, Richard E. Baldwin, Pertti Haaparanta, and Jaakko Kiander, 54–78. ed. Jeffrey J. Schott and Murray G. Smith. Washington, DC: Institute Cambridge, U.K.: Cambridge University Press. for International Economics. Ornelas, Emanuel. 2005a. “Trade Creating Free Trade Areas and the Under- Serra, Jaime, et al. 1997. Reflections on Regionalism: Report of the Study mining of Multilateralism.� European Economic Review 49: 1717–35. Group on International Trade. New York: Carnegie Endowment for ———. 2005b. “Rent Destruction and the Political Viability of Free Trade International Peace. Agreements.� Quarterly Journal of Economics 120: 1475–1506. Shibata, Hirofumi. 1967. “The Theory of Economic Unions: A Compara- ———. 2007. “Exchanging Market Access at the Outsiders’ Expense: tive Analysis of Customs Unions, Free Trade Areas and Tax Unions.� In The Case of Customs Unions.� Canadian Journal of Economics 40 Fiscal Harmonisation in Common Markets, ed. Call S. Shoup, vol. 1, (1): 207–24. 145–264. New York: Columbia University Press. Oye, Kenneth A. 1992. Economic Discrimination and Political Exchange: Sinclair, P., and D. Vines 1995. “Bigger Trade Blocs Need Not Entail More World Political Economy in the 1930s and 1980s. Princeton, NJ: Prince- Protection.� University of Birmingham, U.K. ton University Press. Snape, Richard. 1993. “History and Economics of GATT’s Article XXIV.� Panagariya, Arvind. 2000. “Preferential Trade Liberalization: The Tradi- In Regional Integration and the Global Trading System, ed. Kym tional Theory and New Developments.� Journal of Economic Literature Anderson and Richard Blackhurst, 273–91. New York: Harvester 38: 287–331. Wheatsheaf. Panagariya, Arvind, and Jagdish Bhagwati. 1999, “Preferential Trading Staiger, Robert W. 1995 “A Theory of Gradual Trade Liberalization� In New Areas and Multilateralism: Strangers, Friends or Foes?� In Regionalism Directions in Trade Theory, ed. Alan V. Deardorff, James A. Levinsohn, in Trade Policy: Essays on Preferential Trading, ed. Arvind Panagariya, and Robert M. Stern. Ann Arbor, MI: University of Michigan Press. 47–111. Singapore: World Scientific Publishing. Summers, Lawrence. 1991. “Regionalism and the World Trading Sys- Panagariya, Arvind, and Ronald Findlay. 1994. “A Political Economy tem.� Economic Review, Federal Reserve Bank of Kansas City, Analysis of Free Trade Areas and Customs Unions.� Policy Research 76, no.6 (November–December): 295–302. Working Paper 1261, World Bank, Washington, DC. Trefler, Daniel. 2004. “The Long and Short of the Canada–U.S. Free Trade Richardson, Martin. 1993. “Endogenous Protection and Trade Diversion.� Agreement.� American Economic Review 94 (4): 870–95. Journal of International Economics 34 (3–4): 309–24. Vousden, Neil. 1990. The Economics of Trade Protection. Cambridge, U.K.: ———. 1994. “Why a Free Trade Area? The Tariff also Rises.� Economics Cambridge University Press. and Politics 6 (1): 79–96. Whalley, John. 1996. “Regional Trade Arrangements in North America: ———. 1995. “Tariff Revenue Competition in a Free Trade Area.� Euro- CUSTA and NAFTA.� In New Dimensions in Regional Integration, ed. pean Economic Review 39: 1429–37. Jaime de Melo and Arvind Panagariya, 352–81. Cambridge, U.K.: Riezman, Raymond. 1985. “Customs Unions and the Core.� Journal of Cambridge University Press. International Economics 19 (3–4): 355–65. Winters, L. Alan. 1993. “The European Community: A Case of Successful Romalis, John. 2007. “NAFTA’s and CUSFTA’s Impact on International Integration?� CEPR Discussion Paper 755, Centre for Economic Pol- Trade.� Review of Economics and Statistics 89 (3): 416–35. icy Research, London. Saggi, Kamal, and Halis Murat Yildiz. 2008. “Bilateral Trade Agreements Wonnacott, Paul. 1987. “U.S. and Canadian Auto Policies in a Changing and the Feasibility of Multilateral Free Trade.� MPRA Paper 17561, World Environment.� Canadian-American Committee, C. D. Howe Munich Personal RePEc Archive, Munich, Germany. http://mpra. Institute and National Planning Association, Toronto, Canada. ub.uni-muenchen.de/17561/. WTO (World Trade Organization). 1995. Regionalism and the World Trad- Sapir, André. 1996. “Discussion� of ch. 7, L. Alan Winters, “The European ing System. Geneva: WTO. Community: A Case of Successful Integration?� In New Dimensions Yi, Sang-Seung. 1996. “Endogenous Formation of Customs Unions under in Regional Integration, ed. Jaime de Melo and Arvind Panagariya, Imperfect Competition: Open Regionalism Is Good.� Journal of Inter- Cambridge, U.K.: Cambridge University Press. national Economics 41 (1–2): 151–77. 7 Agriculture Tim Josling The links between regional integration and agricultural viewpoint of developing countries, has been that the terms trade strategy are of increasing interest to developing of the PTAs are largely at the discretion of the preference- countries as they confront the challenge of opening up granting country, with little guarantee that the agreements their economies to competition while mitigating the asso- will not be changed if they lose domestic political backing. ciated adjustment costs. Countries around the world have This asymmetry also implies that the granting country can stepped up their efforts to establish regional preferential continue to extract political benefits from the preference- trade agreements (PTAs) and to coordinate trade relations receiving partner, including support for economic and with other regions. Agricultural trade is a core component political positions. of many of these trade initiatives, and a large part of the The role of PTAs in agricultural development varies gains from regional integration depends on the inclusion greatly, from strategic and deliberate to largely passive and of agricultural and food products in PTAs. Agricultural reactive. Countries can choose to plan their agricultural trade policy and regional integration agreements can strategies on a regional basis to take advantage of scale together serve as instruments for accelerating growth and economies and to overcome some of the constraints facing contributing to poverty alleviation. small national markets. Often, however, agriculture is Although the motivation for PTAs is often political, brought into regional agreements through extraneous these agreements have significant implications for agricul- circumstances rather than careful planning; agriculture ture and other sectors of the economy. Many North-South becomes one element in a broader set of complex trade-offs. agreements were concluded after a country gained inde- PTAs often have an underlying rationale of contributing to pendence in order to maintain trading links developed in increased regional cohesion and political integration, and the colonial era—typically, for trade in raw materials and the treatment of the agriculture sector becomes of inter- agricultural products. Other PTAs were instituted as part of est, beyond strictly commercial considerations. Food the development of alliances and to bolster regimes that security and the coordination of food policies and mar- were under threat. Agricultural products often provided keting infrastructure may be important reasons for the trade opportunities that reinforced such alliances. PTAs inclusion of a strong agricultural component in regional have long been used by developed countries to pursue policies. Thus, the key issues in addressing agricultural overseas developmental objectives through the provision of trade in PTAs are whether regional integration promotes preferential market access, often for primary products. or hinders the development of a sustainable, competitive Such access, however, has frequently been constrained agriculture sector and whether agricultural trade consid- when sensitive domestic agricultural products were erations contribute to or detract from the benefits of involved. regional integration. More recent thinking has cast doubt on the longer-term This chapter attempts to put into perspective what we benefits of PTAs on the grounds that they tend to lock know and do not know about the economic impact of exporting countries into a particular pattern of exports, PTAs on agricultural development. The next section sur- often of unprocessed raw materials and farm products, veys the arguments for and against preferential trade inte- while competitors develop other markets and diversify gration as a development strategy for agriculture. These their range of export products. A further problem, from the arguments relate, in general, to the effects of opening up 143 144 Tim Josling trade among selected trade partners, as opposed to relying which regional integration can provide the scale needed for on either unilateral or multilateral actions. The rationales such cost reductions depends on specific circumstances, also cite possible economies in the joint production of but, in principle, the achievement of economies of scale can public goods that benefit agriculture, ranging from be a positive argument for regional integration. If the part- research and extension to food security reserves. The sec- ner with the inefficient agriculture sector can make use of ond section then reviews what is known empirically about scale economies to become efficient, costs will decrease. the impact of PTAs on agricultural trade. The third dis- But to treat PTAs as a nursery for potentially competitive cusses some ways in which PTAs have dealt in practice with sectors is controversial at best: the infant may become a set of problems that are commonly encountered when dependent on the protected market within the PTA and agriculture is included in PTA provisions. Both regional may not have an incentive to become competitive outside and bilateral PTAs are considered, as the agricultural prob- the area. Moreover, import-competing sectors will tend to lems differ somewhat in the two types of agreements. shrink with regional integration and may lose some bene- fits of scale. Thus, the larger question is whether there is a possibility of a broad restructuring of the agriculture sec- Economics of Agricultural Trade in PTAs tors of each of the PTA partners so that economies of scale In most respects, the economics of agricultural trade in can be exploited and resources redeployed to take advan- PTAs is no different from the economics of nonagricultural tage of regional (as opposed to national) cost advantages. trade. As with trade in nonfarm goods, agricultural trade in Fluctuations in output often mark agricultural markets, PTAs benefits from static gains related to expanded market and trade is a vital means for offsetting the impact on avail- access and from more dynamic gains related to the spread able consumer supplies. The easier trade is, the less is the of ideas, innovations, and know-how (see Baldwin, ch. 3 in cost of market disruption to consumers. Greater regional this volume). food security is thus another plausible argument for inte- Two important initial questions frame any agricultural gration of the regional partners’ food supply network. All trade strategy in PTAs: (a) how high are domestic (tariff and parties to a PTA that includes an open internal agricultural nontariff) barriers relative to those of regional partners and market will enjoy the advantages of more secure access to other countries, and (b) how efficient are the export sectors regional supplies. Even where weather and other related within the region? If the region includes suppliers of agri- determinants of yield variations are regionally correlated, cultural products whose costs are lower than those of more there can still be benefits from pooling risks. Storage facili- distant exporters, then the regional strategy carries benefits ties can be collectively operated, and regionally coherent similar to the unilateral or multilateral lowering of tariff transportation systems can improve distribution. There barriers. Regional supplies can be integrated into a coun- may, however, be a political cost because of loss of the abil- try’s food policy, and ensuring access to those supplies will ity to control domestic markets. be an element in food security policy. The higher the exist- Agricultural trade in PTAs can benefit from some of the ing tariffs (and other trade barriers) that restrict regional considerations of spatial or economic geography that apply trade in these products, the greater will be the benefits of to trade in goods. Some of these have to do with the provi- preferential liberalization to consumers. But this also sion of public goods, where the good concerned is valued implies greater disruption to domestic producers, who pre- (and underprovided) across local jurisdictions. More gen- sumably have not had to face regional competition. When erally, both public and private sector actions can be there are other products that could gain from the export expanded to a regional scale with advantage. Greater coor- opportunities that would open up with regional trade, a dination of export strategies, more reliable supply chains beneficial transfer of resources from the import-competing for buyers, shared control over the quality and safety of to the exporting agriculture sector may be possible. But exports, a better bargaining position with importers in if the agriculture sector in the regional partner is not effi- other countries, and the possibilities for branding and cient, the reduction in trade barriers may merely substitute a labeling of regional products are all likely to result in high-cost partner product for a more efficient third-country expansion of export markets. supply. Under these conditions, the advantages are likely to be small and the costs high. Preferential Agricultural Trade and Multilateral These considerations need to be seen in a dynamic con- Commitments text. Inefficient agricultural suppliers could become low cost if their inefficiencies had been the consequence of lim- As noted by Baldwin and Freund (ch. 6 in this volume), a ited markets and diseconomies of scale. The extent to key tension between bilateral or regional trade rules and Agriculture 145 multilateral trade rules arises from the latter’s requirement the North American Free Trade Agreement (NAFTA), the that PTAs eliminate tariffs and other trade measures on United States is moving toward a free internal agricultural “substantially all trade� and that the level of preference be market with Mexico and Canada, with few exceptions. 100 percent. (The multilateral rules in question are those of More recently, the United States has negotiated agreements the General Agreement on Tariffs and Trade, or GATT, as with Chile, Central America, the Dominican Republic, and now embodied in the World Trade Organization, or Australia, giving those countries preferred, although not WTO.)1 Although there has as yet been no agreement on free, access to U.S. markets. Less sensitive food products are the interpretation of “substantially all trade,� agriculture is also included in the generalized system of preferences (GSP) the sector most often excluded or treated differently; man- schemes. Substantial amounts of agricultural trade thus do ufactures are far more likely to benefit from tariff reduc- face barriers less restrictive than MFN tariffs. Presumably, tions in PTAs than are agricultural goods (Fiorentino much of this trade is diverted from lower-cost suppliers. 2005). The prospect that competing exporters will chal- Yet, the short-term market access gains from a PTA have lenge the exceptional treatment of agriculture in PTAs is to be weighed against the possible adverse effects on the remote, however, because these competitors tend to benefit multilateral trading system. Many of the problems that from it. Exporters within the PTA have implicitly agreed to make the incorporation of agriculture in a PTA regime dif- the exclusion and would be reluctant to challenge a partner ficult also prevail in a multilateral context. A prominent with respect to mutually agreed decisions. issue in this regard is the extent to which disciplines on The requirement for movement toward full internal free domestic farm subsidies can be included in PTAs. It is often trade (100 percent preference) has also been problematic. assumed that the conduct of domestic policy is outside the In some PTAs, partners gain an advantage from preferential realm of PTAs, but this is not always the case. The move- tariffs but still face nonzero rates. Despite the inconsistency ment toward “decoupled� policies, encouraged by the with GATT Article XXIV, several of these preferential tariff WTO Agreement on Agriculture, has the advantage of schemes have been permitted. As noted, there would be lit- making it easier to have free trade in a commodity and still tle outside interest in challenging such schemes, given that maintain domestic support policies (box 7.1). Neverthe- the lack of 100 percent preference works to the advantage less, the existence of an active domestic support policy, of the excluded supplier.2 The WTO requirement that the involving subsidies and market management, complicates free trade area encompass “substantially all trade� (which, the negotiation of free trade in those products. If agricul- for most developing countries, includes agriculture) tural trade can be omitted from PTA provisions, the imposes costs on PTAs that include high-cost agriculture question of domestic support does not arise. Conversely, sectors. It may be therefore better in economic terms for if agriculture cannot be excluded without violating WTO such PTAs to exclude highly protected sectors, including provisions, the potentially problematic issue of domestic agriculture. support policies cannot be avoided. The complementarity of preferential tariff reduction The treatment of export subsidies in PTAs is similarly with multilateral trade developments can be part of a posi- problematic. Various trade agreements include provisions tive strategy for agriculture: the multilateral system could that countries may not employ export subsidies in mutual work to lower most favored nation (MFN) tariffs and trade. Although this stipulation sounds logical, it is not reduce trade-distorting subsidies, making it easier and less easy in practice to ban subsidies paid on internal trade costly to negotiate PTAs. The coordination of multilateral without creating an incentive to import from outside and a strategy among the regional partners also offers other pos- disincentive to export within the PTA. In effect, export sibilities for regional negotiating strategies and negotiating subsidies also have to be controlled at the WTO level. groups and opens an opportunity to develop strategies that The current Doha Round of WTO negotiations would, combine regional and multilateral integration.3 if successfully completed, make a significant difference to There are indications that PTAs may be more successful the ease with which agricultural trade could be opened up than multilateral agreements in opening markets for agri- within PTAs. Under the provisions of the 2008 draft cultural goods. It often seems easier to fine-tune market modalities, tariff levels for developed countries would access within discriminatory agreements, through selective decline by more than 50 percent for agricultural products, inclusions. The European Union (EU) provides limited and tariff-rate quotas (TRQs) for sensitive commodities access for sensitive agricultural products to the many coun- would be expanded. This would reduce both the degree of tries that have signed such agreements, including the preference for partner suppliers (and hence the risk of Mediterranean countries; the African, Caribbean, and trade diversion) and the adjustment for import-competing Pacific (ACP) countries; South Africa; and Mexico.4 Within sectors. Under an agreed schedule of WTO tariff reductions, 146 Tim Josling Box 7.1. The WTO Agreement on Agriculture The Agreement on Agriculture entered into force with the establishment of the WTO on January 1, 1995. The preamble to the document cites the agreed long-term objective of the Uruguay Round reform program: to establish a fair, market-oriented agricultural trading system. The program includes specific commitments to reduce support and protection in the areas of domestic and export subsidies and of market access and through the establishment of strengthened and more operationally effective GATT rules and disciplines. The Agreement on Agriculture also takes into account nontrade concerns, such as food security and the need to protect the environment, and it provides for special and differential treatment for developing countries, including an improvement in the opportunities and terms of access for agricultural products of particular export interest to these members. In principle, all WTO agreements and understandings on trade in goods apply to agriculture. These include GATT (incorporated into the WTO agreements as GATT 1994) and WTO agreements on such matters as customs valuation, import-licensing procedures, preshipment inspection, emergency safeguard measures, subsidies, and technical barriers to trade. Where there is any conflict between these agreements and the Agreement on Agriculture, the provisions of the latter prevail. The WTO General Agreement on Trade in Services and the Trade-Related Aspects of Intellectual Property Rights Agreement are also applicable to agriculture. Source: WTO Agreements series: Agriculture. the marginal impact of quicker reductions for regional or preferred partner with less threat to domestic agriculture bilateral partners may be more tolerable. In addition, sectors. Regional and bilateral PTAs, however, cannot deal export subsidies would finally be eliminated—a step that effectively with agricultural export subsidies and domestic has proved difficult to accomplish within PTAs. But the farm support, and so it is likely that the WTO will continue main contribution that the WTO negotiations on agricul- to be needed as a complement (Josling 2009). Meanwhile, ture can make to the process of regional and bilateral trade regionalism poses problems for the multilateral system. liberalization may be to push countries to abandon price PTAs may pick the easiest agriculture sectors to liberalize, supports in domestic markets and embrace direct pay- leaving the most difficult products to the WTO. PTAs can ments for public goods or for income relief. Such policies also lead to investment in the “wrong� countries, just to get will be more compatible with PTAs, as well as beneficial for access to their agricultural markets. Moreover, too many the multilateral trading system. simultaneous negotiations can overstretch resources. Some Because of the slow pace of the Doha Round negotia- PTAs can be “strategic�: an example is the efforts of the tions on agriculture, regional negotiations may have to Southern Cone Common Market (Mercosur, or Mercado bear the brunt of attempts to further liberalize agricultural Común del Sur) to get the rest of South America into its trade and to open access to new markets for agricultural camp before dealing with the United States and NAFTA. exports. Thus, in agriculture there is a strong degree of Some PTAs can be competitive, as when the EU and the complementarity in trade negotiations. Plurilateral agree- United States compete for the Mercosur market. Such ments can erode market access barriers, but they can also activities are likely to distract from the WTO, or they might foster less advantageous trade flows by discrimination distort the multilateral nature of the Doha Round. among suppliers. Multilateral talks can reduce the scope These issues have been discussed in the literature of for such trade diversion. The multilateral process can han- trade strategy. Analyses by Zissimos and Vines (2000) and dle subsidy reduction, which, in turn, makes it easier for by Andriamananjara (2002) suggest that joining a PTA can countries to agree to opening up regional or bilateral be the best “safe-haven� strategy when other countries are trade. This complementarity, however, depends on also doing so. But this does not imply that the end result is progress at the multilateral level. Currently, that is the one large free trade area, given that PTA membership con- stumbling block. fers a terms-of-trade gain on members at the expense of nonmembers. Some members, at least, will be better off by limiting PTA membership than by allowing expansion to Economic Integration as an Agricultural Strategy cover the world as a whole. The effect is similar to that sug- Is regionalism a better approach to agricultural trade pol- gested by the domino theory of the dynamics of regional icy than reliance on improved market access through the trade blocs (see Baldwin, ch. 3 in this volume). As PTAs multilateral system? PTAs may be better than tariff reduc- expand, the attraction of being within the bloc (or the cost tions through the WTO at improving market access for the of being outside) increases, but the marginal gain to existing Agriculture 147 members of adding one more (small) market to the bloc, with the restrictive rules of origin for many processed and the extra administrative and political cost of a large products, have severely limited the role of trade preferences membership, will act as a brake. The implication is that each in encouraging agricultural diversification in developing such agreement will tend to find its equilibrium size, where countries. the costs and benefits of enlargement are in balance.5 Tariffs introduce a wedge between the world price of a So, is the pursuit of PTAs a short-term or a long-term product and the price on the domestic market. Trade pref- strategy in agricultural liberalization? Pursuing bilateral erences allow products from the beneficiary country to North-South PTA arrangements at the regional level may enter the partner country with lower import duties than lead to short-term benefits of access to agricultural mar- are applied to other countries’ products under the partner kets for participating developing countries. But developing country’s MFN tariffs and hence capture some of the countries should be aware that preferential access is likely wedge. They give suppliers in beneficiary developing coun- to be eroded as more countries sign such deals, reducing tries access to part or all of the price premium that nor- the value of preferences (García-Alvarez-Coque 2002). For mally accrues to the importing country government as tar- PTAs to be beneficial in the longer term, governments and iff revenue. The acquisition of these rents raises returns in stakeholders should implement a set of reforms to help the developing country and, depending on the nature of sustain the growth of domestic agriculture and reduce the competition in domestic product and factor markets, stim- poverty of the agricultural population. ulates expansion of the activity concerned, with implica- As noted, this type of discriminatory trade agreement tions for wages and employment. has both positive and negative aspects. On the positive side Developing countries, especially the least developed is the ease of reaching an agreement among a small group of countries, face much higher trade-related costs than adjacent countries. Often, the countries involved will share other countries in getting their products into interna- historical and social perspectives on trade and agriculture. tional markets. Some of these costs may reflect institu- But this ease of reaching agreement comes at a cost. PTAs tional problems within the countries themselves, such as tend to “cherry-pick� the easiest trade areas in which to con- inefficient practices and corruption, and they require a clude a deal, leaving the more difficult ones to the WTO. domestic policy response. But some reflect weak trans- The ease of reaching an agreement may reflect the willing- portation infrastructure and firms’ lack of access to stan- ness of the parties to avoid hard decisions by excluding sen- dard trade-facilitating measures such as insurance and sitive sectors such as agriculture from the deal. trade finance. Those PTAs that have been most effective in opening up agricultural markets have tended to include as members Empirical Evidence on Agricultural major agricultural exporters that see the advantages of Trade and PTAs expanding markets. Countries that are mainly importers of farm products are less likely to agree to open up markets, What might a theoretical approach to the issue of agricul- and hence the benefits to the sector may be small. tural trade and regional agreements suggest? Would one In PTAs among members in different regions, the temp- expect the proliferation of PTAs to have brought about tation to exclude sensitive sectors of agriculture is even trade expansion in agricultural products? In his analysis greater, as there is less probability that the deal will include of the economic impact of regional integration on agricul- provisions of benefit to agricultural export interests. Given tural trade, Goto (1997) concludes that the higher the level that bilateral PTAs often involve countries that are not geo- of preintegration protection, and the lower the degree of graphically close, there is often an opportunity to negotiate product differentiation, the greater the impact of regional with a country with a complementary agricultural pattern integration. He hypothesizes that “regionalism has more to take advantage of trade opportunities. In practice, this [of an impact] on agricultural trade than on manufactur- often works in reverse, as countries cherry-pick partners so ing [trade], because the initial level of protection is higher as to avoid conflicts over agriculture. The existence of and the degree of product differentiation is lower for agri- political tensions in trade agreements usually indicates cultural products.� On the basis of this theoretical conclu- potential economic benefits that could be realized from sion, PTAs could be expected to have a significant role in changes in trade patterns. agricultural liberalization, and this hypothesis will be The positive and negative aspects of PTAs are com- explored in the brief review of the empirical literature that pounded by the apparent advantage that nonreciprocal follows. preferences give to the recipient country relative to others. The literature on agricultural trade issues in PTAs is But overall, the limiting factors mentioned earlier, together thin and scattered, and there is very little by way of detailed 148 Tim Josling and comparative analyses of the arrangements made for health and safety regulations for regional trade. These agriculture in regional PTAs. Bilateral PTAs are somewhat models do not evaluate the trade policies themselves better documented, as they tend to be focused on a more and cannot indicate whether a particular trade strategy is limited number of products, and the trade flows and con- desirable. ditions of market access are watched closely by the domes- The third group of studies consists of ex ante evalua- tic sectors concerned. Unilateral preferences are again the tions of prospective agreements. These studies often use subject of study, in part because of their dependence on computable general equilibrium (CGE) models to calcu- periodic renewal and in part because of their direct impact late trade flows and the welfare implications of policies. on competing suppliers. An example of a case for which Most of the few studies that focus on agriculture take a there is adequate information and several empirical analy- similar approach, analyzing the significance of trade agree- ses is the EU’s regime for bananas, where the WTO case has ments for agricultural trade. brought much transparency to the way in which the ACP Each of the three sets of studies has strengths and weak- countries sell their bananas to Europe and to the marketing nesses. The study findings are reviewed below, and issues of choices of the excluded suppliers. Sugar sales to the United relevance to the specific question of the empirical evalua- States and the EU under PTAs have also been closely ana- tion of trade preferences are then addressed. lyzed, and adequate data exist for calculating the effects of such trade arrangements. Regional Trade Flows Absent such comprehensive and detailed studies dealing with the amount and type of agricultural products traded The steady growth in world trade in relation to world out- within PTAs, the assessment of the costs and benefits of put is seen as an indicator of the success of the multilateral agricultural trade in PTAs has tended to rely on more con- trade rules put in place by the GATT and reinforced by the ceptual studies. These studies can be grouped by their WTO. The assumption has been that the elimination of focus on one of three questions: trade barriers and the extension of trade rules have stimu- lated trade flows. But has trade tended to be concentrated • Has regional trade increased faster than trade with third among regions? The literature (e.g., Lloyd 1992; Anderson countries? and Blackhurst 1993) tentatively concludes that the trading • Does the existence of PTAs explain trade flows among system has not developed into a series of intensively trad- the partners in such agreements? ing blocs, with decreased interbloc trade.6 Trade among • What are the gains and losses from participation in blocs remained resilient, despite the burst of “regionalism� regional or bilateral PTAs? in trade policy that characterized the decade from 1985 to 1995. Nevertheless, evidence of increased intrabloc trade The first group of studies essentially consists of explo- has been a widespread, if not a dominant, feature of the rations of the extent to which world trade is becoming more trading system. or less regionalized. It is difficult to derive direct implica- Several studies find that intraregional trade in agricul- tions from the outcome of such studies because regional tural and food products grew during the 1980s and 1990s trade could well increase rapidly even in the absence of (Vollrath 1998; dell’Aquila and Kuiper 2003). With regard regional agreements. The nature of agricultural and food to the effects of particular trade blocs, Diao, Roe, and trade itself is changing over time, and the goods and serv- Somwaru (2001) find that, on average, agricultural trade ices that are traded across continents may vary with the under NAFTA, the EU (then consisting of 15 countries), regional composition of trade. But it is still useful to have Mercosur, and Asia-Pacific Economic Cooperation grew these studies as a way of putting the regionalization of agri- more rapidly than did total world agricultural trade.7 In cultural trade policy in context. particular, growth in intraregional agricultural trade The second group of studies generally involves ex post exceeded the growth in extraregional agricultural trade explanations of trade flows. The most commonly used for these PTAs. technique is a gravity model. By inserting dummy variables for the existence of PTAs in regression equations, this Trade Flows and Preferences method aims to determine the significance of such trade policies in the explanation of trade flows. Agricultural A PTA increases trade among members through preferen- trade flows can be isolated in these studies, and the impor- tial treatment. The question is whether that growth comes tance of the trade policy for agriculture can be determined. at the expense of the rest of the world. Despite a number of One application has been to look at the implications of theoretical and empirical contributions in recent years, the Agriculture 149 effects of PTAs on trade in agrifood products have not been to meet import requirements at a reasonable cost. Lowering evident from the literature because most of the studies tariffs against third countries, even if done unilaterally, is have dealt with merchandise trade. To what extent agrifood a strategy that would minimize the cost of giving prefer- trade among PTA partners has increased and how much ence to high-cost imports. Making exceptions for tariff- of the increase could be attributed to trade diversion free access for high-cost partner supplies would seem less remained for some time an open empirical question. desirable. Too stringent rules of origin will also add to the Recently, researchers have tried to answer this question possibility of trade diversion because they will discourage but have come to mixed conclusions. Jayasinghe and Sarker processing of third-country raw materials in the partner (2008) analyzed NAFTA’s trade creation and trade diver- country. sion effects on trade in six selected agrifood products from The main drawback to relying on preferences for export 1985 to 2000. The results show that the share of intrare- products for agricultural development is the effect on the gional trade within NAFTA is growing and that NAFTA has pattern of domestic agricultural production. A few export displaced trade with the rest of the world. Countries par- commodities will benefit from preferences, but this will set ticipating in NAFTA have moved toward a diminished the economic context for other products that have to com- degree of relative openness in their agrifood trade with the pete for land and labor. In addition, the guarantee of access rest of the world. Grant and Hertel (2005), however, find, under unilateral preference schemes may be a Faustian bar- with only a few commodity exceptions, that PTAs have gain, as the supplying economy becomes more dependent increased trade with nonmembers even as members have over time on the continuation of the preference margin. increased trade among themselves. As these examples Preferences can create a degree of dependence that con- show, the impact of PTAs on agriculture varies among strains flexibility and diversification and results in high-cost regions and among agricultural products. production of preferred products (Topp 2001). Moreover, other countries will have an interest in reducing those preference margins over time. The most highly protected Evaluating Trade Preferences products have the highest potential margins of preference, A major problem with the standard databases on tariff lev- but these are also the products that are likely to lose the els is that they rarely incorporate preferential tariffs. Given most protection through WTO negotiations. the amount of world trade that enters countries under Many preferences are, in any case, quantity constrained. preferential (or zero) tariffs, the picture of market access When preferences are granted on products for which presented in these databases is misleading. This omission is domestic prices in industrial countries are much higher gradually being rectified; one notable advance has been the than world prices, such as sugar in the EU and the United development by the French research organization Centre States, quantities are limited, to avoid undermining the dis- d’Études Prospectives et d’Informations Internationales of tortionary policies that generate the large divergence a database that includes full information on the preferen- between domestic and world prices. In these instances, tial tariffs accorded developing countries. Analysis of the preferential access can lead to substantial gains for pre- extent to which PTAs promote the development of agricul- ferred suppliers but little hope of market expansion and ture by opening up markets remains difficult but is now high probability that the gains will be eroded. subject to empirical exploration. In addition, some of the preference rents may not be Trade preferences, both reciprocal and nonreciprocal, available for development. How much of the available rent can provide the premium over the normal rate of return is actually obtained by suppliers in developing countries that is required to encourage investment in developing depends on the nature of competition in the industry and economies. The increase in agricultural trade attributable the regulations governing the granting of preferential to preferences leads to more output and, if there are scale access, among other factors. If there is little effective com- economies, to lower costs, stimulating further trade. Thus, petition among buyers, exporters may be unable to capture the search for preferential access to foreign markets is nat- much of the price premium. Olarreaga and Özden (2005) urally a component of national trade policy. The degree of find that only a third of the available rents for African preference, however, can be fleeting if others are also nego- exports of clothing to the United States under the African tiating market access. The benefits may be those of the first Growth and Opportunity Act actually accrue to exporters. mover and can be eroded steadily over time. Furthermore, satisfying the rules governing preferences The granting of trade preferences is also a policy deci- raises costs and reduces the extent to which the preferences sion subject to evaluation. The key question is whether the increase actual returns. The costs of satisfying the rules of partner receiving the preference is able or likely to be able origin in preference schemes have been cited as a major 150 Tim Josling reason for low use of preferential access in some cases whereas for others, it will be a move toward greater protec- (UNCTAD 2001; Brenton and Ikezuki 2005; and Brenton, tion (see Andriamananjara, ch.5 in this volume). ch.8 in this volume). Gravity models and other models that seek to “explain� Tariff preferences can lead to other adverse effects that trade flows are of limited use in evaluating the value of need to be taken into account. Negotiations in the Doha preferences. Ex post analysis of why trade has taken place Round have shown that existing preferences can lead to does not answer the fundamental question of whether dif- support for highly protectionist policies in industrial coun- ferentiating one’s own tariff schedule in exchange for simi- tries and can weaken proposals that would substantially lar differentiation by others is beneficial. If the gravity reduce such levels of protection. This not only causes a rift model identifies “natural� trade partners, then the question among developing-country negotiating positions but also is, why is that natural trade, based on proximity and perpetuates policies that depress world markets and rein- income, not taking place already? The answer could be that force dependence on preferences for export revenue. the trade policy of the partner country precludes such Differences and inconsistencies between preference trade, but many other explanations could intrude. And the schemes can prevent developing-country suppliers from “best� trade partners may well be on other continents. So, evolving global market strategies. Furthermore, the prefer- gravity models are an interesting way of looking at trade ence schemes may not be directly consistent with poverty patterns, but a shaky guide to policy action. reduction objectives: beneficiaries of trade preferences are CGE models also have drawbacks for the evaluation of not always the poorest constituents in developing coun- preferences, although they do address the key issue of the tries. Although rents do accrue to the developing country, economic benefits and costs. The difficulties stem from they will tend to benefit the owners of the most intensively whether the situation to be analyzed can be specified in used and the most limiting factors. enough detail. The trade policy question may require Relatively few studies have directly measured the value knowledge of particular market conditions, such as quality of preferences. The value depends on the difference in and production cost. The device of assuming that each returns in different markets. The rents accrue to the holder country produces a somewhat different version of traded of the preference, but those rents are usually subject to the products masks the question of whether and how such dif- trade policy of the preference-giving country. A recent ferentiation can be created and exploited. An exporter will example of this view of preferences is given in Paggi, ask, “What are the regional markets that can open up for Yamazaki, and Josling (2005): the value of improved access my product, and how can I adjust to meet the market to Central American markets by U.S. exporters as a result requirements?� The CGE model will reply that the relevant of the Central America Free Trade Agreement (CAFTA) substitution elasticity is already in the model and is not a and its extension to the Dominican Republic (CAFTA–DR) part of the policy space. depends on who else has such preferences in those markets Therein lies the dilemma facing analysts in this area. and how long any advantage over other competitors will Every PTA is different in its coverage and treatment of last. The United States competes with the Mercosur coun- products. Moreover, members’ motivations and interests tries and with the EU in Central American agricultural can differ widely, making the overall analysis of the agree- markets, and evaluation of the value of CAFTA for U.S. ment of limited use to individual countries. And within exporters is as much a function of the state of trade rela- each country, the calculation of costs and benefits will be tions among these other countries as of the details of specific to conditions in particular sectors. In brief, the task CAFTA as such.8 of analyzing any particular decision for a country on the One problem with calculating the value of regional and basis of the benefits to be gained from a trade agreement is bilateral preferences in agriculture is that the models tradi- heavily data intensive, context specific, and time related. It tionally used do not adapt well to such questions. The is not surprising that the models do not produce satisfying study of trade flows and the impact of regionalism may not answers to such questions. capture the strategic and dynamic aspects of PTAs. A coun- Does this mean that the study of PTAs and their varia- try contemplating joining a PTA does not need to know tions is pointless? Clearly, one needs to continue to evalu- whether that PTA has been trade creating or trade divert- ate the overall impact of a fragmentation of the rule system ing; the issue is whether acceding to the agreement creates in world trade and the ways in which regional and multilat- beneficial trade flows, either from better market access or eral trade can coexist and become more complementary. from reliable low-cost imports. In the case of a customs But in addition to that work, there is considerable scope for union, the height of the common tariff holds the key—for focusing on the practical issues of decision making in the some countries, accession will lead toward liberalization, area of trade policy. Such work would help countries—in Agriculture 151 particular, those with limited internal analytical capacity— The decision to leave agriculture and fisheries out of the face the challenges of the day. EFTA agreement led to the exclusion of the sector from the terms of the European Economic Area (EEA), the set of bilateral PTAs that the EU negotiated with EFTA members The Practice of Agricultural Trade in PTAs as a way of keeping them close to the EU in terms of eco- Regional PTAs have the capacity to develop strong nomic regulation and price levels. The EEA allowed for free regional agricultural systems, but the path may be politi- trade in manufactured goods and cooperation in regula- cally difficult. Bilateral PTAs are free of the problem of tory issues. In effect, it extended the previous bilateral regional competition, but they often have issues with the PTAs to several aspects of trade that had been incorporated liberalization of trade in particular products, where there into the 1992 Single Market of the EU. Although some may be links between the two economies concerned. quotas on agricultural goods were expanded, there was no Regional PTAs are considered first because most of the progress toward incorporation of the rural sector into eco- difficult questions surrounding the incorporation of agri- nomic integration, as would be stimulated by enlargement. culture arise in these cases. The EU, the European Free Later, EFTA countries found themselves unable to include Trade Association (EFTA), NAFTA, and Mercosur provide agriculture in bilateral agreements that they negotiated rich examples of the ways in which the issues have been with countries such as Canada and had to settle again for tackled. The recent growth of bilateral PTAs across regions small bilateral trade deals. offers many other cases of the treatment of agriculture. In At the other extreme, the countries that formed the EEC particular, the bilateral PTAs negotiated with the EU and (later, the European Union) made a conscious decision to with the United States represent (different) standards for include all trade, including agriculture, in their trade liber- the way in which developing countries can seek to gain alization. As integration progressed, more internal agricul- secure market access in major developed-country mar- tural trade took place, some of it displacing lower-cost kets. These bilateral agreements tend to be “lighter� in the imports. In addition, the agricultural market became more area of agricultural policy, avoiding the problems that integrated as firms were able to locate in other member accompany the development of regional agricultural and states, and a European food industry began to emerge. The food markets (Josling 2009). development of a Common Agricultural Policy (CAP), with common financing and uniform support mecha- nisms, advanced further in the EU than in other PTAs. Agricultural Provisions in Regional PTAs More recently, harmonized regulations on food safety and The inclusion of agricultural trade in a regional PTA is a quality controls have reinforced the development of a challenge for negotiators. Relatively high levels of protec- regional industry. tion in agricultural markets, combined with a heightened The polar cases of EFTA and the EU bracket the degree sensitivity to issues bearing on the maintenance of a of incorporation of agriculture in PTAs. Almost all other domestic production base for staple foods, make for ten- PTAs have included agricultural trade in the liberaliza- sions. Countries in the same region are likely to have sim- tion process, to varying degrees. The agricultural content ilar production patterns. Where there has been a history of the PTAs can be explored by identifying some issues of agricultural trade among the countries, the tensions that arise in most cases. These categories are not confined may be a minor political problem, but in many cases, to agricultural trade, but they do form a set of negotiating trade with neighbors in a regional group may raise major issues that frame the agricultural agenda. They include concerns. the schedule for cutting tariffs and the use of TRQs as a Some PTAs have chosen to omit agriculture from their way of increasing access; safeguards against import surges; provisions. EFTA was created in 1960 by seven countries subsidies to domestic firms and to firms dependent on that had opted out of the European Economic Community exports; the provision of public goods, both environmen- (EEC, the precursor of the EU). Several members (Austria, tal and related to food security; and market structures Finland, Sweden, and Switzerland) had high-cost farming and institutions. sectors because of climate or topography and did not wish Tariff cutting. Elimination of tariffs among partners is to compete directly with the United Kingdom or Denmark. the defining feature of a PTA, and the inclusion of agricul- EFTA accordingly chose to exclude agriculture (and fish- tural tariff lines in the reduction schedules is a key decision. eries) from the free trade provisions. Each country was able For some products, the tariff cuts are made at the time the to maintain its own agricultural policy through tariff pro- PTA enters into force; for others, a schedule of reductions visions and domestic support.9 is agreed on. Agricultural tariff cuts, at least for sensitive 152 Tim Josling products, are usually introduced over time. The timetable in tariffs should imports surge (see Prusa, ch.9 in this for liberalization in NAFTA provides an example. volume). Agricultural products in PTAs are often subject NAFTA set in process the removal of all trade barriers to specific safeguard provisions to help guard against sud- to goods moving between countries in North America. den shifts in trade patterns. The nature of the safeguards The detailed market access provisions were embedded in for agriculture is usually in the form of a “snapback� to a three bilateral agreements (the one between the United previous tariff, no higher than the MFN tariff rate, for a States and Canada essentially continued a previous bilat- limited period of time. Similar provisions in NAFTA were eral agreement). For the U.S.–Mexico bilateral agree- used on several occasions during the transition period to ment, the time period for most sectors to achieve market react to trade surges. Countries generally reserve the right integration was 10 years, but markets for some sensitive to take action under WTO safeguard provisions, although agricultural products (beans and corn for Mexico, and not in addition to regional safeguards. The EU is again an tomatoes and citrus products for the United States) were exception; its regulations prohibit safeguard action against given 15 years to adjust. The adjustment period has trade from another member state. ended, and the U.S.–Mexico agricultural market is now Domestic and export subsidies. The thorny issue of effectively open.10 domestic subsidies in PTAs has been dealt with in two dif- Another success in removing tariff barriers on trade in ferent ways. Generally, the decision is made in negotiations farm products has been in the Australia–New Zealand not to attempt any constraints on subsidies. Indeed, it is regional market.11 The Australia–New Zealand Closer Eco- usually assumed that PTAs could not regulate domestic nomic Relations Trade Agreement (ANZCERTA) takes the subsidies because to do so on inter-PTA trade but not on two countries further toward effective market integration extra-PTA sales would be impractical at best and self- than does NAFTA in North America, although by no defeating at worst. The emphasis in several PTAs is, accord- means as far as the EU. Both countries are major agricul- ingly, on acknowledging the multilateral process as the tural exporters. The product mix of exports is somewhat location of decisions on subsidy reduction. Hence, NAFTA similar, reducing the scope for trans-Tasman trade, but contains the injunction to “endeavor to work towards there are natural trade flows based on climatic differences, domestic support measures� that have minimal trade- such as sales of Australian wheat to New Zealand and distorting effects or that would be exempt under a future exports of New Zealand dairy goods to Australia. These GATT agreement (the so-called “green box� policy instru- flows were hampered by tight restrictions on trade within ments). It recognizes, however, the right of parties to change domestic marketing legislation. It took bold political deci- domestic support measures subject to GATT obligations. sions, coupled with a significant reduction in the power of This light treatment of a contentious area enabled nego- the marketing agencies, to allow trade in agricultural prod- tiators to say that they were not altering domestic policy. ucts to flow freely. The EU took a different tack. All subsidies by member In most respects, tariff cutting in agricultural markets states are constrained by the competition regulations of the has been successful in Latin American PTAs. Among Mer- EU, and farm subsidies are not excluded from this provi- cosur countries, agricultural trade is nominally free; sion. The CAP became (in principle) the only vehicle for indeed, agricultural products are widely traded among the granting agricultural subsidies, although some exceptions member states, notably from Argentina to the others. Mer- have survived the attempts by the European Commission cosur has relatively few provisions that apply specifically to to enforce this regulation. One result of the common agriculture. There are two likely reasons for this relatively nature of the CAP has been that the EU can negotiate liberal treatment of the sector. First, Mercosur includes reductions in domestic support as a single WTO member, major exporters of temperate agricultural products, each which individual members of other PTAs are not in a posi- of which would like to strengthen its agriculture industry tion to do. and promote regional exports. Second, as a result of sweep- Export subsidies for agricultural products pose similar ing structural reforms, the countries concerned have elimi- issues. Several PTAs have contemplated banning export nated many of the state marketing monopolies that previ- subsidies on intrabloc trade, but this is easier said than ously controlled trade.12 This, together with the reduction done. The NAFTA provisions again give a good example of of subsidies and support prices, has allowed a fuller incor- the dilemma facing PTA negotiators. The text states that poration of agriculture within Mercosur than in many parties “share the objective of the multilateral elimination other PTAs. of export subsidies for agricultural goods� and promises Safeguards. As a complement to tariff cutting, PTAs fre- cooperation in the GATT to this end. The zeal for multilat- quently include safeguards that allow temporary increases eral elimination of such policies does not, however, extend Agriculture 153 to their internal use. Article 705.2 of NAFTA merely holds form the basis for improvement of the economic capacity it “inappropriate� for a party to grant export subsidies on of an agricultural area. sales to another party unless the importing country is ben- One particular public good associated with agriculture efiting from export subsidies paid by other countries.13 In deserves separate mention. Food security refers to the abil- other words, matching of EU export subsidies in Mexico is ity of a country to provide the conditions under which allowable by the United States and Canada until such food is available to (and relatively affordable by) the popu- practices are stopped multilaterally. Indeed, if the export- lation. Economic, social, and political imperatives converge ing and importing parties agree to an export subsidy on here. The contribution of PTAs to the attainment of this intra-NAFTA trade, that subsidy is allowed. This provi- objective is generally positive, as discussed earlier, but the sion, no doubt, was included to take account of the con- issue does pose some challenges for negotiators. A balance siderable importance to the United States of retaining the has to be struck between the benefits of open trade for the means to stay competitive with EU export subsidies in the regionwide sharing of risk and the ultimate national Mexican market. responsibility for ensuring food supplies. The issue that PTA discussions about the impact of different market- may cause regional friction is whether a partner in a PTA ing structures and institutions raise some interesting may restrict supplies to another partner when its own sup- issues. This is particularly true for state marketing insti- plies are scarce. The stronger PTAs, with regional food tutions in agricultural products, where historical differ- markets and coordinated policy reactions, will tend to ences in policies can lead to problems for integration. An restrain the ability of one country to impose an export ban early example was the difficulty posed for the EU at the on a partner, whereas the weaker agreements tend to leave time of U.K. accession (1973) by the existence in England this possibility open. and Wales of the Milk Marketing Board (MMB), which Many PTAs include provisions that relate to the health held a monopoly on milk sales and on imports of milk and safety aspects of agricultural and food trade, such as products. This situation was clearly inconsistent with the the harmonization (or the mutual recognition of) health competition regulations of the EU, and so the MMB had and safety regulations. (See, in this volume, Maur and to change its policies and give up its control over the milk Shepherd, ch. 10, and Stoler, ch. 11, on standards in PTAs.) market. A more recent example appears in the 1986 Most such provisions are based on the WTO Sanitary and U.S.–Canada free trade agreement. Canada was able to Phytosanitary Agreement and do not require members to exclude from the free trade provisions the products of its go far beyond those standards. In some cases, however, supply-managed sectors, primarily dairy and poultry, such as ANZCERTA, the establishment of joint agencies which were managed largely by provincial marketing to oversee such regulations acts as a guard against trade boards. Neither the United States nor Canada wished to frictions (Almeida, Gutierrez, and Shearer 2009). face the task of harmonizing marketing systems for these Institutions. Institutional innovations are also com- products, and it was felt that the operation of the boards mon, although some of the bodies set up seem to have required control of all imports, including those from the little role in policy decisions. The NAFTA trilateral agri- United States. As a result, the integration of these sectors cultural agreement, for instance, set up a Committee on was delayed indefinitely. Agricultural Trade to administer the arrangements and an Public goods. It is widely recognized that agriculture Advisory Committee on Private Commercial Disputes provides certain public goods (as well as negative exter- regarding Agricultural Goods to deal with private dis- nalities in the way of water and soil pollution). In rich putes. But there is little evidence that these bodies have countries, these public goods are often identified as the had any significant impact on agricultural trade policy stewardship of the landscape and the provision of locally over the 15 years of their existence. grown healthy foods; in less affluent societies, the benefits are food security, rural development, and poverty allevia- Agricultural Provisions in Bilateral PTAs tion. Whether society is adequately compensating the farm sector for the provision of these public goods is a The treatment of agriculture in bilateral PTAs is often subject of debate in many countries. The collective provi- markedly different from that in regional agreements.14 sion of a public good within a PTA can sometimes be The motivation for such PTAs ranges from strategic to advantageous, in particular where agricultural and envi- practical, but most often it is the exchange of preferential ronmental conditions are defined more by geography and access for goods and services, with little regard for the climate than by political boundaries. Similarly, coordina- longer-run economic relationship. No regional integra- tion of rural infrastructure within a regional PTA could tion of the agriculture sectors is envisaged, and many of 154 Tim Josling the tensions around farm policies that occur in regional liberalization road map has been defined for the agricul- pacts are absent. The main characteristics of bilateral ture sector as a whole; only for certain products have spe- PTAs are usually determined by the dominant partner, cific concessions for liberalization been determined. A often a developed country, and the discussion of these concern for the non-EU Mediterranean countries is that, PTAs is therefore conveniently organized according to with the conclusion of PTAs between the EU and other the dominant partner—in these examples, the EU and countries in Asia and Latin America, the competitive the United States. advantage that they themselves used to enjoy in EU mar- EU agreements. In the network of agreements involving kets may be eroded, and they may become marginalized. the EU and nonmembers, agriculture is still treated as The deferral of substantive negotiations on liberaliza- being largely outside the realm of unrestricted free trade. tion of trade in agricultural products has been a constant The Euro-Mediterranean free trade agreements (Euromed feature of the Euro-Mediterranean partnership (Asbil FTAs) now being finalized between the EU and the coun- 2005). The principal reason has been the reluctance on the tries of North Africa and the Middle East have so far part of European farmers to compete with Mediterranean avoided including unrestricted access for sensitive agricul- countries that are not EU members. The southern enlarge- tural products, and the same is true for the customs union ment of the EU in the 1980s redefined its relation with the that was negotiated with Turkey. The negotiation of a free Mediterranean partners. Greece, Portugal, and Spain com- trade agreement between the EU and South Africa was pete directly in agricultural products with the countries of held up by the reluctance of the EU to grant improved North Africa, and these members’ political influence access to goods that would have directly competed with largely explains the limits on trade concessions through those covered by the CAP. The agreement between the EU tariff quotas and reference quantities (García-Alvarez- and Mexico was also difficult to negotiate until Mexico Coque 2002). Agriculture has become a key sector in the abandoned its attempt to win easy access to the EU market debate between the EU and its Mediterranean trade part- for a full range of agricultural products. ners because it is seen as a necessary element in the Similarly, Mercosur and the EU are finding it difficult establishment of a balance of commercial opportunities to overcome the problems that improved access to the EU through increases in both industrial and agricultural market would seem to pose for European agriculture. The exports from the region. Cotonou Agreement between the EU and ACP countries, The other problem that Mediterranean countries need which mandated the negotiation of a transformation of to consider is the shortcomings in the diversification and the existing nonreciprocal agreements into full free trade competitiveness of their production structures. Several areas after eight years, attempted to address agricultural Mediterranean countries have very similar product com- trade issues, but these negotiations were hampered by positions of exports. Algeria, Cyprus, Israel, Morocco, and inconsistency with the CAP. The unilateral PTA between Spain all have agriculture sectors oriented toward specialty the EU and the least developed countries (the Everything products—mainly, fresh fruit and nuts, olive oil, and wine. But Arms agreement) broke significant new ground in The similarity of agricultural products means that, espe- this respect by providing duty-free and quota-free access cially since the enlargement of the EU to include Cyprus, for agricultural goods, with only temporary derogations Greece, Malta, Portugal, and Spain, EU members find it for the most sensitive commodities—rice, sugar, and easy to replace supplies from nonmember Mediterranean bananas. countries with supplies from EU members. Traditionally, the EU has used the policy of trade pref- The current negotiations between the EU and the ACP erences as a strategy of cooperation for development and countries (the signatories to the Lomé and Cotonou Agree- has unilaterally granted trade concessions to other coun- ments) have advanced through six regional talks. The EU tries. Now, Euromed agreements are taking further steps has succeeded in agreeing on a comprehensive economic toward trade liberalization on a bilateral and reciprocal partnership agreement (EPA) with Caribbean ACP coun- basis. Since the first Euro-Mediterranean Conference in tries through CARIFORUM and the Caribbean Commu- November 1995, the EU and 12 Mediterranean countries nity (CARICOM) Regional Negotiating Mechanism.15 have been engaged in negotiating association agreements In the case of the African countries, negotiations are (the Barcelona process). The overall objective is to form, being channeled through four of the main regional agree- eventually, a single Euro-Mediterranean free trade area ments. The Economic Community of West African States from the separate agreements in place. Yet, trade in agri- (ECOWAS), in collaboration with the West African Eco- culture is subjected to weak liberalization within the pres- nomic and Monetary Union (WAEMU), is the negotiating ent framework of the association agreements. No explicit partner for 16 West African states. Eight Central African Agriculture 155 states have been negotiating through the Economic and was concluded in 2001, again as a show of political support Monetary Community of Central Africa (CEMAC, Com- and economic assistance. munauté Économique et Monétaire de l’Afrique Centrale) The United States began to negotiate additional bilateral in conjunction with the Economic Community of Central PTAs in 2002 as an expression of a policy of “competitive lib- African States (ECCAS), which CEMAC has plans to merge eralization� articulated by the U.S. trade representative. This with. Seven Southern African states are negotiating policy consisted of offering swift negotiations to any country through the Southern African Development Community that was willing to conform to terms consistent with the (SADC), although some of those states are not SADC mandate of the U.S. administration, as specified in the Trade members. Another 15 are represented by the Common Promotion Authority Act. The list of willing trade partners Market for Eastern and Southern Africa (COMESA), even with which PTAs were concluded includes Bahrain, though some of them do not participate in other COMESA Morocco, Oman, Peru, and Singapore. Among other com- activities. With the exception of the agreement with the pleted bilateral PTAs with a more significant agricultural Caribbean, the EPAs are still not fully in operation. Some component were those with Chile and Australia. Talks with countries have signed partial (“goods only�) agreements, Bolivia, Ecuador, Peru, the Southern African Customs but more than half of the ACP countries failed to reach an Union, and the United Arab Emirates (UAE) are currently agreement before the January 1, 2008, deadline, when the suspended.18 Agreements with Colombia, the Republic of WTO waiver that allowed the EU to negotiate these agree- Korea, and Panama await ratification. A new front has been ments expired. Renewal of the waiver would encounter opened up in the Asia-Pacific region as the United States has some opposition. Countries that have shown opposition to begun to explore the possibility of a Trans-Pacific Partner- the EPAs include South Africa, which already has a free ship (TPP) agreement to include Australia, Brunei Darus- trade agreement with the EU, and Nigeria, with its oil- salam, Chile, New Zealand, Singapore, and possibly Malaysia based economy. A bold move by China to develop trade and Thailand. Recent agreements have often been designed and investment links with African countries appears to be as “templates� for future PTAs within a region. Thus, the causing a rethinking of the desirability of continuing close PTAs with Bahrain, Oman, and the UAE are seen as building ties with the EU if those ties come with political con- blocks toward a Middle East free trade area, and the negotia- straints. tions with Malaysia and Thailand (along with the one U.S. agreements. U.S. policy toward regional and bilat- already in place with Singapore) were originally supposed to eral PTAs changed dramatically in the mid-1980s. Long a pave the way for other bilateral PTAs with countries in the champion of the multilateral system and of nondiscrimi- Association of Southeast Asian Nations—although this nation, the United States has now become an active prospect has been overtaken by the TPP. The PTAs them- supporter of bilateral PTAs as a complement to its commit- selves usually follow from trade and investment framework ment to the WTO and its membership in NAFTA. The agreements (TIFAs) and bilateral investment treaties (BITs). United States has completed, or is currently in the midst of, The United States has a considerable number of TIFAs and trade negotiations with 27 countries aimed at creating BITS in place that would form the basis for bilateral PTAs. about 20 separate PTAs.16 The United States has economic Although all the PTAs have provisions for tariff reduc- and geopolitical reasons for expanding its commercial ties; tions that affect many food and agricultural goods, the the attraction for other countries is to secure preferred agreements, with few exceptions, control trade in a range of access to the large U.S. market.17 products considered politically sensitive in one or both The first of these recent PTAs was signed with Israel in partners. For the United States, these sensitivities include 1985 as an expression of political and economic support sugar, citrus fruits, peanuts, and dairy products; for the for that country. The free trade agreement with Canada partners, the list includes corn, beans, and rice. followed in 1986, largely at Canada’s request. It was Three current agreements have the greatest actual or designed to consolidate existing sector agreements, potential impact on U.S. agricultural markets and hence on encourage U.S. investment north of the border, and give the environment in which policy is formed: the recent Canadian firms some protection from aggressive use of PTAs with Chile and Australia, and the CAFTA–DR agree- trade remedy provisions (i.e., antidumping and counter- ment. Table 7.1 summarizes the main characteristics of vailing duty measures). In 1990, Mexico requested similar each agreement. conditions, to ensure overseas investors’ access to the large The United States and five Central American countries— U.S. market. Canada opted to join the United States and Costa Rica, El Salvador, Guatemala, Honduras, and Mexico in NAFTA, which incorporated the earlier bilateral Nicaragua—began negotiations on CAFTA in 2003, and agreement with Canada. A free trade agreement with Jordan the agreement took effect in 2006. Negotiations with the Table 7.1. Summary of Provisions Affecting Agriculture in NAFTA, U.S.–Chile, U.S.–Australia, and CAFTA Agreements Provision NAFTA Chile FTA Australia FTA CAFTA 156 Tariff cuts Some tariffs eliminated; others staged Some tariffs eliminated; others staged Most tariffs eliminated; others staged Some tariffs eliminated; others staged over 5, 10, and 15 years over 4, 8, 10, and 12 years; some cuts over 4, 10, and 18 years over 5, 10, and 15 years; other cuts delayed for 2 and 4 years delayed for 6 or 10 years; duty-free status after 15 or 20 years TRQs TRQs introduced during transition No use of TRQs introduced TRQs for imports of avocados, cotton, TRQs for sensitive products in Annex 3.3 period for sensitive products peanuts, tobacco, beef, and dairy imposed; rules on administration of products into the United States TRQs (in addition to GATT Article XIII) expanded; above-quota duties for beef put in place phased out over 18-year period; current sugar TRQs not increased: no cuts in above-quota tariff Agricultural TRQs allowed as special safeguard for Additional duties linked to price trigger Additional customs duties linked to Additional duties linked to trigger safeguards horticultural crops (Annex 703.3) (Article 3.18) for goods listed in Annex price trigger for horticultural products quantities (Article 3.14) for products 3.18; total duties not to exceed MFN (Annex 3-A, section A) and to quantity listed in Annex 3.14; total duties not to rate; safeguard not operative after 12 triggers for beef (Annex 3-A, section B); exceed MFN rate; safeguard not years, or when zero-duty stage reached price triggers used for beef in 19th year operative when zero-duty stage reached of agreement (Annex 3-A, section C); safeguard not operative when zero-duty stage reached Other Safeguards (Chapter 8A) stipulated: Trade remedies (Chapter 8) stipulated; Safeguards (Chapter 9) stipulated; GATT Trade remedies (Chapter 8) stipulated; safeguards snapback to previous year’s tariff on GATT 1994 Article XIX safeguards 1994 Article XIX safeguards adhered to GATT 1994 Article XIX safeguards bilateral trade or MFN tariff adhered to adhered to Export subsidies Agreement reached to avoid use of Agreement reached to avoid use of Agreement reached to avoid use of Agreement reached to avoid use of export subsidies on bilateral trade unless export subsidies on bilateral trade unless export subsidies on bilateral trade unless export subsidies on bilateral trade unless third countries subsidized exports to third countries subsidized exports to third countries subsidized exports to third countries subsidized exports to NAFTA markets; agreement reached to Chile or the United States; agreement Australia; agreement reached to work CAFTA markets; agreement reached to work together for elimination in the reached to work together for together for elimination in the WTO work together for elimination in the GATT elimination in the WTO WTO Domestic Agreement reached to work together in Agreement reached to work together in Agreement reached to work together in Agreement reached to work together in support GATT for reduction of domestic support WTO for reduction of domestic support WTO for reduction of domestic support WTO for reduction of domestic support levels and to shift to less-trade-distorting levels and to shift to less-trade-distorting levels and to shift to less-trade-distorting levels and to shift to less-trade-distorting instruments instruments instruments instruments SPS measures Precursor of WTO SPS agreement Affirm commitment to SPS agreement Work to resolve trade conflicts over Affirm commitment to SPS agreement (Chapter 7B) SPS barriers Dispute Dispute settlement mechanism for Dispute settlement mechanism for Dispute settlement mechanism for Dispute settlement mechanism for settlement matters arising from agreement (Chapter matters arising from agreement matters arising from agreement matters arising from agreement put in 20) put in place; separate procedures for (Chapter 22) put in place (Chapter 21) put in place; provision for place review of antidumping and countervailing monetary penalties put in place actions (Chapter 19) put in place Institutions Committee on Agricultural Trade; Working Group on Agricultural Trade; Committee on Agriculture; Standing Committee on Agricultural Trade; Working Group on Agricultural Committee on Sanitary and Technical Working Group on Animal Agricultural Review Commission; Subsidies; Advisory Committee on Phytosanitary Matters and Plant Health Measures Committee on Sanitary and Private Commerce Disputes Regarding Phytosanitary Matters Agricultural Goods Source: Author’s compilation from texts of agreements. Notes: CAFTA = Central America Free Trade Agreement; FTA = free trade agreement; GATT = General Agreement on Tariffs and Trade; MFN = most favored nation; NAFTA = North American Free Trade Agreement; SPS = sanitary and phytosanitary; TRQ = tariff-rate quota; WTO = World Trade Organization. Citations of specific provisions refer to the respective agreement. Agriculture 157 Dominican Republic that would fully integrate that country exporter of agricultural products, particularly fruits, veg- into CAFTA were concluded in 2004. In addition, ratifica- etables, and wine, but the different seasonality makes the tion of the pending agreement with Panama, if successful, produce complementary to rather than competitive with would round off the establishment of free trade agreements U.S. production. The beneficiaries were supermarkets, between the United States and almost all of the countries of which gained the assurance of year-round supplies. Chile is Central America (see Paggi, Yamazaki, and Josling 2005). one of the more liberal Latin American countries, even on CAFTA is intended to help foster economic growth and agricultural products, so that opening up to U.S. exporters improve living standards in the Central American region was not such a big move for its farmers. Aside from some by reducing and eliminating barriers to trade and invest- controversy over wine labels, the talks went smoothly. It ment. It essentially converts the nonreciprocal and discre- may have helped that Chile is not a significant sugar tionary benefits that these countries derive from the exporter. Caribbean Basin Initiative (CBI) into permanent and The U.S. agreement with Australia also involved a reciprocal access to the U.S. market. CAFTA covers all Southern Hemisphere country and thus offered some trade, but the agricultural component is one of its most advantages of complementary production. Australia, how- important aspects (see table 7.1). The key to the agricul- ever, is a major exporter of meats, dairy products, cereals, tural agreement is market access; the arrangement contains and sugar, and so tight rules had to be built in to the agree- relatively few provisions in the areas of export subsidies ment to protect U.S. farmers from competition from and sanitary and phytosanitary regulations, and it does not imports. Reluctantly, Australia accepted long transition cover domestic subsidies. periods for dairy products and beef, and an exclusion alto- Agricultural trade barriers in the Central American gether of any relaxation of protection for the U.S. sugar countries are higher than those for manufactured goods, sector. This decision may have an effect on the politics of and CAFTA will create improved market opportunities for future bilateral PTAs. U.S. agricultural products and for related goods and serv- The most ambitious bilateral agreement to have been ices. CAFTA locks in the applied duty rates for many prod- negotiated since NAFTA is still awaiting approval by Con- ucts and ensures that permanent U.S. access to the market gress at present. The Korea–U.S. Free Trade Agreement is preserved. Its short-term impact on U.S. exports may, (KORUS) would establish a free trade area between the however, be modest because the terms of the agreement United States and a major economy in East Asia. There was delay the full benefits of increased access to the countries of no doubt from the beginning of the talks that agriculture the region for U.S. agricultural products of interest. The would be a stumbling block, with the Korean government, lengthy phase-in period for increased market access and in particular, not wishing to open up its highly protected the back-loading of commitment levels suggest that the rice market to U.S. exports. The U.S. position had been to benefits of the agreement may only be realized many years include rice, even if access for U.S. rice were to be intro- in the future. duced slowly over a transition period. In the end, the rice Increased access to the U.S. market for Central Ameri- sector was essentially excluded from the agreement. This can goods will also be a consequence of CAFTA. Here, establishes a precedent in case KORUS were to act as a tem- however, the effect is likely to be even more limited plate for an agreement with Japan. because most CAFTA countries have had permanent duty- free access to the U.S. market since the late 1960s under Conclusions the GSP and since 1990 under the provisions of the Caribbean Basin Initiative (CBI) and the Caribbean Basin Agricultural trade is becoming increasingly governed by Economic Recovery Act (CBERA), which implements the conditions negotiated in preferential trade agreements CBI. The CBI was enhanced in 2000 under the terms of (PTAs), whether regional or bilateral. Regional integration the Caribbean Basin Trade Partnership Act (CBTPA) to of agricultural markets through open trade can have a pos- grant access more equivalent to that enjoyed by Mexico itive effect on the development of a competitive and sus- under NAFTA. In fact, approximately 99 percent of tainable agriculture sector, although complementary poli- CAFTA exports already enter the U.S. market duty-free. cies at the multilateral level are needed to prevent trade Duties are paid only on over-quota imports, as part of the diversion. PTAs can yield benefits in the area of food secu- U.S. tariff-rate quota regimes for sugar, dairy products, rity and the provision of public goods, but the empirical cotton, meats, and peanuts.19 analyses required to quantify these benefits are scarce, in The bilateral PTA with Chile was easier to negotiate part because of the very diverse treatment of agriculture in than either NAFTA or CAFTA. Chile is an important PTAs. If concluded with the right partner countries, PTAs 158 Tim Josling can avoid the disadvantages of trade diversion. If the terms oilseed exports remained, together with a fee to pay for research. Argentina has essentially liberalized imports of agricultural goods, of the agreements are appropriate, such PTAs can further although some export taxes have reappeared in recent years. full integration into the global economy and stimulate 13. The language about and treatment of export subsidies is much needed investment and the transfer of technology. softer in NAFTA than in the U.S.–Canada Free Trade Agreement, which banned the use of export subsidies between the two countries. 14. Bilateral PTAs involving Singapore do not face the same political Notes problems in including agriculture and food products as do most other agreements. Singapore has no significant agricultural production and no 1. A full account of the interpretation of GATT Article XXIV is pro- import barriers. Other countries, however, may be concerned about the vided in Hudec and Southwick (1999). Srinivasan (1998) presents a possibility of trade deflection through Singapore. Japan, for instance, has critique of the systemic issues posed by PTAs. A type of agreement not not agreed to open its agricultural market to imports from Singapore in considered in this chapter is the partial-scope agreement, which covers their bilateral free trade agreement. only a subset of goods. These agreements are not generally notified under 15. CARIFORUM includes the Dominican Republic, as well as the Article XXIV. CARICOM countries. 2. It is perhaps ironic that the insistence on both the inclusion of 16. It is worth recalling that in the late 1930s, U.S. trade policy took a substantially all trade and 100 percent preference increases the likelihood similar direction. The Reciprocal Trade Agreements Act of 1934 was an of trade diversion. Exclusion of those sectors where protection is high, open-ended mandate to negotiate bilateral trade agreements with other and the partial liberalization of internal trade in those high-cost sectors countries, and about 30 such agreements were signed. that are covered, would reduce trade diversion, although it would also 17. In many cases, access is already covered by existing agreements, limit trade creation. The strict rules were presumably inserted to discour- but the negotiation of a formal PTA reduces uncertainty as to whether age trade agreements that merely “picked low-hanging fruit� by confining these preferences will continue. themselves to nonsensitive sectors or small reductions in tariffs. 18. SACU is made up of Botswana, Lesotho, Namibia, South Africa, 3. For a discussion of the benefits of coordinating trade strategies and Swaziland. for small countries, see Schiff (1997, 2002); Andriamananjara and Schiff 19. For more details on CAFTA and its potential impact on U.S. agri- (1998). culture, see Paggi, Yamazaki, and Josling (2005). 4. Adjustment of market access provisions is made easier by the negotiation in the Uruguay Round of tariff-rate quotas (TRQs) that ensure continued access for those countries with negotiated quotas. Allo- References cation to preferred partners of TRQs allowed in WTO schedules is a prin- cipal link between the multilateral and preferential trade regimes. These Almeida, Juliana Salles, Carlos M. Gutierrez, Jr., and Matthew Shearer. allocations appear contrary to the spirit of Article XIII of the GATT, 2009. “The Treatment of Agriculture in Regional Trade Agreements in which provides that the distribution of quotas should be nondiscrimina- the Americas.� Integration and Trade Sector, Inter-American Develop- tory and should reflect market conditions. ment Bank, Washington, DC. 5. Analytical devices such as that applied here are sometimes difficult Anderson, Kym, and Richard Blackhurst, eds. 1993. Regional Integration to reconcile with a world of actual countries and diverse trade strategies. In and the Global Trading System. New York: St. Martin’s Press. fact, some new members may be willing to pay the price for entry, and Andriamananjara, Soamiely. 2002. “On the Size and Number of Preferen- some old members may have strong reasons for excluding newcomers. tial Trading Arrangements.� Journal of International Trade and Eco- 6. The “completion� of the European Community’s (EC’s) internal nomic Development 11 (3): 279–95. market reform, which was incorporated in the 1992 program, was the Andriamananjara, Soamiely, and Maurice Schiff. 1998, “Regional Group- product of an attempt to remove remaining barriers between EC mem- ings among Microstates.� Policy Research Working Paper 1922, World bers in goods, services, labor markets, and capital markets. The success of Bank, Washington, DC. http://econ.worldbank.org/external/default/ the European Commission in pushing through this program caused a stir main?pagePK=64165259&theSitePK=469372&piPK=64165421&menu of interest in other regions. PK=64166093&entityID=000009265_3980625102911. 7. The inclusion of APEC in this list is of interest, as no preferential Asbil, Alexandre. 2005. “The Barcelona Process and Trade Liberalisation in tariff reductions were specified in that agreement. APEC’s modus Agriculture.� Presentation to conference of the Food and Agriculture operandi was and is a loosely coordinated process of unilateral liberaliza- Organization of the United Nations (FAO), Rome, October 10–14. tion among its members, with no discrimination against outsiders. Brenton, Paul, and Takako Ikezuki. 2005. “The Impact of Agricultural 8. This analysis also points up the importance of timing in trade Trade Preferences, with Particular Attention to the LDCs.� In Global agreements. The advantage of being first to reach a deal may be consider- Agricultural Trade and Developing Countries, ed. M. Ataman Askoy able. And the negotiation of long transition periods may be of little value and John C. Beghin. Washington, DC: World Bank. in easing adjustments if other countries can enter into agreements with a dell’Aquila, Crescenzo, and Marijke Kuiper. 2003. “Which Road to Liber- shorter time horizon. alisation? A First Assessment of the EuroMed Association Agree- 9. A side agreement guaranteed access for Denmark to the British ments.� Working Paper 2, European Network of Agricultural and market for pigmeat products and for limited quantities of dairy goods. Rural Policy Research Institutes (ENARPRI), Centre for European 10. Canada declined to join the open market between Mexico and the Policy Studies, Brussels. United States because that would have required liberalization of the sup- Diao, Xinshen, Terry Roe, and A. Somwaru. 2001. “What Is the Cause of ply-managed sectors (dairy products and poultry), as well as sharply Growth in Regional Trade? Trade Liberalization or RTAs? The Case of increased prices for sugar in Canada. Agriculture.� The World Economy 24(1): 51–79. 11. The original ANZCERTA suffered from a lack of ambition and Fiorentino, Roberto. 2005. “Regional Trade Agreements and the WTO: political support (Lloyd 1991). A 1988 review of its operations led to an Theory and Practice of WTO Provisions on RTAs.� Presentation to acceleration of the process of liberalization, with July 1990 set as the date conference of the Food and Agriculture Organization of the United for the removal of remaining tariff and nontariff barriers. In addition, Nations (FAO), Rome, October 10–14. export subsidies on intrabloc trade were to be removed. García-Alvarez-Coque, José María. 2002. “Agricultural Trade and the 12. In 1991, Argentina eliminated export taxes on agricultural goods, Barcelona Process: Is Full Liberalization Possible?� European Review of which had long been a source of government revenue. A small tax on Agricultural Economics 29 (3): 399–422. Agriculture 159 Goto, Junichi. 1997. “Regional Economic Integration and Agricultural Trade.� Paggi, Mechel S., Fumiko Yamazaki, and Tim Josling. 2005. “The Central Policy Research Working Paper 1805, World Bank, Washington, DC. American Free Trade Agreement: What’s at Stake for California Agri- http://www-wds.worldbank.org/servlet/WDSContentServer/WDSP/IB/ culture?� Center for Agricultural Business, California State University, 2000/02/24/000009265_3971110141344/Rendered/PDF/multi_page.pdf. Fresno, CA. Grant, J., and Tom Hertel. 2005. “Regionalism in World Agricultural Schiff, Maurice. 1997. “Small Is Beautiful: Preferential Trade Agreements Trade: Lessons from Gravity Model Estimation.� Global Economic and the Impact of Country Size, Market Share, Trade Policy, and Analysis conference, Lübeck, Germany, June. Smuggling.� Journal of Economic Integration 12, no.3 (September): Hudec, Robert, and James D. Southwick. 1999. “Regionalism and the 359–87. WTO Rules: Problems in the Fine Art of Discriminating Fairly.� In ———. 2002. “Regional Integration and Development in Small States.� Trade Rules in the Making: Challenges in Regional and Multilateral Development Research Group, Trade, World Bank, Washington, DC. Negotiations, ed. Miguel Rodríguez Mendoza, Patrick Low, and Barbara Srinivasan, T. N. 1998. “Regionalism and the WTO: Is Nondiscrimination Kotschwar. Washington, DC: Brookings Institution Press. Passé?� In The WTO as an International Organization, ed. Anne Jayasinghe, Sampath, and Rakhal Sarker. 2008. “Effects of Regional Trade O. Krueger, 329–49. Chicago, Il: University of Chicago Press. Agreements on Trade in Agrifood Products: Evidence from Gravity Topp, Vernon. 2001. “Trade Preferences: Are They Helpful in Advancing Modeling Using Disaggregated Data.� Review of Agricultural Econom- Economic Development in Poor Countries?� Australian Bureau of ics 30 (1): 61–81. Agriculture and Resource Economics (ABARE), Canberra. Josling, Tim. 2009. “Agriculture.� In Commentary and Analysis. Book 1 of UNCTAD (United Nations Conference on Trade and Development). Bilateral and Regional Trade Agreements: Commentary, Analysis and 2001. “Trade Negotiation Issues in the Cotonou Agreement: Case Studies, ed. Simon Lester and Bryan Mercurio. Cambridge, U.K.: Agriculture and Economic Partnership Agreements.� UNCTAD, Cambridge University Press. Geneva. Lloyd, Peter. J. 1991. “The Future of CER: A Single Market for Australia Vollrath, Thomas L. 1998. “RTAs and Agricultural Trade: A Retrospective and New Zealand.� Committee for Economic Development of Australia Assessment.� In Regional Trade Agreements and U.S. Agriculture, ed. (CEDA) and Institute for Policy Studies, Wellington. Mary Burfisher and Elizabeth Jones. Agricultural Economic Report ———. 1992. “Regionalisation and World Trade.� OECD Economic Studies 771. Economic Research Service, U.S. Department of Agriculture, 18 (Spring). Organisation for Economic Co-operation and Development Washington, DC. (OECD), Paris. Zissimos, Ben, and David Vines. 2000. “Is the WTO’s Article XXIV Olarreaga, Marcelo, and Çaglar Özden. 2005. “AGOA and Apparel: Who a Free Trade Barrier?� CSGR Working Paper 49/00, Centre for Captures the Tariff Rent in the Presence of Preferential Market the Study of Globalisation and Regionalisation, University of War- Access?� World Economy 28, no.1 (January): 63–77. wick, U.K. 8 Preferential Rules of Origin Paul Brenton Preferential rules of origin are applied by countries that various elements in the production of a good. The issue offer certain trade partners zero-duty or reduced-duty becomes which of these stages of production defines the access for their imports as a means of determining the eli- country of origin of the good. This chapter argues that com- gibility of products to receive such preferential access.1 plex rules of origin, which differ across countries and agree- These rules of origin are required to prevent trade deflec- ments, can be a significant constraint on trade, a substantial tion or simple transshipment, whereby products from burden on customs, and a hindrance to trade facilitation. nonpreferred countries are redirected through a free trade The nature of the rules of origin can undermine the stated partner to avoid the payment of customs duties. They are intentions of preferential trade agreements. meant to ensure that only goods originating in participat- The next section discusses what is meant by “origin� and ing countries enjoy duty preferences. Rules of origin are examines methods for determining where a substantial thus integral to preferential trade agreements such as transformation of a product takes place. The second sec- bilateral and regional free trade agreements and to the tion elaborates on the definition of preferential rules of nonreciprocal preferences that industrial countries offer origin and looks at the rules of origin in existing preferen- to developing countries. tial trade agreements. The third reviews the economic The nature of rules of origin and their application can implications of rules of origin, examines the links between have profound implications for trade flows and for the rules of origin and the use of trade preferences, estimates work of customs authorities. Rules of origin can be the costs of complying with rules of origin, and analyses designed in such a way as to restrict trade and therefore can the use of rules of origin as a tool for economic develop- be used, and are being used, as instruments of trade policy. ment. The final section provides some conclusions. The proliferation of free trade agreements throughout the world, with the accompanying preferential rules of origin, Defining Origin is increasing the burden on customs services in many countries, with implications for the ease of trade. Perhaps When a product is produced in a single stage or is wholly surprisingly, given their potential to influence trade flows, obtained in one country, such that there are no imported preferential rules of origin are one area of trade policy that components, the country origin of the product is relatively has been subject to very little discipline during the almost easy to establish. This applies mainly to “natural products� 50 years of the multilateral rules-based system now gov- and to goods made entirely from them. Proof that the erned by the World Trade Organization (WTO). product was produced or obtained in the preferential trade Determining the country of origin of products has partner is normally sufficient. For all other cases in which become more difficult over the past four decades as techno- two or more countries have taken part in the production of logical change, declining transport costs, and the process of the good, the rules of origin define the methods for ascer- globalization have led to the splitting up of production taining in which country the particular product has chains and the distribution to different locations of the undergone sufficient working or processing or has been 161 162 Paul Brenton subjected to a substantial transformation. (In general, • Under certain conditions, the EU will accept vessels these terms can be used interchangeably.) A substantial chartered or leased by the ACP state under the Cotonou transformation is one that conveys to the product its essen- Agreement. tial character. Unfortunately, there is no simple and stan- • Under Cotonou, the master, the officers, and 50 percent dard rule of origin for identifying the “nationality� of a of the crew must be nationals of any ACP state or the EU. product. Although rules relating to products that are wholly So, identifying the nationality of fish can be a demand- obtained are usually relatively straightforward, this is not ing task! More important, these rules have important eco- always the case. A good example of how rules for products nomic implications for countries that wish to attract for- that appear to be in their natural state, and are therefore eign direct investment (FDI) into their fisheries sectors and apparently wholly obtained, can become complex and for small island economies that may have great difficulty in restrictive is the case of fish—typically, a sensitive sector meeting the crew and officer requirements. in many countries. There is no apparent import content of fish caught in the sea or ocean, yet the European Union (EU) rules for fish caught outside a country’s territorial Methods for Determining Substantial Transformation waters but within the exclusive economic zone of a coun- The three main criteria for determining origin are change try can be very complex and difficult to satisfy and to of tariff classification, value added, and specific manufac- prove compliance with. To receive preferential access to turing process. We discuss each in turn. the EU under the generalized system of preferences Change of tariff classification. Origin is granted if the (GSP), all of the following conditions relating to the ves- exported product falls into a different part of the tariff clas- sel that catches the fish and the crew that sails on it must sification from any imported inputs that are used in its pro- be satisfied: duction. Application of this “tariff-shift� method has been facilitated by the widespread adoption of the Harmonized • The vessel must be registered in the beneficiary country System (HS), under which most of the world’s more than or in the EU. 190 countries are now classifying goods according to the • The vessel must sail under the flag of the beneficiary or same harmonized categories. The level of classification of of a member state of the EU. the HS at which change is required remains an issue, how- • The vessel must be at least 60 percent owned by nation- ever. Typically, it is specified that the change should take als of the beneficiary country or the EU or by compa- place at the heading level (that is, at the four-digit level of nies with a head office in either the beneficiary country the HS).2 Examples of simple HS headings are “beer made or in an EU state, and the chairman and a majority of from malt� (HS 2203) and “umbrellas and sun umbrellas� the board members must be nationals of the beneficiary (HS 6601). But headings can be more sophisticated: country or the EU. • The master and the officers must be nationals of the Machinery, plant or laboratory equipment, whether or not beneficiary country or of an EU member, and at least 75 electrically heated (excluding furnaces, ovens and other percent of the crew must be nationals of the beneficiary equipment of heading 8514), for the treatment of materials by a process involving a change of temperature such as country or of an EU country. heating, cooking, roasting, distilling, rectifying, sterilizing, pasteurising, steaming, drying, evaporating, vaporizing, Under the EU’s Cotonou Agreement, which gives pref- condensing or cooling, other than machinery or plant of a erential access to the EU market to African, Caribbean, and kind used for domestic purposes; instantaneous or storage Pacific (ACP) countries, the rules of origin for fish are water heaters, non-electric. (HS 8419) slightly different and a little more liberal than those for GSP countries: The HS, however, was not designed specifically as a vehicle for determining country of origin; its purpose is to • The vessel must be registered in the EU or in any ACP state. provide a unified commodity classification for definition • The vessel must sail under the flag of any ACP or EU of tariff schedules and for statistical purposes. Thus, in par- country. ticular cases it can be argued that a change of tariff heading • The vessel must be at least 50 percent owned by nation- will not identify substantial transformation, whereas in als of any ACP or EU state, and the chairman and the other cases, substantial transformation may occur without majority of the board members must be nationals of change of tariff heading. As a result, schemes utilizing the any of those countries. criterion of change of tariff heading usually provide for a Preferential Rules of Origin 163 wide range of exceptions that need to satisfy other criteria unambiguous as stated. In application, however, it can for determining country of origin. become complex and uncertain. First, there is the issue of The change of tariff classification may be used to define the valuation of materials, which may be based on several both a positive test of origin, by stating the tariff classifica- prices: ex works (from factory); free on board (FOB); cost, tion of imported inputs that can be used in the production insurance, and freight (CIF); or into factory. Each method of the exported good (for example, those under a different yields a different (in this instance, ascending) value of heading), and a negative test, by stating cases in which nonoriginating materials. Second, the application of this change of tariff classification will not confer origin. For method can be costly for firms that will require sophisti- example, the North American Free Trade Agreement cated accounting systems and the ability to resolve often- (NAFTA) rule of origin for tomato ketchup states that a complex accounting questions. Finally, under the value change to ketchup (HS 210320) from imported inputs of added method, origin is sensitive to changes in the factors any chapter except subheading 200290 (tomato paste) will that determine production cost differentials across coun- confer origin. In other words, any ketchup made from tries, such as exchange rates, wages, and commodity prices. imported fresh tomatoes will confer origin, but ketchup For example, operations that confer origin in one location made from tomato paste imported from outside the area may not do so in another because of differences in wage will not qualify for preferential treatment, even though the costs, and an operation that confers origin today may not basic change of tariff classification requirement has been do so tomorrow if exchange rates change. satisfied.3 In the EU’s preferential rules of origin, bread, Specific manufacturing process. This criterion delineates biscuits, and pastry products (HS 1905) can be made from for each product or product group certain manufacturing any imported products except those in chapter 11, which or processing operations that define origin (positive test) include flour—the basic input to these products. and manufacturing or processing procedures that do not The WTO Agreement on Rules of Origin stipulates that confer origin (negative test). The rules may require the use preferential and nonpreferential rules of origin should be of certain originating inputs or prohibit the use of certain based on a positive standard, but it allows the use of nega- nonoriginating inputs. For example, EU rules of origin for tive standards (a definition of what does not confer origin) clothing products stipulate manufacture from yarn, while if they “clarify a positive standard.� The latter condition is the rule for sodium perborate requires manufacture from so vague as to have had very little effect, and EU and disodium tetraborate pentahydrate.4 NAFTA rules of origin, for example, are rife with negative The main advantage of specific manufacturing process standards. rules is that, once defined, they are clear and unambiguous Thus, although in principle the change of tariff classifi- so that, from the outset, producers are able to clearly ascer- cation can provide a simple, uniform method of determin- tain whether their product is originating or not. There are, ing origin, in practice, instead of a general rule, there are however, a number of drawbacks to this system, including often many individual rules. Nevertheless, the rule on obsolescence (as a consequence of changes in technology) change of tariff classification, once defined, is clear, unam- and documentary requirements, such as an up-to-date biguous, and easy for traders to learn, and it is relatively inventory of production processes, which may be burden- straightforward to implement. In terms of documentation, some and difficult to comply with. it requires that traders keep records showing the tariff clas- Table 8.1 summarizes the main advantages and disad- sification of the final product and of all the imported vantages of these various methods of determining suffi- inputs. This may not be a demanding requirement if the cient processing or substantial transformation. No one rule exporter directly imports the inputs, but it may be more dominates others as a mechanism for formally identifying difficult if inputs are purchased from intermediaries in the the nationality of all products, and each has its advantages domestic market. and disadvantages. It is clear, however, that different rules Value added. When the value added to a product in a of origin can lead to different determinations of origin. particular country exceeds a specified percentage, the goods Producers who are eligible for preferential access to dif- are defined as originating in that country. This criterion can ferent markets under different schemes with different rules be defined in two ways: as the minimum percentage of the of origin may find that their product qualifies under some value of the product that must be added in the country of schemes but not others. For example, a company in a origin, or as the maximum percentage of imported inputs developing country may find that the product it produces in total inputs or in the value of the product. qualifies for preferential access to the EU market under the As in the case of change of tariff classification, the value EU’s GSP scheme but that the same product does not sat- added rule has the advantage of being clear, simple, and isfy the rules of origin of the U.S. GSP scheme. 164 Paul Brenton Table 8.1. Summary of Methods for Determining Origin Rule Advantages Disadvantages Key issues Change of tariff Consistency with nonpreferential Harmonized System not designed Level of classification at which classification rules of origin. for conferring origin; as a result, change required; the higher the in the Once defined, rule is clear, there are often many individual level, the more restrictive. Harmonized unambiguous, and easy to learn. product-specific rules, which can Test may be positive (e.g., which System be influenced by domestic imported inputs may be used) or Relatively straightforward to industries. negative (e.g., definition of cases implement. Documentary requirements may be where change of classification will difficult to comply with. not confer origin); negative test Conflicts over the classification of more restrictive.a goods can introduce uncertainty about market access. Value added Clear, simple to specify, and Complex to apply; requires firms to Level of value added required to unambiguous. have sophisticated accounting confer origin. Allows for general rather than systems. Valuation method for imported product-specific rules. Uncertainty resulting from sensitivity materials: methods that assign a to changes in exchange rates, higher value (e.g., CIF) will be wages, commodity prices, and more restrictive with respect to so on. use of imported inputs. Specific Once defined, clear and Documentary requirements can be Formulation of the specific processes manufacturing unambiguous. burdensome and difficult to required; the more procedures process Provides for certainty if rules can be comply with. required, the more restrictive. complied with. Leads to product-specific rules. Use of negative test (processes or Domestic industries can influence inputs which cannot be used) or the specification of the rules. positive test (what can be used); negative test more restrictive. Can quickly become obsolete due to technological progress and therefore require frequent modification. Source: Author’s compilation. Note: CIF, cost, insurance, and freight. a. A positive determination of origin typically takes the form of “change from any other heading,� as opposed to a negative determination of origin, such as “change from any other heading except for the headings of chapter XX.� It is worth noting that change of tariff classification, particularly with a negative determination of origin, can be specified to have an effect identical to that of a specific manufacturing process. Best-Practice Suggestions for the Design of Rules of Origin Rules of Origin and Trade Preferences Although it is difficult to derive specific recommendations Preferential rules of origin define the conditions that a prod- with regard to the best-practice approach to the design of uct must satisfy to be deemed as originating in a country rules of origin, certain general propositions can be that is eligible for preferential access to a partner’s market— advanced that apply to both preferential and nonpreferen- not simply transshipped from a nonqualifying country or tial rules: subject to only minimal processing. In practice, the greater the level of work that is required by the rules of origin, the • The rules of origin should be simple, precise, easy to more difficult it is to satisfy those rules, and the more restric- understand, transparent, predictable, and stable. They tive the rules are in constraining market access relative to should avoid or minimize scope for interpretation and what is required simply to prevent trade deflection. This is administrative discretion. particularly true for small, less diversified developing • The rules should be designed to have the least trade- economies. The higher the amount of domestic value added distorting impact and should not become disguised that is required by a value added rule, the more difficult nontariff barriers to trade. Protectionist lobbying should compliance will be, since there will be less scope for the use not compromise the specification of the rules of origin. of imported parts and materials. A rule of origin that pre- • As much as possible, the rules should be consistent vents the use of imported flour in the production of pastry across products and across agreements. The greater are products such as biscuits, for example, will be very restrictive the inconsistencies, the greater will be the complexity of for countries that do not have a competitive milling industry. the system of rules of origin, both for companies and for With regard to requirements relating to sufficient pro- officials administering the various trade schemes. cessing, change of tariff classification is the most frequently Preferential Rules of Origin 165 used criterion in current preferential trade agreements, With respect to value added requirements, the WTO and it features in both EU agreements and NAFTA. WTO views a threshold for domestic content of between 40 and research (WTO 2002) shows that of 87 free trade agree- 60 percent as the norm, with a typical average import ments (FTAs) and other preferential trade agreements requirement of between 60 and 40 percent. The EU agree- investigated, 83 used change of tariff classification in the ments identify various thresholds on import content, rang- determination of origin. Most agreements specify that the ing from 30 to 50 percent. NAFTA has a domestic content change should take place at the heading level (that is, at requirement of either 50 or 60 percent, according to the the four-digit level), but in many agreements, especially method used to value the product. A value added require- those involving the EU and NAFTA, the tariff-shift require- ment of 50 percent can be very demanding in the global- ment varies by product. For example, Estevadeordal and ized world of today, in which production has become split Suominen (2003) show that, although in NAFTA about among (perhaps many) countries. 40 percent of tariff lines requires change of tariff heading, A further feature of globalization is that, for such prod- most tariff lines (54 percent) are subject to the more ucts as clothing, computers, and telecommunication equip- restrictive requirement of change of chapter (two-digit ment, much of the value added lies in the intermediate level). For a small number of products, only a change of products. High value added requirements therefore become subheading (six-digit level) is required. particularly difficult for developing countries to satisfy, Although change of tariff heading is used in most prefer- since it is the final, labor-intensive stage that they host. In ential trading agreements, it is seldom the only method this way, restrictive rules of origin act to constrain special- applied. It is also important to note that in some agree- ization at the country level. The available evidence suggests ments, such as those involving the EU, change of tariff clas- that for many products, value added in low-income coun- sification is applied to some products, but the value added tries is substantially less than 30 percent. When the final and specific manufacturing process methods are used for stage of production involves labor-intensive activities others. In NAFTA, rules of origin tend to require at least applied to relatively high-value imported inputs, it is more change of tariff classification, but the level at which change difficult for low-wage countries to satisfy a particular value is required varies across products. This typically leads to added requirement than it is for higher-wage countries. considerable complication for customs officials in deter- In general, these percentage value rules are rarely mining origin in preferential agreements. By contrast, many applied as the sole test of origin and are typically employed agreements between developing countries tend to specify with the change of tariff classification. Exceptions among general rules of origin and eschew the detailed product- agreements are the Australia–New Zealand Closer Eco- by-product approach adopted by the EU and NAFTA. nomic Relations Trade Agreement (ANZCERTA), the Furthermore, in EU agreements and in NAFTA for cer- South Pacific Trade and Economic Co-operation Agree- tain products, rules are stipulated that require satisfaction ment (SPARTECA), and the Association of Southeast Asian of more than one method to confer origin. This is clearly Nations (ASEAN) Free Trade Agreement (AFTA), which more restrictive than a requirement to satisfy a single have percentage requirements without any additional need method. For example, in NAFTA’s rules of origin, the for change of tariff heading. All three agreements do requirement for passenger motor vehicles (HS 870321) require that the last process of manufacture be undertaken reads, “A change to subheading 8703.21 from any other in the exporting country. heading provided there is a regional value content of not As noted earlier, under the value added method, origin less than 50 percent under the net cost method.� is sensitive to changes in factors such as exchange rates, In some agreements for certain products, two or more wages, and commodity prices. The value added method methods will be stipulated, and satisfaction of any one of thus tends to penalize low-labor-cost locations, which will the methods will be sufficient to confer origin. For exam- find it more difficult than higher-cost locations to add the ple, in the EU rules of origin, the requirements for wooden necessary value. It is likely to cause particular problems of office furniture (HS 940330) are “manufacture in which all compliance for companies in developing countries that the material used is classified within a heading other than lack the sophisticated accounting systems necessary under that of the product� or “manufacture in which the value of this method. all the materials used does not exceed 40 percent of the ex- Rules based on specific manufacturing processes are works price of the product.� The provision of alternative widely used (in 74 of the 83 preferential trade agreements means of satisfying origin requirements gives exporters, analyzed by the WTO), often in conjunction with the especially small firms, greater flexibility and will facilitate change of tariff classification criterion, the value added cri- trade under preferential trade agreements. terion, or both. They are a particular feature in the textiles 166 Paul Brenton and clothing sectors. Some examples of the application of without undermining the origin of the final product. In the rules follow: effect, the imported materials from the identified countries are treated as being of domestic origin in the country • A producer imports cotton fabric (HS5208), which requesting preferential access. There are three types of is then dyed, cut, and made up into cotton shirts cumulation: bilateral, diagonal (or partial), and full. (HS6105). The value of the imported materials The most basic form, bilateral cumulation, applies to amounts to 65 percent of the value of the shirts. In materials provided by either of two partners of a preferen- this case, origin would come under a change of tariff tial trade agreement. In this case, originating inputs (i.e., heading rule, but not under a value added rule, which materials) that have been produced in accordance with the requires an import content of not more than 60 per- relevant rules of origin and imported from the partner, cent or a domestic content of more than 40 percent. A qualify as originating materials when used in a country’s specific manufacturing process requirement that the exports to that partner. For example, under the EU’s products have been manufactured from yarn (the pro- GSP scheme, the rule of origin for cotton shirts states that duction stage before fabric) would mean that the origin is conferred to a beneficiary country if the shirt is product would not be originating.5 manufactured from yarn. Nonoriginating yarn may be • A doll (HS9502) is made from imported plastics and imported, but the weaving into fabric, the cutting, and the imported ready-made garments and footwear. The making up into a shirt must take place in the beneficiary. value of the imported materials amounts to 50 percent The EU’s GSP scheme allows for bilateral cumulation so of the value of the doll. In this case, the doll would be that fabric that originates in the EU (that is, fabric pro- originating under a value added rule requiring an duced in accordance with the rule of origin for fabric—in import content of no more than 60 percent; it would this case, produced from the stage of fibers) can be treated not be originating under the change of tariff heading as originating in the beneficiary country. Thus, originating because garments and accessories for dolls are classified fabrics can be imported from the EU and used in the pro- under the same tariff heading as dolls. duction of shirts for export that will qualify for preferential access to the EU. The EU, however, is often not the least- Most preferential trade agreements also specify types of cost supplier of inputs, and so the benefits of this type of operations that are deemed to be insufficient in working or cumulation can be limited. If the extra cost of using EU- processing to confer origin. Typically, they include (a) sim- sourced inputs rather than the lowest-cost inputs from ple packaging operations, such as bottling, placing in elsewhere exceeds the available benefit from preferential boxes, bags, and cases, and simple attachment on cards and access, cumulation will have no effect, and there will be no boards; (b) simple mixing of products and simple assembly improvement in market access. of parts; and (c) operations to ensure the preservation of Diagonal cumulation takes place on a regional basis. products during transport and storage. The requirements Qualifying materials from anywhere in the specified region act to ensure that these basic operations do not confer ori- can be used without undermining preferential access. In gin even if the basic rule of origin, such as change of tariff other words, parts and materials from anywhere in the heading, has been satisfied. region that qualify as originating can be used in the manu- Several other typical features of the rules of origin of facture of a final product, which can then be exported with preferential trade schemes can influence whether origin preferences to the partner country’s market. Diagonal is conferred on a product and can hence determine the cumulation is widely used in EU agreements but is not effect of the scheme on trade flows. These are cumula- applied by NAFTA. In Europe, a pan-European system of tion, tolerance rules, and absorption. The treatment of rules of origin with diagonal cumulation has been devel- duty drawback and of outward processing outside the oped to govern EU free trade agreements with countries of free trade partners or preferential trade partners can also the European Free Trade Association (EFTA) and with be important. countries in Central and Eastern Europe. Diagonal cumu- lation is allowed under the EU’s GSP scheme, but within a limited set of regional groups that have pursued their own Cumulation regional trade agreements. For example, diagonal cumula- The basic rules of origin define the processing that has to tion can take place within four regional groupings: be done in the individual beneficiary or partner to confer ASEAN, the Central American Common Market (CACM), origin. Cumulation allows producers to import materials the Andean Community, and the South Asian Association from a specific country or regional group of countries for Regional Cooperation (SAARC). Preferential Rules of Origin 167 Diagonal cumulation allows originating materials from with Algeria, Morocco, and Tunisia; and, under the Cotonou regional partners to be further processed in another coun- Agreement, with the ACP countries. It is also available in try in the group and treated as though the materials origi- the GSP schemes of Japan and the United States; among nated in the country where the processing is undertaken.6 countries within specified groupings; and, on a global This flexibility in sourcing is, however, constrained by the basis, among all developing-country beneficiaries in the further requirement that the value added in the final stage schemes of Australia, Canada, and New Zealand, as well as of production exceed the highest customs value of any of the ANZCERTA and SPARTECA regional agreements. the inputs used from countries in the regional grouping. Under full cumulation, it may be easier for more Thus, for example, with diagonal cumulation, shirt pro- developed, higher-labor-cost countries to outsource ducers in Cambodia can use fabrics from Indonesia (pro- labor-intensive, low-technology production stages to less vided that they are originating—that is, produced from the developed, lower-wage partners while maintaining the fiber stage) and still receive duty-free access to the EU, but preferential status of the good produced in low-cost loca- the value added in Cambodia must exceed the value of the tions. Diagonal cumulation, by requiring that more stages imported fabric from Indonesia. Similarly, producers in of production or higher value added be undertaken in the Nepal can import originating fabric from India and still lower-cost country, may make it more difficult for the qualify for preferential access to the EU if the value added products produced by outsourcing to qualify for preferen- in Nepal is sufficient. tial access. The documentary requirements of full cumula- As demonstrated in a report by the United Nations Con- tion may be more onerous than those required under ference on Trade and Development (UNCTAD) and the diagonal cumulation. Detailed information from suppliers Commonwealth Secretariat (2001), the value added of inputs may be required under full cumulation, whereas requirement can render regional cumulation of little value. the certificates of origin that accompany imported materi- For example, value added in the making up of clothing in als may suffice to show conformity under diagonal cumu- Bangladesh ranges from between 25 and 35 percent of the lation. For this reason, it is desirable that traders be offered value of the product, so the import content of the fabrics a choice between diagonal or full cumulation. that come from India is around 65 to 75 percent. In this To illustrate the alternatives, a clothing product made in case, the value added requirement placed on regional one country from fabric produced in a regional partner cumulation is not met, and origin of the made-up clothing and made from nonoriginating yarn would be eligible for is conferred not on Bangladesh, but on India. Regional duty-free access to the EU under full cumulation but not cumulation still allows clothing that is produced in under diagonal cumulation, since the fabric would not be Bangladesh from Indian fabrics preferential access to the deemed to be originating. (The rule of origin for the fabric EU, but not at the zero rate for which Bangladesh is eligible. requires manufacture from fibers.) Or, country A provides Rather, the rate for which India is eligible—a 20 percent parts (say, chassis for bicycles) to country B, where they are reduction from the most favored nation (MFN) rate—is processed (painted and prepared) and sent to country C applied. Thus, instead of the zero duty, which is in principle for final assembly, using locally produced parts (tires and available to Bangladesh under the Everything But Arms seat), before being exported to country D. Countries B, C, regulation, a tariff of more than 9 percent would be levied and D participate in the same FTA; country A is not a on these exports from Bangladesh to the EU. member. The value of the final product (the bicycle) In full cumulation, any processing activities carried out exported from country C to country D consists of 25 per- in any participating country in a regional group can be cent parts from country A, 25 percent value added counted as qualifying content, regardless of whether the in country B, and 50 percent parts from and value added processing is sufficient to confer originating status on in country C. The value of parts from country A makes the materials themselves. Full cumulation allows for up 50 percent of the value of the intermediate product greater fragmentation of production processes among exported from country B to country C. If there were a the members of the regional group and so stimulates 40 percent maximum import content for all products, the increased economic linkages and trade within the region. bicycle exported from country C to country D would qual- Under full cumulation, all the processing carried out in ify for preferential access under full cumulation. (Only the participating countries is assessed in deciding whether 25 percent of parts from country A is nonoriginating.) It there has been substantial transformation. Full cumulation would not, however, qualify under diagonal cumulation therefore encourages deeper integration among participat- because the value of nonoriginating materials in the prod- ing countries. Full cumulation is rare. It is currently uct exported by country B exceeds 40 percent. This inter- applied in the EU agreements with the EFTA countries; mediate product would not be treated as originating, and 168 Paul Brenton the total of nonoriginating materials in the final product import content of the part is not taken into account when would now be calculated as 50 percent of the final price of assessing the import content of the final product. The con- the bicycle (the value from both country A and country B). verse of this is that if the part does not satisfy the relevant rule of origin, it is deemed to be 100 percent nonoriginat- ing (“roll-down�). Ideally, if the part or the materials fail to Tolerance Rules satisfy the relevant rule of origin, the portion of value Tolerance, or de minimis, rules allow a certain percentage added domestically should still be counted in the determi- of nonoriginating materials to be used without affecting nation of the origin of the final product. the origin of the final product. The tolerance rule can make it easier for products with nonoriginating inputs to qualify Duty Drawback and Outward Processing for preferences under the change of tariff heading rule and the specific manufacturing process rule. This provision Provisions relating to duty drawback can lead to the repay- does not affect the value added rules. The tolerance rule ment of duties on nonoriginating inputs used in the pro- does not act to lower the limitation on the value of duction of a final product that is exported to a free trade or imported materials; the nonoriginating materials will preferential trade partner. Some agreements contain always be counted in calculating import value content. explicit no-drawback rules that will affect decisions relat- In NAFTA, nonoriginating materials can be used even ing to the sourcing of inputs by firms exporting within the if the rule on sufficient processing is not fulfilled, provided trade area, reducing the incentives for the use of imported that the value of these materials does not exceed 7 percent inputs from nonparticipating countries and encouraging of the value of the final product. Under the EU’s GSP the use of originating inputs from participating ones. scheme, the threshold is 10 percent, but under the Cotonou Increasingly important are rules concerning territoriality Agreement between the EU and the ACP countries, the tol- and the treatment of outward processing by companies erance rule allows 15 percent of nonoriginating materials based within the free trade area that is undertaken in coun- that would otherwise not be accepted. For example, in the tries that are not members of the agreement. These rules case of the doll described earlier, in which the use of dolls’ determine whether processing outside the area undermines clothing accessories denied origin to the final product under the originating status of the final product exported from the change of heading rule (since the accessories are classi- one partner to another. fied under the same heading), origin would be conferred under the EU GSP if the value of the dolls’ clothing and Rules of Origin in Existing Preferential Trade Agreements accessories is less than 10 percent of the value of the doll. Preferential rules in EU and U.S. agreements. All three The tolerance rules applied to the textiles and clothing methods of determining origin are employed in agree- sector are often different and are generally less favorable ments involving the EU and NAFTA. A key feature of the than the general rules on tolerance. In many cases, the rule EU and NAFTA models of rules of origin is that these rules is applied in terms of the maximum weight rather than are specified at a very detailed level on a product-by- the value of the nonoriginating materials that are tolerated, product basis and can be very complex—they often run to and in cases in which the value threshold is maintained, it well over 200 pages! The rules for clothing products under is set at a lower level than in the general rule. NAFTA provide an example of very complex and restrictive rules of origin. The following summarizes the rules for Absorption (Roll-Up) Principle men’s or boys’ overcoats made of wool (HS620111), which According to the absorption principle, parts or materials are typical of the nature of the rules for a wide range of that have acquired originating status by satisfying the rele- clothing products: vant rules of origin can be treated as being of domestic ori- A change to subheading 620111 from any other chapter, gin in any further processing and transformation. This is of except from heading 5106 through 5113, 5204 through particular relevance to the value added test. For example, in 5212, 5307 through 5308 or 5310 through 5311, Chapter the production of a particular part, origin is conferred 54 or heading 5508 through 5516, 5801 through 5802 or because imported materials constitute 20 percent of the 6001 through 6006, provided that the good is both cut and sewn or otherwise assembled in the territory of one or final price of the part and are less than the maximum, say, more of the Parties. 30 percent required by an import content rule of origin. This part will then be treated as 100 percent originating The basic rule of origin stipulates change of chapter but when incorporated into a final product. The 20 percent then provides a list of headings and chapters from which Preferential Rules of Origin 169 imported inputs cannot be used. In effect, the overcoat NAFTA can be contrasted with most of the agreements must be manufactured from the stage of wool fibers for- involving developing countries, such as AFTA, the Com- ward, since neither imported woolen yarn (HS5106–5110) mon Market for Eastern and Southern Africa (COMESA), nor imported woolen fabric (HS5111–5113) can be used. and the Southern Cone Common Market (Mercosur, Mer- However, the rule also states that neither imported cotton cado Común del Sur), where rules are typically general and thread (HS5204) nor imported thread of man-made there are no, or very few, product-specific rules of origin. fibers (HS54) can be used to sew the coat together. This This suggests that domestic industry did not play a signifi- rule in itself is very restrictive, and the rule for this prod- cant role in the specification of these rules. Some agree- uct is further complicated by requirements relating to the ments, such as AFTA, rely solely on the value added visible lining: method. The COMESA rules of origin require satisfaction of a value criterion (either the CIF value of imports must Except for fabrics classified in 54082210, 54082311, not exceed 60 percent of the value of all materials used, or 54082321, and 54082410, the fabrics identified in the fol- domestic value added should be at least 35 percent of the lowing sub-headings and headings, when used as visible ex-factory cost of the goods) or a change of tariff heading.7 lining material in certain men’s and women’s suits, suit- type jackets, skirts, overcoats, car coats, anoraks, wind- What are the merits of these different approaches to the breakers, and similar articles, must be formed from yarn specification of preferential rules of origin? Detailed prod- and finished in the territory of a party: 5111 through 5112, uct-by-product rules can leave very little scope for interpre- 520831 through 520859, 520931 through 520959, 521031 tation. Indeed some argue that a product-by-product through 521059, 521131 through 521159, 521213 through approach based on input from domestic producers is the 521215, 521223 through 521225, 540742 through 540744, best way of dealing with the specification of rules of origin. 540752 through 540754, 540761, 540772 through 540774, But, as the examples of fish and clothing show, product- 540782 through 540784, 540792 through 540794, 540822 specific rules can become very complex and restrictive. The through 540824 (excluding tariff item 540822aa, 540823aa more complex and the more technical the rules become, the or 540824aa), 540832 through 540834, 551219, 551229, greater is the scope for the participation of domestic indus- 551299, 551321 through 551349, 551421 through 551599, tries in setting restrictive rules of origin (see Hoekman 551612 through 551614, 551622 through 551624, 551632 1993). Indeed, “the formulation of product specific rules of through 551634, 551642 through 551644, 551692 through 551694, 600110, 600192, 600531 through 600544 or origin is, by its nature, very much out of the practical con- 600610 through 600644. trol of generalists, which is to say government officials at the policy level, and very much in the practical control of spe- This stipulates that the visible lining used must be pro- cialists, which is to say the representatives of concerned duced from yarn and finished in either party. The rule may industries� (Palmeter 2003, 159). Other interests, such as well have been introduced to constrain the effect of the tol- consumers of the relevant product, are effectively excluded erance rule, which would normally allow 7 percent of the from discussion concerning the rules of origin. weight of the article to be of nonoriginating materials. In Those who lobby hardest for trade policy interventions overcoats and suits, the lining is probably less than 7 per- are not altruistic, and their objectives with regard to rules cent of the total weight. Finally, it is interesting to note that of origin are likely to be to restrict competition from the rules of origin also provide very specific exemptions to imports and to expand their own exports within a free the rules of origin for materials that are in short supply or trade area at the expense of third-country suppliers. Such are not produced in the United States—reflecting firm- objectives can be more effectively pursued when policy is specific lobbying to mitigate the restrictiveness of the orig- determined in an environment that lacks transparency and inal NAFTA rules of origin. The most specific example is openness, as can easily occur when rules of origin are where apparel is deemed to be originating if assembled determined product by product. from imported inputs of “fabrics of subheading 511111 or From a trade policy perspective, the restrictiveness of a 511119, if hand-woven, with a loom width of less than value added rule, in terms of its impact on trade, is clearer 76 cm, woven in the United Kingdom in accordance with and more apparent than is the case for the change of tariff the rules and regulations of the Harris Tweed Association, classification and specific manufacturing process rules. It Ltd., and so certified by the Association.� The task facing is relatively straightforward to compare alternative propos- importers, and the relevant customs officials, in checking als concerning a value added rule. The extent of protection consistency with such rules is clearly not a simple one! engendered by complex and technical rules of origin that Preferential rules in other agreements. This detailed, differ across products is much more difficult to detect. This product-specific approach to rules of origin of the EU and asymmetry of information is one reason why those groups 170 Paul Brenton seeking protection will push for complex rules of origin and in the United States because of the more liberal rules of why the change of tariff classification and specific manufac- origin applied in the Canada–Chile agreement for those turing process rules may be more susceptible to capture by products, whereas other products may satisfy U.S. but not protectionist domestic interest groups (see Hirsch 2002). Canadian rules of origin requirements. Adopting a product-by-product approach to rules of origin Most free trade and nonreciprocal trade preference will tend to lead to rules that are more restrictive than is schemes contain provisions for cumulation, but there is con- necessary to prevent trade deflection—to protectionist rules siderable variation in its nature. For example, the EU allows of origin—and that can lead to an overly complex system for diagonal cumulation in the Pan-European Area of that is difficult for traders to implement and that adds con- Cumulation, encompassing the EFTA, Central and Eastern siderably to the burden on customs services. European, and Balkan countries, whereas under the Coto- More general rules of origin can allow greater scope for nou Agreement, there is full cumulation among African and interpretation, as noted by Izam (2003). In Asia, there are Caribbean countries. Similarly, for tolerance rules, which are some suggestions of underutilization of AFTA preferences, widely applied in agreements that are not based on the sole reflecting uncertainties concerning the rules of origin. It use of the value added method, there are considerable differ- appears that differing interpretations of the rules in ences across agreements, even those involving the same ASEAN countries lead to inconsistent application of the country. Under the EU–Mexico Free Trade Agreement, rules within the region. This suggests the need for more nonoriginating materials can constitute up to 10 percent of effective coordination between customs and other relevant the value of the final product, while under the agreement authorities in partner countries, with the aim of clarifying between the EU and South Africa, the level of tolerance is set existing rules and regulations rather than applying more at 15 percent. Different rules of tolerance are often estab- restrictive rules of origin. It is also important that alterna- lished for certain sectors, especially textiles and clothing. tive rules be considered so that producers are allowed some Measures of the restrictiveness of different origin regimes. flexibility in proving origin. Giving producers the option of Estevadeordal (2000) introduced an ordinal index (the satisfying either a value added rule or a change of tariff R-index) to capture the overall restrictiveness of a set of classification rule is likely to be trade facilitating. rules of origin. The index is derived from data at the tariff- There is substantial variation in the permitted amount line level and is based on an observation rule, with the fol- of nonoriginating import content under value added lowing two assumptions: requirements in different agreements. In the Canada–Chile agreement, for example, products are typically subject to a • For a change of tariff classification (CTC), change at the change of tariff classification (where the level of change chapter level is more difficult to satisfy than a change at required varies by product) and a domestic value added the heading level; a change at the heading level is stricter requirement that varies between 25 and 60 percent, accord- than at the subheading level; and a change at the sub- ing to the product and the method of valuation used. In the heading level is more stringent than at the tariff-line or U.S.–Chile agreement, where the rules are similar to those item level. of NAFTA but not identical for all products, the required • Additional criteria usually imply a more restrictive rule. domestic content is between 35 and 55 percent. Under the When a CTC is accompanied by a value added require- Canada–Chile agreement, plastic products (HS39), for ment, a specific technical process, or both, the rule is example, must satisfy the requirements of change of tariff more difficult to meet. Allowances (tolerance) and heading and between 50 and 60 percent of domestic value cumulation, by contrast, will tend to diminish the added, depending on the method of valuation. Under the restrictiveness of a given rule. U.S.–Chile agreement, most plastic products need only satisfy the requirement of change of subheading to be Higher values are assigned more demanding rules, with originating. To be originating under the U.S.–Chile agree- a maximum value of 7. Figure 8.1 provides a simple sum- ment, nonelectrical engineering products (HS84) must mary of the outcome of applying this index to a number of satisfy a change of subheading and a domestic value con- free trade agreements by using information presented in tent of between 35 and 45 percent, whereas under the Estevadeordal and Suominen (2006). The index shows that Canada–Chile agreement, such products need to satisfy agreements involving the United States and the EU tend to change of subheading but only a 25 to 35 percent content have more restrictive rules of origin than do agreements requirement (always depending on the valuation method). among developing countries, such as COMESA and the Thus, certain products produced in Chile that are granted Economic Community of West African States (ECOWAS). duty-free access to Canada may not receive such treatment Agreements involving the EU and the United States tend to Preferential Rules of Origin 171 Figure 8.1. Restrictiveness (R-Index) of Rules of Origin in imports from GSP beneficiaries in 1994 qualified for pref- Free Trade Agreements erential access to the EU market, yet only 38 percent actu- ally entered the EU market with a duty less than the MFN 6 rate. The reasons for this difference are the effects of rules of origin and tariff quotas for particular products, which 5 set limits on the amount of imports that can receive bene- 4 ficial access to the EU market. The gap also reflects the treatment of textiles and clothing products, which R-index 3 accounted for more than 70 percent of EU imports from countries covered by the GSP but for which the utilization 2 rate (the ratio of imports receiving preferences to eligible 1 imports) was only 31 percent. 0 Rules of Origin and the Utilization of Trade Preferences TA co le C M TA A AS ES D hi AC i W AF AF ex SA –C M O –C N –M CO EU EC Under the EU’s Everything But Arms Agreement for the ile EU Ch least developed countries, which offers duty-free access Source: Derived from Estevadeordal and Suominen 2006. for all products, almost all of Cambodia’s exports to the Note: AFTA, ASEAN Free Trade Agreement; ASEAN, Association of Southeast Asian Nations; CACM, Central American Common Market; COMESA, EU are eligible for zero duty preferences, yet in 2001, only Common Market for Eastern and Southern Africa; ECOWAS, Economic 36 percent of those exports obtained duty-free access. Community of West African States; EU, European Union; NAFTA, North American Free Trade Agreement; SADC, Southern African Development Brenton (2003) shows that this failure to utilize prefer- Community. For derivation of the R-index, see the discussion in the text. ences meant that, on average, Cambodia’s exports to the EU paid a tariff equivalent to 7.7 percent of the value of have complex product-specific rules of origin, whereas the total exports. Again, the main suspect for this underuti- COMESA and ECOWAS agreements have simpler rules lization of trade preferences is the rules of origin, particu- that are common across products. larly since Cambodia specializes in the production of Although agreements between developing countries clothing products, for which EU rules of origin are very often have less restrictive rules of origin on paper, in prac- restrictive, requiring production from yarn. tice their implementation can be highly restrictive. For Brenton and Manchin (2003) show that large amounts example, to be able to use the ECOWAS trade liberalization of EU imports from Eastern European countries of cloth- scheme, companies must obtain, for each and every prod- ing products made from EU-produced fabrics still enter the uct that they wish to export, approval from their national EU market under an alternative customs regime—outward ministry and then from ECOWAS. This is in addition to processing—even though there is no fiscal incentive to do the requirement of a certificate of origin for each ship- so, since EU tariffs have been removed under free trade ment. The whole process apparently takes between four to agreements. This probably reflects the costs and uncertain- six months. Indeed, exporters registering for the first time ties in proving origin that would be necessary under the are advised to state the names of future products to be normal preferential customs procedures. Estevadeordal exported under the scheme. If not, they will have to apply and Miller (2002) demonstrate that, in the transition from again for each new product they wish to export! the U.S.–Canada free trade agreement to NAFTA, rules of origin for certain sectors, such as textiles, became more restrictive and that, as a result, the utilization of the avail- Economic Implications of Rules of Origin able preferences declined. The specification and implementation of rules of origin Compliance with rules of origin entails costs that can can be a major determinant of the impact of free trade and affect the sourcing and investment decisions of compa- preferential trade agreements. In practice, rules of origin nies.9 If the optimal input mix for a firm involves the use of are controversial because the available evidence suggests imported inputs that are proscribed by the rules of origin that the utilization of preferences tends to be less than full. of a free trade agreement in which the country participates, That is, substantial proportions of actual exports that are then the rules of origin will reduce the value of the available eligible for preferences do not enter the partner’s market preferences. The firm will have to shift from the lowest-cost with zero or reduced duties but actually pay the MFN tar- source of inputs to a higher-cost source in the domestic iff.8 Sapir (1998) shows that 79 percent of EU dutiable economy, reducing the benefits of exporting under a lower 172 Paul Brenton tariff. In the extreme, if the cost difference exceeds the size the use of what are, for small companies in developing of the tariff preference, the firm will prefer to source inter- and transition economies, sophisticated and expensive nationally and pay the MFN tariff. The ability to cumulate accounting procedures. Without such procedures, it is dif- inputs from a partner under bilateral, diagonal, or full ficult for companies to show precisely the geographic cumulation will tend, in increasing order, to open the pos- breakdown of the inputs they have used. sibilities for identifying low-cost sources of inputs that do An important feature of most preferential trade schemes not compromise the qualifying nature of the final product. is the requirement of direct consignment or direct trans- Nevertheless, if the lowest-cost supplier is not a member of port. It stipulates that goods for which preferences are the area of cumulation, the benefits of the preferential requested are shipped directly to the destination market. If scheme will always be less than indicated by the size of the they are in transit through another country, documentary preferential tariff. evidence may be requested to show that the goods remained Rules of origin can also distort the relative prospects of under the supervision of the customs authorities of the similar firms within a country. For example, a clothing pro- country of transit, that they did not enter the domestic mar- ducer in Moldova may have established an efficient manu- ket there, and that they did not undergo operations other facturing process on the basis of importing fabrics from than unloading and reloading. In practice, it may be very Turkey. A less efficient producer that uses imported EU difficult to obtain the necessary documentation from fabrics may be able to expand production on the basis of foreign customs offices. preferential access to the EU market under the GSP (with Finally, it is important to note that customs authorities bilateral cumulation). The more efficient firm may not be are typically responsible for implementing the system of able to expand, since its product does not qualify for prefer- rules of origin. Customs usually has the responsibility for ences because of the use of nonqualifying fabrics and there checking the certificate of origin and may also be may be substantial costs in changing suppliers of fabrics. involved in issuing origin certificates for local exporters. These problems will be exacerbated in sectors in which Rules of origin, although an essential element of prefer- economies of scale are important. Producers that supply ential trade agreements, add considerable complexity to both preferential and nonpreferential trade partners, or the trading system for traders, customs officials, and trade that face different rules of origin in different preferential policy officials. partners, will have to produce with a different input mix Implementation of preferential trade agreements for different markets if they are to receive preferential increases the burden on customs. Limited resources and access. This may undermine the benefits from lower aver- weak administrative capacity in many developing countries age costs that would arise if total production were to be mean that these trade agreements have inevitable repercus- based on a single set of material inputs and a single pro- sions for trade facilitation. At the very least, when designing duction process. trade agreements, issues of administrative capacity in cus- Rules of origin may be an important factor in determin- toms need to be taken into account. Complicated systems of ing the investment decisions of multinational firms. Such rules of origin increase the complexity of customs proce- firms often rely on imported inputs from broad interna- dures and the burden on origin-certifying institutions. tional networks that are vital for supporting firm-specific In general, rules of origin that are clear, straightforward, advantages such as a technological edge in the production transparent, and predictable and that require little or no of certain inputs. More generally, if the nature and applica- administrative discretion will place less of a burden on cus- tion of a given set of rules of origin increase the uncer- toms than will complex rules. The use of general rather tainty concerning the extent to which preferential access than product-specific rules appears to be most appropriate will actually be provided, the level of investment will be less for preferential rules of origin applied by and to developing than if such uncertainty were reduced. countries. Less complicated rules of origin stimulate trade For companies, there is not only the issue of complying between regional partners by reducing the transaction with the rules on sufficient processing but also the cost of costs of undertaking such trade, in comparison with more obtaining the certificate of origin, including any delays that complex and restrictive rules of origin. arise in obtaining it. The costs of proving origin include WTO members have recognized that rules of origin are satisfying a number of administrative procedures, so as to an important factor affecting the ability of exporters to be able to provide the required documentation, and main- exploit market access opportunities. At the sixth WTO taining systems that accurately account for imported ministerial meeting, held in Hong Kong SAR, China, in inputs from different sources, in order to prove consistency December 2005, ministers declared that “developed-country with the rules. The ability to prove origin may well require Members, and developing-country Members declaring Preferential Rules of Origin 173 themselves in a position to do so, agree to implement duty- costs for NAFTA are around 2 percent and those for EU free and quota-free market access for products originating rules, about 6.8 percent, which reflects the more demand- from [least-developed countries (LDCs)]� and that “mem- ing certification procedures of EU schemes. bers shall take additional measures to provide effective Finally, Estevadeordal and Suominen (2006) include the market access, both at the border and otherwise, including R-index in a standard gravity model of bilateral trade flows. simplified and transparent rules of origin so as to facilitate Their econometric analysis leads them to conclude that exports from LDCs.� restrictive, product-specific rules of origin undermine over- all trade between the partners in a free trade agreement and that provisions such as cumulation and tolerance rules, Quantifying the Costs Associated with Rules of Origin which increase the flexibility of application of a given set of The costs of complying with rules of origin can be decom- processing requirements, act to boost intraregional trade. posed into distortionary costs (caused by changes in the By applying this approach at the sectoral level, they find production structure to enable compliance) and adminis- support for the hypothesis that the restrictiveness of rules trative costs (to prove origin). Information on these costs is of origin for final goods stimulates trade in intermediate limited. Early studies suggested that the costs of providing products between preferential partners. the appropriate documentation to prove origin could be It is useful to complement these econometric studies about 3 percent or more of the value of the export ship- with case studies, and there is one product-specific case ment for companies in developed countries (Herin that clearly highlights how restrictive rules of origin can 1986).10 constrain the ability of beneficiaries to exploit trade pref- Recently, efforts have been made to derive cost estimates erences (Brenton 2006). Both the EU and the United for various product-specific rules of origin by linking the States have schemes that offer duty-free access to low- index of the restrictiveness of rules of origin developed by income countries in Africa. A key sector is clothing. Most Estevadeordal (2000) to rates of utilization of preferences, of today’s developed countries and newly industrialized after controlling for the size of the preferential margin. countries have used the clothing sector as a gateway to Cadot et al. (2006) find utilization rates of preferences to industrial development. The sector has very low entry bar- be positively related to preferential margins and negatively riers: it is labor intensive, the technology is relatively sim- related to the restrictiveness of the rules of origin, as prox- ple, start-up costs are comparatively low, and scale ied by the R-index. They then proceed to use the R-index economies are negligible. The industry generates employ- and the information on utilization of preferences to carry ment for large magnitudes of unskilled labor. Finally, the out nonparametric estimation of the upper and lower clothing sector is still subject to high tariffs in rich coun- bounds of the costs of complying with the rules of origin. tries, so that there are large margins of preference for low- By revealed preference, when utilization rates are 100 per- income countries in Africa. cent, the preference margin provides an upper bound for Exports of apparel from African least-developed coun- compliance costs. When utilization rates are zero, the pref- tries to the EU and to the United States were almost equal erence margin provides a lower bound of the costs of com- in 2000, but, by 2005, the value of exports to the United plying with the rules of origin. For intermediate rates of States was more than three times greater than the value of utilization, the average rate of preference is taken to cap- exports to the EU. The key factor explaining this increase is ture the costs of compliance. The trade-weighted average the rules of origin. EU rules stipulate production from of compliance costs is found to be 6.8 percent for NAFTA yarn. This means that a double-transformation process and 8 percent for EU rules of origin. must take place in the beneficiary; the yarn is woven into Cadot et al. (2006) also use the information on utiliza- fabric, and then the fabric is cut and made up into apparel. tion rates to break down the estimate of compliance costs The rules prohibit the use of imported fabric, although into the costs attributable to the costs of administration cumulation provisions allow for the use of inputs pro- and those attributable to the distortionary element. They duced in other ACP countries. To obtain preferences, assume that low values of the R-index will tend to be asso- apparel producers must use local, EU, or ACP fabrics; they ciated with low administrative costs. (For example, the may not use fabrics from the main fabric-producing coun- requirement to satisfy only change of tariff heading will tries in Asia and still qualify for EU preferences. This is a require little paperwork.) Hence, preference margins for binding restriction, since few countries in Africa have com- high utilization rates and a low value of the R-index will set petitive fabric industries. The rules of origin under the U.S. an upper bound on the distortionary element of the com- African Growth and Opportunity Act (AGOA) allow pliance costs. The authors conclude that administrative African clothing exporters to use fabrics from any country 174 Paul Brenton (the so-called third-country fabric rule). The EU rules do fabrics from any country and to qualify for preferential not allow producers in African least-developed countries access to the EU market. the flexibility they currently have under the U.S. scheme to source fabrics globally. Rules of Origin and Economic Development De Melo and Portugal-Pérez (2008), controlling for other relevant factors, find that, although Sub-Saharan Can and should rules of origin be used as tools for stimu- African countries were offered similar preferential margins lating economic development within a regional grouping? of around 10 percent in both EU and U.S. markets under The draft ministerial text for the Cancún meeting of WTO Everything But Arms (EBA) and AGOA, the U.S. third- members as part of the Doha Development Round of country fabric rule was associated with an increase in trade negotiations proposes, under provisions for special apparel exports from the seven main African exporters of and differential treatment, that “developing and least- about 300 percent. The removal of tariffs on imports of developed country Members shall have the right to adopt apparel from Sub-Saharan African countries was estimated preferential rules of origin designed to achieve trade policy to have led to a 96 percent increase in exports. objectives relating to their rapid economic development, It is worth remembering that the EU has granted pref- particularly through generating regional trade.� Strict rules erences to African countries for apparel, subject to these of origin are viewed by some as a mechanism for encourag- strict rules of origin, for more than 20 years under ing the development of integrated production structures the Lomé and Cotonou Agreements and now EBA. These within developing countries to maximize the impact on strict rules, however, have done little to encourage the employment and to ensure that it is not just low value development of an efficient fabric industry in Africa and added activities that are undertaken. are likely to have severely constrained the impact of pref- There are problems with this view. First, such rules dis- erences in stimulating the apparel industry. (See Brenton criminate against small countries where the possibilities and Özden 2009 for a more detailed analysis of the for local sourcing are limited or nonexistent. Since most impact of the EBA and AGOA on apparel exports from developing countries are small, they are particularly dis- African least-developed countries and the role of the advantaged by restrictive rules of origin relative to larger rules of origin.) countries. Second, there is no evidence that the applica- The specific justification for constraining access to tion of strict rules of origin over the past 30 years has third-country fabrics through the use of restrictive rules of done anything to stimulate the development of integrated origin is to encourage the expansion of fabric production production structures in developing countries. In fact, in Africa, consistent with the view that vertical integration such arguments have become redundant in the light of in Africa is crucial to survival in a world in which competi- technological changes and global trade liberalizations tors in Asia are no longer constrained by quotas. However, that have led to the fragmentation of production the basis for this view is not well founded (see Stevens and processes and the development of global networks of Kennan 2004), since restrictive rules of origin will not lead sourcing. Globalization and the splitting up of the pro- to the emergence of competitive textile producers in Africa duction chain do not allow the luxury of establishing and will actually undermine the prospects of the sector. integrated production structures within countries. Strict Textile capacity will only emerge if production of apparel rules of origin act to constrain the ability of firms to inte- continues. Lack of access to competitively produced fabrics grate into global and regional production networks and, undermines the viability of the apparel sector, so that there in effect, act to dampen the location of any value added will be no demand for locally produced yarns and fabrics. activities. In the modern world economy, flexibility in the Substantial improvements in infrastructure, especially in sourcing of inputs is a key element in international com- power and transport, together with a better climate for petitiveness. Thus, it is most likely that restrictive rules of investment, are essential requirements for significant origin, rather than stimulating economic development, investments in textile production. will raise costs of production by constraining access to The European Commission now appears to have cheap inputs and will undermine the ability of local firms accepted the need for less restrictive rules of origin for to compete in overseas markets. clothing in its negotiations on economic partnership Flatters (2002) and Flatters and Kirk (2003), document- agreements (EPAs) with countries in Africa, the Caribbean, ing the evolution of the rules of origin in the South African and the Pacific. For the interim EPAs that have been signed Development Community (SADC), show that the adop- with African countries, the rules of origin for clothing have tion of restrictive rules of origin is more likely to constrain been relaxed to allow African exporters to use imported than to stimulate regional economic development. This Preferential Rules of Origin 175 example provides a salutary lesson on how sectoral inter- go beyond this function and seek to force use of local con- ests and misperceptions of the role and impact of rules of tent are likely to prove counterproductive, since they origin can undermine regional trade agreements. undermine the competitiveness of downstream industries SADC initially agreed to simple, general, and consistent (see Flatters 2001). If the objective is to stimulate regional rules of origin similar to those of the neighboring and trade, it is best achieved by adopting simple, clear, consis- overlapping COMESA. The initial rules required a change tent, and predictable rules of origin that avoid administra- of tariff heading, a minimum of 35 percent of value tive discretion and onerous burdens on customs and that added within the region, or a maximum import content minimize the costs of compliance for businesses. of 60 percent of the value of total inputs. Simple packaging and the like were defined to be insufficient to confer origin. Conclusions Subsequently, however, these rules were revised, and there are now more restrictive sector- and product-specific rules, The nature of rules of origin typically reflects the purpose with the change of tariff heading requirement being sup- that is set for them, the transparency of the process by planted by detailed technical process requirements and which they are determined, and the composition of the rules with much higher domestic value added and lower group involved in that process. Within preferential trade permitted import content. The rules became much more areas, complex and restrictive rules of origin act to dampen similar to those of the EU and of NAFTA, reflecting, in competition for final producers within a country from part, the influence of the recently negotiated EU–South suppliers in partner countries and to stimulate intra-area Africa agreement and the rules of origin governing EU exports of intermediate products by diverting demand preferences for ACP countries: away from third-country suppliers. Such rules typically emerge when the process by which they are determined The EU–South Africa rules were often invoked by special lacks transparency and openness and is dominated by interests in South Africa as models for SADC. Such claims input from domestic industry. If the purpose of preferen- were too often accepted at face value and not recognized as tial rules of origin is simply to prevent trade deflection, self-interested pleading for protection by already heavily pro- then a simple and less restrictive set of rules of origin tected domestic producers. There were few questions about the appropriateness of the underlying economic model implemented through general rather than product-specific (whatever it might be) for SADC. (Flatters and Kirk 2003, 7) rules is appropriate. In the current globalized world mar- ket, less restrictive rules are more likely to stimulate trade Flatters (2002) points out that, in the SADC case, it has and investment in the partner region by giving producers been argued that the weakness of customs administra- as much flexibility as possible in sourcing their inputs tions in the region makes it likely that low-cost products without compromising the ability to prevent transship- from Asia could enter through porous borders and then ment of goods from third countries that are not members claim tariff preferences when exported to another mem- of the agreement. If the objective of the trade agreement is ber state. It is then suggested that restrictive rules of ori- to foster trade and development, it is best achieved gin are required to prevent this from happening. There is through simple and liberal rules of origin, rather than by no reason, however, to expect that weak customs admin- using rules of origin as opaque measures of trade protec- istrations would be better able to enforce strict rules of tion. The analysis in this chapter leads to the following origin than less restrictive rules. In fact, in many cases, the broad conclusions: rules of origin are so strict that no producers in the region can satisfy them and therefore no discretion on the part 1. Rules of origin that vary across products and agree- of customs is required; preferences are not granted, and ments add considerably to the complexity and costs of the preferential trade agreement has no impact. A better participating in and administering trade agreements. approach is to adopt economically sensible rules of origin The incidence of such costs falls particularly heavily on and a program for improving administrative capacities small and medium-size firms and on firms in low- in customs. Clearly designed safeguard measures can also income countries. Complex systems of rules of origin be adopted to deal with surges of imports entering via add to the burdens on customs and may compromise partner countries. progress on trade facilitation. To conclude, rules of origin are an inefficient tool for 2. Restrictive rules of origin constrain international spe- achieving development objectives; better policies are avail- cialization and discriminate against small, low-income able. Rules of origin should be used as a mechanism for countries, where the possibilities for local sourcing are preventing trade deflection. Restrictive rules of origin that limited. 176 Paul Brenton 3. Simple, consistent, and predictable rules of origin are developing countries. For a recent discussion of utilization rates of GSP schemes and rules of origin, see Inama (2002). more likely to foster the growth of trade and develop- 9. Economists have generally given little attention to rules of origin ment. Specifying generally applicable rules of origin, within the voluminous literature on free trade areas. The key initial con- with a limited number of clearly defined and justified tributions on rules of origin are Krueger (1997) and Krishna and Krueger exceptions, is appropriate if the objective is to stimulate (1995); these authors demonstrate how rules of origin can act as “hidden protectionism� and can induce a switch in demand in free trade partners integration and minimize the burdens on firms and from low-cost external inputs to higher-cost partner inputs to ensure that customs in complying with and administering the rules. final products actually receive duty-free access. Falvey and Reed (1998) 4. Producers should be accorded flexibility in meeting show how rules of origin can be used to protect a domestic industry from unwanted competition from a partner, even in conditions where trade origin rules, for example, by specifying that either a deflection is unlikely. change of tariff requirement or a value added rule can 10. Herin (1986) also found that as a result of the costs to EFTA pro- be satisfied. ducers of proving origin, one-quarter of EFTA exports to the EU paid the applied MFN duties. 5. Preferences granted by developed countries would be more effective in stimulating exports from developing countries if they were governed by less restrictive References rules of origin. Ideally, these schemes should have Brenton, Paul. 2003. “Integrating the Least Developed Countries into the common rules of origin. Producers in developing coun- World Trading System: The Current Impact of EU Preferences under tries should be able to gain preferential access to all Everything But Arms.� Journal of World Trade 37: 623–46. developed-country markets if their product satisfies a ———. 2006. “Enhancing Trade Preferences for LDCs: Reducing the Restrictiveness of Rules of Origin.� In Trade, Doha and Development, single origin test. ed. Richard Newfarmer. Washington, DC: World Bank. 6. Restrictive rules of origin should not be used as tools for Brenton, Paul, and Hiroshi Imagawa. 2005. “Rules of Origin, Trade and achieving economic development objectives; they are Customs.� In The Customs Modernization Handbook, ed. Luc De Wulf and José B. Sokol. Washington, DC: World Bank. http://sitere likely to be counterproductive. The potential benefits of sources.worldbank.org/INTEXPCOMNET/Resources/Customs_Mod trade agreements among developing countries can be ernization_Handbook.pdf. substantially undermined if those agreements contain Brenton, Paul, and Miriam Manchin. 2003. “Making EU Trade Agree- ments Work: The Role of Rules of Origin.� World Economy 26 (5): restrictive rules of origin. 755–69. ˇlar Brenton, Paul, and Çag Özden. 2009. “Trade Preferences for Apparel and the Role of Rules of Origin: The Case of Africa.� In Trade Prefer- Notes ence Erosion: Measurement and Policy Response, ed. Bernard Hoekman, Will Martin, and Carlos A. Primo Braga, 401–24. Washington, DC: This chapter is based on Brenton and Imagawa (2005). Palgrave Macmillan and World Bank. 1. Countries also apply another set of often different nonpreferential Cadot, Olivier, Celine Carrère, Jaime de Melo, and Bolormaa rules of origin in applying basic trade policy measures such as tariffs, quan- Tumurchudur. 2006. “Product-Specific Rules of Origin in EU and US titative restrictions, antidumping measures and countervailing duties, and Preferential Trading Arrangements: An Assessment.� World Trade safeguard measures, as well as to fulfill requirements relating to country- Review 5 (2): 199–224. of-origin marking and public procurement and for statistical purposes. De Melo, Jaime, and Alberto Portugal-Pérez. 2008. “Rules of Origin, 2. The Harmonized System comprises 96 chapters (two-digit level), Preferences and Diversification in Apparel: African Exports to the 1,241 headings (four-digit level), and about 5,000 subheadings (six-digit US and to the EU.� CEPR Discussion Paper 7072, Centre for Eco- level). nomic Policy Research, London. http://www.cepr.org/pubs/new-dps/ 3. The apparent reason for this rule in NAFTA is to protect producers dplist.asp?dpno=7072. of tomato paste in Mexico from competition from producers in Chile; see Estevadeordal, Antoni, 2000. “Negotiating Preferential Market Access: The Palmeter (2003). Case of NAFTA.� Journal of World Trade 34 (1): 141–66. 4. The EU rule of origin for sodium perborate also allows satisfaction Estevadeordal, Antoni, and Eric Miller. 2002. “Rules of Origin and the Pat- of a maximum import content rule of 40 percent. tern of Trade between US and Canada.� Inter-American Development 5. This yarn-forward rule is common in EU agreements for all cloth- Bank (IADB), Washington, DC. ing products. The United States typically applies an even stricter process Estevadeordal, Antoni, and Kati Suominen. 2003. “Rules of Origin in FTAs rule requiring that the clothing be made from fibers, which means that the in Europe and the Americas: Issues and Implications for the EU–MER- processes of spinning fiber into yarn and weaving yarn into fabric, as well COSUR Inter-Regional Association Agreement.� In Market Access for as making up fabric into clothing, have to be undertaken in the exporting Goods and Services in the EU–Mercosur Negotiations, ed. Alfredo G. A. country to confer origin on the product. Valladão and Roberto Bouzas. Paris: Chaire Mercosur de Sciences Po. 6. For both bilateral and regional cumulation, there can be an addi- ———. 2006. “Rules of Origin: A World Map and Trade Effects.� In The tional requirement that the processing carried out be more than “insuf- Origin of Goods: Rules of Origin in Preferential Trade Agreements, ed. ficient working or processing.� This addition, which is typical in EU Olivier Cadot, Antoni Estevadeordal, Akiko Suwa-Eisenmann, and agreements but not in those of other countries, requires that more than Thierry Verdier. Oxford, U.K.: Oxford University Press. packing, mixing, cleaning and preserving, and simple assembly of parts Falvey, Rod, and Geoff Reed. 1998. “Economic Effects of Rules of Origin.� take place. Weltwirtschaftliches Archiv 134: 209–29. http://www.springerlink.com/ 7. The COMESA agreement also specifies that a range of goods content/g638k36612659n59/. deemed to be of particular importance to economic development need Flatters, Frank. 2001. “The SADC Trade Protocol: Which Way Ahead?� only satisfy a 25 percent domestic value added criterion. Southern African Update 10: 1–4. 8. For many years, UNCTAD has been highlighting the relatively low ———. 2002. “SADC Rules of Origin: Undermining Regional Free levels of utilization of preferences granted by developed countries to Trade.� Presented at the Trade and Industrial Strategies (TIPS) Forum, Preferential Rules of Origin 177 Johannesburg, September. http://qed.econ.queensu.ca/faculty/flat- Krishna, Kala, and Anne Krueger. 1995. “Implementing Free Trade Areas: ters/writings/ff_sadc_roo_tips_forum.pdf. Rules of Origin and Hidden Protection.� NBER Working Paper 4983, Flatters, Frank, and Robert Kirk. 2003. “Rules of Origin as Tools of Devel- National Bureau of Economic Research, Cambridge, MA. http://www opment? Some Lessons from SADC.� Presented at Institut National de .nber.org/papers/w4983. la Recherche Agronomique (INRA) Conference on Rules of Origin, Krueger, Anne. 1997. “Free Trade Agreements versus Customs Unions.� Paris, May. Journal of Development Economics 54: 169–87. Herin, Jan. 1986. “Rules of Origin and Differences between Tariff Levels in Palmeter, David. 2003. The WTO as a Legal System: Essays on International EFTA and in the EC.� European Free Trade Association (EFTA) Secre- Trade Law and Policy. London: Cameron May. tariat, Geneva. Sapir, André. 1998. “The Political Economy of EC Regionalism.� European Hirsch, Moshe. 2002. “International Trade Law, Political Economy and Economic Review 42: 717–32. Rules of Origin: A Plea for a Reform of the WTO Regime on Rules of Stevens, Christopher, and Jane Kennan. 2004. “Comparative Study of Origin.� Journal of World Trade 36 (2): 171–88. G8 Preferential Access Schemes for Africa: Report on a DFID- Hoekman, Bernard. 1993. “Rules of Origin for Goods and Services: Commissioned Study.� Institute of Development Studies, Brighton, U.K. Conceptual and Economic Considerations.� Journal of World Trade UNCTAD (United Nations Conference on Trade and Development) and 27: 81–99. Commonwealth Secretariat. 2001. “Duty and Quota Free Market Inama, Stefano. 2002. “Market Access for LDCs: Issues to be Addressed.� Access for LDCs: An Analysis of Quad Initiatives.� UNCTAD, Geneva; Journal of World Trade 36 (1): 85–116. Commonwealth Secretariat, London. http://r0.unctad.org/ditc/tab/ Izam, Miguel. 2003. “Rules of Origin and Trade Facilitation in Preferential publications/duty_quota_free.pdf. Trade Agreements in Latin America.� Presented at the International WTO (World Trade Organization). 2002. “Rules of Origin Regimes in Forum on Trade Facilitation, Geneva, May. Regional Trade Agreements.� WT/REG/W/45, WTO, Geneva. 9 Trade Remedy Provisions Thomas J. Prusa The question of whether preferential trade agreements injury. Because this material injury standard is considered (PTAs) are good or bad for the global trade system has weaker, and because antidumping measures and CVD pro- always been contentious. As experience has accumulated, tection are country specific, sanctions against unfair trade our understanding of the consequences of PTAs has practices are generally easier to apply. World Trade Organi- evolved. The traditional worries about welfare impacts zation (WTO) rules require that, for all of the trade reme- associated with trade creation and diversion have been dies discussed here, there be a link between change in trade augmented by a multitude of new and unanticipated volume and the imposition of trade protection; the exis- concerns.1 tence of a causal link is generally determined by an admin- Most countries have been reducing tariffs across the istrative body in the importing country. board for all partners on a nondiscriminatory basis, and An important precursor to the analysis reported here the value of PTA preferences has, as a consequence, steadily was the development of a database of PTA provisions. As of fallen. How much trade creation or diversion can be early 2009, the database contained detailed information on expected when preferential rates are essentially the same as trade remedy provisions in 74 PTAs. The longer-run goal is most favored nation (MFN) rates?2 For many PTAs, there- to survey all PTAs reported to the WTO. fore, the main welfare consequences are likely to stem from Some PTAs include no language concerning specific nontariff provisions. Indeed, as emphasized in this hand- trade remedies; others prohibit trade remedies against book, PTAs have increasingly come to address many issues members. Often, PTAs allow trade remedies but add extra beyond tariffs—government procurement, labor stan- rules. The database permits us to dig deeper and look at dards, environmental protection, and so on. which rules were actually included in the agreements. We This chapter examines the potential effects of one such are then able to determine whether certain rules are more beyond-tariffs area, trade remedy provisions. The focus is common than others and whether countries that are on the most frequently used of these provisions: antidump- involved in many PTAs are consistent in the provisions ing measures, countervailing duties (CVDs), and safeguard they enact in different agreements. measures. Antidumping measures and countervailing The next section surveys some of the political and eco- duties are designed to sanction exporters who engage in nomic justifications for including trade remedy provisions “unfair� trading practices that cause material injury to in PTAs. The subsequent sections take advantage of the domestic producers. These unfair practices can take the database to survey the provisions contained in the PTAs form of selling products below their “normal� price included. As a first step, the PTAs are divided into three (dumping) or of benefiting from government-provided sub- groups: those with trade remedy rules, those that prohibit sidies (the situation that CVDs are meant to address). By the use of trade remedies, and those without any trade contrast, safeguard actions are designed to deal with unex- remedy rules. Specific provisions are then scrutinized in pected circumstances arising in the course of “fair� trade. greater detail. Next, the hub-and-spoke pattern of PTAs is They can be imposed even if there has been no unfair trade discussed, along with the hubs’ use of trade remedy pro- practice, as long as imports have increased to such an visions. Despite considerable variation in rules within extent that domestic producers have suffered serious hubs, there is evidence of different North American and 179 180 Thomas J. Prusa European philosophies regarding trade remedies. The set will be fewer subsidies and thus less need for countervail- of PTAs that have managed to prohibit one or more trade ing duties is not supported. remedies is then examined. Finally, the chapter concludes Each of the three explanations suggests that PTAs may with an analysis of the important issue of protection alter the demand for trade remedy protection. On the one diversion. hand, import-competing sectors need to be assured that Before moving to the main discussion, a comment on they can protect themselves from the unanticipated conse- terminology is useful: the terms “trade remedies,� “con- quences of the regional liberalization program. Retaining tingent protection,� and “administered protection� are trade remedies in the PTA helps maintain political support employed interchangeably in this chapter. for the agreement. On the other hand, regional liberaliza- tion might eliminate unfair trade. To the extent that PTA trade remedy provisions offer The Political Economy of the Need for new forms of protection or make existing forms of protec- Trade Remedies in PTAs tion easier to obtain, they are similar to provisions in PTAs The rationale for the inclusion of preferential tariff for long transition periods, complicated rules of origin, and schedules and definitions of rules of origin in PTAs seems carve-outs for sensitive sectors—all of which result in clear. It is less obvious why most PTAs devote significant slower liberalization for import-competing sectors. Instead language to amending and qualifying the use of trade of directly cushioning the effects of the PTA by drawing remedies. out the process of tariff elimination, trade remedies One explanation for the widespread presence of trade achieve a different cushioning effect by specifying a set of remedies in PTAs is the political economy of protection- conditions—injury to the domestic industry—under which ism (Tharakan 1995). The long-term process of tariff lib- regional liberalization may be temporarily suspended or eralization in the post–World War II era has reduced tar- partially reversed. Bilateral safeguard rules are an example iff rates to very low levels worldwide. Import-competing of rules that temporarily reverse preferential concessions. sectors, however, still have an incentive to secure protec- Such rules may hurt PTA partners and moderate beneficial tion through whatever means they can find. With the trade creation, but they may be beneficial from a global most direct route (tariffs) eliminated, these interests turn perspective if they serve to lessen trade diversion. to the next best alternative, contingent protection (trade PTA provisions that make contingent protection more remedies). difficult to grant have more subtle effects. Abolishing or A second, related argument is that contingent protec- restricting the use of trade remedies with respect to PTA tion acts as a pressure-release valve that enables continued partners’ trade will most likely increase intrabloc trade. liberalization (Jackson 1997). Trade liberalization often The welfare effects, however, are uncertain. The ambiguity imposes costs of adjustment on uncompetitive industries, stems from the well-known insight that preferential trade and the incorporation of trade remedy measures in PTAs arrangements have both trade creation and trade diversion may be thought of as a way of managing the political con- effects (Viner 1950). Rules on contingent protection can sequences of these costs through a temporary reversal of clearly both create and divert trade (Bown and Crowley liberalization. 2007). Empirically, it turns out that the trade remedy rules in The danger is that, as intraregional trade expands PTAs often make granting protection more difficult. A because of preferential tariffs, contingent protection will be third rationale explains why this might be so. The inclusion increasingly directed at the imports of nonmembers. of PTA provisions that restrict the use of trade remedies is Bhagwati (1996) and Bhagwati and Panagariya (1996) fore- consistent with the view that contingent protection is nec- saw this danger, arguing that the elastic and selective nature essary because countries are insufficiently open to trade. of contingent protection increases the risk that PTAs will For example, Mastel (1998) argues that dumping is driven lead to trade diversion.4 As specific provisions are discussed by closed home markets. The elimination of barriers to in what follows, it is important to consider the conflicting intra-PTA trade reduces the ability of firms to dump, as motivations countries may have when negotiating agree- they no longer have a protected home market where they ments. If trade remedies serve primarily as pressure-release can earn supernormal profits.3 This third explanation is valves, PTAs should include provisions that make it easier also consistent with the lack of rules on countervailing for domestic industries to raise barriers, but if PTAs duties in PTAs. Specifically, given that most PTAs have open up closed home markets, then, arguably, some trade failed to strengthen antisubsidy rules, the notion that there provisions are not needed. Trade Remedy Provisions 181 Incidence of Trade Remedy Actions and countervailing duty measures, but it is also a major concern for global safeguards because provisions in PTAs Before discussing the role of trade remedies in PTAs, it is often allow PTA members to be excluded from these safe- useful to review the incidence of trade remedy actions over guards (Bown 2004). the past decade. Table 9.1 presents data from notifications made by members to the WTO over the 1995–2007 period. Whereas the other WTO exceptions—infant industries, Trade Remedy Provisions in PTAs balance of payments, national security, and so on—are PTAs vary in size, degree of integration, geographic scope, rarely invoked, the provisions studied in this paper have and members’ level of economic development, and the been used literally hundreds (or, in the case of antidump- political and economic demands for trade remedy provi- ing measures, thousands!) of times.5 sions across PTAs also necessarily vary. The proliferation Countries’ clear preference for using antidumping and diversity of PTAs has produced a complicated pattern rather than countervailing duty measures or safeguards is in the use and inclusion of trade remedy provisions across striking. As shown, there were nearly nine times more initi- PTAs that defies simple characterization. ations of antidumping measures (3,220) than of counter- Some PTAs contain long discussions of trade remedy vailing duty (201) and safeguard (163) actions combined.6 rules; others do not even mention trade remedies. For A similar discrepancy is seen in the number of measures some PTAs, the trade remedy provisions make protection applied. easier, but in most cases they make it more difficult to There has been a significant change in the use of these impose. remedies. The four major users—Australia, Canada, the A simple characterization is impossible, not only European Union (EU), and the United States—accounted because trade remedy provisions vary from one PTA to for more than 90 percent of the contingent trade initia- the next but also because provisions differ for the same tions during the 1980s and were the targets in more than country across different PTAs. For example, PTAs entered 75 percent of the investigations (Prusa 2001).7 By contrast, into by the United States have no specific antidumping countries from all parts of the world are now active users provisions except for the North American Free Trade and targets of contingent protection (Prusa 2005). Since Agreement (NAFTA), which contains a number of these 1995, 43 countries have initiated antidumping cases, 18 provisions—notably, the creation of binational panels have initiated countervailing duty cases, and 30 have initi- that review antidumping determinations made by ated global safeguard cases. Nearly 100 countries have national authorities. Similarly, the EU has entered into been the subject of antidumping investigations, and 40 PTAs that have no antidumping rules, others that contain have been targeted in countervailing duty investigations.8 many antidumping rules, and even some that prohibit the The broadened set of uses and targets of trade remedies use of antidumping. reflects increased globalization. Trade remedies can reinforce the trade diversion effects of a PTA: on average, the imposition of antidumping and Diversity among PTAs countervailing duty measures reduces subject imports Tables 9.2 and 9.3 present a summary of the 74 PTAs sur- from the targeted country by about half (Prusa 2001). veyed and their characteristics. With only four exceptions, When faced with contingent protection measures, non-PTA the PTAs mapped were notified to the WTO. The database members will be at an even greater disadvantage than under includes PTAs with members in Europe, North America, preferential tariffs. The potential for such discrimination the Caribbean, Latin America, Asia and the Pacific, Africa, is clear for country-specific measures such as antidumping and the Middle East. The sample reflects the economic diversity of PTAs, covering as it does North-North, South- South, and North-South agreements. Most (46) of the Table 9.1. Trade-Contingent Initiations and Measures in PTAs, 1995–2007 sampled PTAs have a mix of developed and developing countries in their membership; 22 have only developing Trade-contingent instrument Initiations Measures countries as members, and 6 have developed members Antidumping measures 3,220 2,052 only.9 The sample is dominated by free trade agreements: Countervailing duties 201 119 80 percent of the PTAs in the sample are free trade areas, Safeguards 163 83 10 percent are customs unions, and 10 percent are Source: WTO Secretariat. preferential trade areas.10 182 Thomas J. Prusa Table 9.2. Contingent Protection Rules in Selected PTAs Trade remedy Development Entry into Relevant GATT status of Countervailing Global Bilateral PTA force provision Typea members Antidumping duty safeguards safeguards AFTA 1992 Enabling Clause FTA Developing No rules No rules No rules Rules Andean Community 1993 CU Developing Rules Rules No rules Rules ANZCERTA 1990 Article XXIV FTA Developed Disallowed Rules No rules Rules Australia–Singapore 2003 Article XXIV FTA Mixed Rules Rules No rules Disallowed Australia–Thailand 2005 Article XXIV FTA Mixed Rules Rules Rules Rules Australia–United States 2005 Article XXIV FTA Developed No rules No rules Rules Rules CACM 1961 Article XXIV CU Developing Rules Rules No rules No rules Canada–Chile 1997 Article XXIV FTA Mixed Disallowed No rules Rules Rules Canada–Costa Rica 2002 Article XXIV FTA Mixed Rules No rules Rules Rules Canada–Israel 1997 Article XXIV FTA Mixed No rules Rules Rules Disallowed CARICOM 1973 Article XXIV CU Developing Rules Rules No rules Rules CEMAC 1999 Enabling Clause PTA Developing No rules No rules No rules No rules China–Hong Kong SAR, China 2004 Article XXIV FTA Developing Disallowed Disallowed No rules Rules China–Macao SAR, China 2004 Article XXIV FTA Developing Disallowed Disallowed No rules Rules COMESA 1994 Enabling Clause PTA Developing Rules Rules No rules Rules EEA 1994 Article XXIV FTA Developed Disallowed Disallowed No rules Rules EFTA 2001 Article XXIV FTA Developed Disallowed Disallowed No rules Rules EFTA–Chile 2004 Article XXIV FTA Mixed Disallowed Rules Rules Rules EFTA–Croatia 2002 Article XXIV FTA Mixed Rules Rules No rules Rules EFTA–Israel 1993 Article XXIV FTA Mixed Rules Rules No rules Rules EFTA–Jordan 2002 Article XXIV FTA Mixed Rules Rules No rules Rules EFTA–Macedonia, FYR 2001 Article XXIV FTA Mixed Rules Rules No rules Rules EFTA–Morocco 1999 Article XXIV FTA Mixed Rules Rules No rules Rules EFTA–Palestinian Authority 1999 Article XXIV FTA Mixed Rules Rules No rules Rules EFTA–Singapore 2003 Article XXIV FTA Mixed Disallowed Rules No rules Rules EFTA–Tunisia 2005 Article XXIV FTA Mixed Rules Rules Rules Rules EFTA–Turkey 1992 Article XXIV FTA Mixed Rules Rules No rules Rules EU 1958 Article XXIV CU Developed Disallowed Disallowed No rules Disallowed EU–Algeria 1976 Article XXIV FTA Mixed Rules Rules Rules Rules EU–Andorra 1991 Article XXIV CU Mixed No rules No rules No rules No rules EU–Chile 2003 Article XXIV FTA Mixed Rules Rules Rules Rules EU–Croatia 2002 Article XXIV FTA Mixed Rules Rules No rules Rules EU–Egypt, Arab Rep. 2004 Article XXIV FTA Mixed Rules Rules Rules Rules EU–Faeroe Islands 1997 Article XXIV FTA Mixed Rules No rules No rules Rules EU–Macedonia, FYR 2001 Article XXIV FTA Mixed Rules No rules No rules Rules EU–Israel 2000 Article XXIV FTA Mixed Rules No rules No rules Rules EU–Jordan 2002 Article XXIV FTA Mixed Rules No rules No rules Rules EU–Lebanon 2003 Article XXIV FTA Mixed Rules Rules Rules Rules EU–Mexico 2000 Article XXIV FTA Mixed Rules Rules No rules Rules EU–Morocco 2000 Article XXIV FTA Mixed Rules No rules No rules Rules EU–OCT 1971 Article XXIV FTA Mixed No rules No rules No rules Rules EU–Palestinian Authority 1997 Article XXIV FTA Mixed Rules No rules No rules Rules EU–South Africa 2000 Article XXIV FTA Mixed Rules Rules Rules Rules EU–Switzerland– Liechtenstein 1973 Article XXIV FTA Developed Rules No rules No rules Rules (continued next page) Trade Remedy Provisions 183 Table 9.2. (continued) Trade remedy Development Entry into Relevant GATT status of Countervailing Global Bilateral PTA force provision Typea members Antidumping duty safeguards safeguards EU–Syrian Arab Republic 1977 Article XXIV FTA Mixed Rules Rules No rules Rules EU–Tunisia 1998 Article XXIV FTA Mixed Rules No rules No rules Rules EU–Turkey 1996 Article XXIV CU Mixed Rules No rules No rules Rules GCC 1981 Enabling Clause PTA Developing No rules No rules No rules No rules Group of Three (Colombia, Mexico, and Venezuela, RB) 1995 FTA Developing Rules Rules Rules Rules Japan–Singapore 2002 Article XXIV FTA Mixed No rules No rules Rules Rules Korea, Rep.–Chile 2004 Article XXIV FTA Developing Rules Rules Rules Rules LAIA/ALADI 1981 Enabling Clause PTA Developing No rules No rules No rules Rules Mercosur 1991 Enabling Clause CU Developing Rules Rules No rules Disallowed Mexico–Chile 1999 Article XXIV FTA Developing No rules No rules Rules Rules Mexico–EFTA 2001 Article XXIV FTA Mixed Rules Rules No rules Rules Mexico–Israel 2000 Article XXIV FTA Developing Rules Rules Rules Rules Mexico–Japan 2005 Article XXIV FTA Mixed No rules No rules Rules Rules Mexico–Nicaragua 1998 Article XXIV FTA Developing Rules Rules Rules Rules Mexico–Northern Triangle 2001 FTA Developing Rules Rules Rules Rules Mexico–Uruguay 2004 FTA Developing Rules Rules Rules Rules NAFTA 1994 Article XXIV FTA Mixed Rules Rules Rules Rules New Zealand– Singapore 2001 Article XXIV FTA Mixed Rules No rules No rules Disallowed SADC 2000 Article XXIV FTA Developing Rules Rules No rules Rules SAFTA 1995 Enabling Clause PTA Developing Rules Rules No rules Rules SPARTECA 1981 Enabling Clause PTA Mixed Rules No rules No rules Rules Turkey–Israel 1997 Article XXIV FTA Developing Rules No rules No rules Rules United States– Bahrain 2006 Article XXIV FTA Mixed No rules No rules Rules Rules United States– CAFTA-DR 2006 Article XXIV FTA Mixed No rules Rules Rules Rules United States–Chile 2004 Article XXIV FTA Mixed No rules Rules Rules Rules United States–Israel 1985 Article XXIV FTA Mixed No rules No rules No rules Rules United States–Jordan 2001 Article XXIV FTA Mixed No rules No rules Rules Rules United States– Morocco 2006 Article XXIV FTA Mixed No rules No rules Rules Rules United States– Singapore 2004 Article XXIV FTA Mixed No rules No rules Rules Rules WAEMU/UEMOA 2000 Enabling Clause PTA Developing Rules No rules Rules Rules Source: Author’s compilation. Note: Blank cells under “Relevant GATT provision� indicate PTAs not notified to the WTO. AFTA, ASEAN Free Trade Area; ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Agreement; ASEAN, Association of Southeast Asian Nations; CACM, Central American Common Market; CAFTA-DR, Dominican Republic–Central America Free Trade Agreement; CARICOM, Caribbean Community; CEMAC, Economic and Monetary Community of Central Africa (Communauté Économique et Monétaire de l’Afrique Centrale); COMESA, Common Market for Eastern and Southern Africa; EEA, European Economic Area; EFTA, European Free Trade Association; EU, European Union; GCC, Gulf Cooperation Council; LAIA/ALADI, Latin American Integration Association/Asociación Latinoamericana de Integración; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement; OCT, Overseas Countries and Territories; SADC, Southern African Development Community; SAFTA, South Asian Free Trade Area; SAR, special administrative region; SPARTECA, South Pacific Regional Trade and Economic Cooperation Agreement; WAEMU/UEMOA, West African Economic and Monetary Union/Union Économique et Monétaire Ouest-Africaine. a. CU, customs union; FTA, free trade area; PTA, preferential trade agreement. 184 Thomas J. Prusa Table 9.3. Characteristics of PTAs Development status. Many PTAs have a mixed member- Intra-PTA ship of developed and developing countries; others have imports, 2005 only developing or developed countries as members. Percentage (billions of Developed countries such as Canada, the United States, Characteristic Number of total U.S. dollars) and EU members have a long history with trade remedies Relevant GATT provision that might change their political willingness to restrict or Article XXIV 61 82.4 prohibit the use of these instruments. Enabling Clause 9 12.2 Legal basis. The PTAs in our survey represent about half Unknown 4 5.4 of the total number of PTAs notified to the WTO under Type of agreement the General Agreement on Tariffs and Trade (GATT) Customs union 7 9.5 Free trade agreement 60 81.1 Article XXIV and the Enabling Clause of 1979.11 As shown Preferential trade in table 9.3, about 82 percent of the PTAs in our sample agreement 7 9.5 were notified under Article XXIV of GATT and about 12 Development status of percent under the Enabling Clause.12 Given its roots, PTAs members notified under the Enabling Clause may have fewer rules. Developed 6 8.1 2,932.4 Developing 22 29.7 501.0 Mixed 46 62.2 1,307.7 A First Look at the Provisions Source: Author’s compilation. For this analysis, three key trade remedy provisions were Note: GATT, General Agreement on Tariffs and Trade. Numbers may not mapped: antidumping measures, countervailing duties, and sum to totals because of rounding. global safeguards. In addition, information was gathered on The substantial variation in trade remedy provisions the provisions regarding bilateral safeguards. These are found across PTAs likely reflects the diversity among PTAs. in most PTAs and are meant to apply only to the trade of Size. Many PTAs involve fairly small amounts of intrare- other PTA members. They provide a temporary escape hatch gional trade; these include the Economic and Monetary from PTA commitments. Generally, bilateral safeguards Community of Central Africa (CEMAC, Communauté require evidence of increased imports from regional partners Économique et Monétaire de l’Afrique Centrale); the agree- and of serious injury to the domestic industry. In this sense, ments between the European Free Trade Association (EFTA) they are similar to the other forms of contingent protection. and Tunisia, Mexico, and Uruguay and between the United A two-level template was adopted to facilitate the analy- States and Bahrain; and the West African Economic and sis and comparison of PTAs. First, for each provision, PTAs Monetary Union/Union Économique et Monétaire Ouest- were mapped into three distinct groups: Africaine (WAEMU/UEMOA). Others, including the EU, • Those that disallow the remedy among the members the European Economic Area (EEA), and NAFTA, involve • Those with no language regarding the remedy substantial amounts of trade. PTAs with a great deal of trade • Those with specific rules regarding the remedy. may well have greater political demands for trade remedies than those with less trade. It is clear, however, that although A quick glance at table 9.2 shows that there is considerable size may matter, it is not the only determinant of trade rem- variation, both among PTAs and within provisions. Some edy rules: the two largest PTAs, the EU and NAFTA, have PTAs prohibit the use of antidumping and countervailing very different philosophies about trade remedy rules, as do duty remedies against members; in this group are the two of the smallest PTAs, CEMAC and EFTA–Tunisia. agreements between China and Hong Kong SAR, China, Integration. There is no clear definition of integration and between China and Macao SAR, China; the EEA; with respect to PTAs, and PTAs differ greatly as to their EFTA; and the EU. Others have no language for any of the degree of integration. Clearly, however, the extent to which main trade remedy laws, and some have no language for PTAs go beyond simple tariff reductions influences the antidumping and countervailing duty measures but con- type of trade remedy provisions and the approach toward tain special provisions for global safeguards. Finally, many them. PTAs with deeper integration have adopted harmo- PTAs have extra rules for all the provisions studied. nized or common behind-the-border measures, they have Some key differences among PTAs and trade remedy allowed for free or freer movement of capital and labor, and provisions emerge from the analysis: some have even adopted a single currency. PTAs that aim at deeper integration are more likely to do away with trade • Antidumping is by far the most likely provision to be remedy measures. prohibited (see table 9.4).13 Trade Remedy Provisions 185 Table 9.4. Summary of Contingent Protection Rules in PTAs Countervailing Global Bilateral Provision Antidumping duty safeguards safeguards Disallowed Number 9 5 0 5 Percent 12.2 6.8 0.0 6.8 No rules Number 18 30 45 4 Percent 24.3 40.5 60.8 5.4 Rules Number 47 39 29 65 Percent 63.5 52.7 39.2 87.8 Source: Author’s compilation. • PTAs are most likely to have no special rules concerning if the margin turns out to be higher than the WTO global safeguards. benchmark but less than or equal to that prescribed in • PTAs often have additional rules for antidumping and the PTA. bilateral safeguards. Multilateral rules encourage but do not mandate the • Most PTAs either have no specific countervailing duty application of an antidumping duty that is less than the provisions or have very weak CVD rules. A reason may dumping margin if a lesser duty would be adequate to be that the economic impact of subsidies is rarely con- remove the injury to the domestic industry. A lesser-duty fined to intra-PTA trade; subsidies affect global trade. rule or mandate in a PTA can provide a significant advan- Accordingly, there may be little economic justification tage to members. In the event that an antidumping action for their inclusion in a PTA. is taken by a country against a group of suppliers, some of which happen to be PTA members and others not, PTA partners will face a lower antidumping duty, even though What Provisions Are Included? the antidumping investigation might have found the same As seen in table 9.4, PTAs often include rules for trade dumping margin against all suppliers. remedies. Details about the key rules contained in the PTAs Under multilateral rules, definitive antidumping duties are presented next, following which I discuss how often are to be terminated within five years from their imposi- specific rules are included in the agreements. tion. Thus, PTAs that impose a shorter termination period Antidumping. Box 9.1, point C, outlines specific cate- on regional partners will give an advantage to exporters gories of rules related to antidumping that may be from those countries. Antidumping duties against exports included in PTAs. PTAs have modified four key require- from PTA partners will already have been phased out, ments in antidumping investigations: de minimis dumping while exports from non-PTA partners can continue to be margins, de minimis dumping volumes, the lesser-duty restrained by the duties. rule, and the duration of final antidumping duties. These four provisions (C.2–C.5 in box 9.1) modify Under WTO rules, an antidumping investigation is existing WTO antidumping provisions. By contrast, the to be terminated immediately if the dumping margin is establishment of a regional body that has the power to con- found to be less than 2 percent of the export price or if duct investigations, or the authority to review or remand the volume of dumped imports from a particular coun- final determinations of national authorities, is a unique try is less than 3 percent of imports. PTA provisions that innovation in PTAs. The PTA literature suggests that a specify higher de minimis dumping margins or higher regional institution can have a significant effect on the fre- de minimis volumes than the WTO benchmarks will quency of antidumping initiations and measures against treat PTA partners more favorably. This is because even PTA partners. The best-known example of such a regional though exports from PTA and non-PTA sources may institution occurs in Chapter 19 of NAFTA, which allows a be found to have the same dumping margin, the investi- binational panel to review the final antidumping or coun- gation against the PTA member will terminate, while tervailing duty determination made by the authority of the investigation against non-PTA sources will continue another NAFTA partner. 186 Thomas J. Prusa Box 9.1. Antidumping Template Box 9.2. Countervailing Duties Template A. Antidumping actions disallowed A. Subsidies: Export subsidies on agriculture prohibited B. Antidumping actions allowed, but with no specific B. State aid: Incompatible if it distorts competition provisions C. Countervailing duties C. Antidumping actions allowed, with specific provisions 1. Disallowed 1. Mutually acceptable solution 2. Allowed, but with no specific provisions 2. Different de minimis dumping margin 3. Allowed, with specific provisions 3. Different de minimis dumping volume a. Mutually acceptable solution 4. Lesser-duty rule b. Regional body or committee 5. Different duration of antidumping duty • Conducts investigations and decides on 6. Regional body or committee countervailing duties a. Conducts investigations and decides on • Reviews or remands final determinations antidumping duties • Other b. Reviews or remands final determinations c. Other Box 9.3. Global Safeguards Template There are differing views on the impact of this specific provision. Using a time dummy to control for the pre-PTA A. Rights and obligations under GATT Article XIX/ Safeguards Agreement retained versus post-PTA effect, Jones (2000) finds a statistically sig- B. PTA members excluded from global actions under nificant reduction in both U.S. antidumping filings against defined conditions Canada and Canadian antidumping filings against the 1. Grounds for exclusion a. Imports from the other party do not account for a United States after NAFTA took effect. Blonigen (2005), substantial share of total imports however, incorporates information on actual panel activity b. Imports from the other party do not contribute to serious injury or threat thereof and finds no evidence that binational reviews under 2. Definitions Chapter 19 of NAFTA affected the frequency of U.S. filings a. Substantial share or affirmative determinations against Canada and Mexico. • Imports are among the top five suppliers during the most recent three-year period The fact that the United States has refused to include a • Exports jointly account for 80 percent of the similar provision in any subsequent PTAs suggests that total imports of the importing country U.S. policy makers believe that the binational panels have b. Contribute importantly to serious injury • Growth rate of imports from a party is lower altered the pattern of protection. than the growth rate of imports from all Countervailing duties. In contrast with the provisions on sources antidumping, which address key statutory criteria, the pro- visions concerning countervailing duties include very few substantive rules (see box 9.2). As previously discussed, this excluded from a global safeguard action if those imports absence is probably related to the lack of limits on state aid do not account for a substantial share of total imports and and subsidies in PTAs. if they do not contribute to serious injury or the threat Two provisions regarding countervailing duties appear thereof (box 9.3). Most PTAs describe very precisely what noteworthy. First, some PTAs specify a series of steps that is meant by “substantial share� of total imports and “con- members are first required to take to try to reach a mutu- tribute importantly to serious injury.� For example, a num- ally satisfactory outcome before the countervailing duty ber of PTAs state that imports from a PTA partner do not investigation begins. Such provisions might lead to fewer constitute a substantial share of total imports if that part- disputes, although to date there is no empirical evidence ner is not among the top five suppliers during the most regarding their impact. recent three-year period. Similarly, imports from a PTA Second, provisions giving regional bodies the ability to partner do not contribute importantly to serious injury or conduct countervailing duty investigations or to review threat thereof if the growth rate of those imports during and remand final determinations have received significant the period of serious injury is appreciably lower than the attention, and there is some empirical support for the growth rate of total imports from all sources. hypothesis that they do reduce the number of disputes. The WTO Appellate Body has repeatedly rejected safe- Global safeguards. Provisions that allow PTAs to exclude guard actions from which PTA partners are excluded. As members from global safeguard actions have received con- noted earlier, these exclusions are highly contentious and siderable attention. Imports from PTA members may be have prompted non-PTA members to file multiple WTO Trade Remedy Provisions 187 dispute cases. In each case, the Appellate Body had ruled The special safeguard provisions in the PTAs are usually against the WTO member’s exclusion of PTA partners. applied to agricultural products and textiles and clothing, Bilateral safeguards. There are two types of bilateral which in many countries are the most difficult sectors to safeguards: transition safeguards and special safeguards. liberalize. Products or sectors that are hard to liberalize at Transition safeguards are designed to mitigate the costs the multilateral level are also hard to liberalize in PTAs and incurred as industries adjust to the preferential tariffs and require special safeguard treatment. often can only be imposed during the transition period. Special safeguards are provisions for products or sectors Hub-and-Spoke Pattern that are politically sensitive. PTAs often include extensive language defining when A review of the list of PTAs in table 9.2 shows that the pro- and for how long bilateral safeguards can be imposed liferation of PTAs has not happened by chance; rather, a (see box 9.4). Part of the reason for this detail might be small set of countries recurs as members of most PTAs. Put the absence of analogous WTO provisions specifying differently, there is a pronounced hub-and-spoke and the default behavior—unlike the case with the other trade cross-regional pattern in the PTAs in the sample. The remedies. largest constellations are grouped around the EU, EFTA, The role of regional bodies in bilateral safeguard actions and the United States (figure 9.1), but there are other active is noteworthy. Regional institutions might have a coordi- PTA players, including Mexico, with 9 PTAs, Singapore (6), nating function, serving for example, as clearinghouses for Australia (5), Chile (5), and Canada (4). information on emergency action. Alternatively, regional The prominent hub-and-spoke and cross-regional pat- authorities could conduct safeguard investigations or tern of the PTAs in the sample raises the question of review safeguard measures taken by national authorities. whether there are identifiable features in the trade remedy provisions negotiated by the hubs. The hypothesis is that each major hub negotiates according to certain key princi- ples. The rules and philosophy may vary across hubs, but Box 9.4. Bilateral Safeguards Template we expect consistency within a hub. Table 9.5 presents a summary of the provisions in each A. Safeguard measures disallowed PTA. Looking first at the antidumping provisions, we B. Safeguard measures allowed, but with no specific see that EFTA and the EU have a different philosophy than provisions C. Safeguard measures allowed, with specific provisions the other major hubs, notably the United States. Among 1. Conditions for application of safeguard the EU’s PTAs, 90 percent either prohibit antidumping a. Increasing imports cause serious injury to measures or limit their use. All of EFTA’s agreements either domestic industry b. During transition period, reductions in tariffs lead prohibit or limit the use of antidumping. By contrast, to increased imports and to serious injury almost 90 percent of U.S. PTAs contain no language on c. Other antidumping, and, indeed, it is clear that the United States 2. Mutually acceptable solution 3. Investigation is the least open of all hubs to the inclusion of antidumping 4. Application of safeguard measures provisions in PTAs. a. Only to the extent necessary to remedy serious This does not imply that any of the hubs consistently injury and facilitate adjustment b. Suspension of concessions, tariff reduction, or incorporates the same rules in each (or even most) agree- reversion to most favored nation rates ments. Table 9.6 shows whether a rule is included in the c. Other 5. Provisional measures majority of each hub’s agreements. When the hubs are 6. Duration and review of safeguard measures inspected more closely, there is little evidence that any of a. Less than four years’ duration them negotiate the same rules in all their PTAs. For b. Not allowed beyond transition period 7. Maintenance of equivalent level of concessions instance, for only five hubs—the EC, EFTA, Mexico, (compensation) Australia, and Canada—are there antidumping rules in 8. Suspension of equivalent concessions (retaliation) most of the agreements, and only in the EC and EFTA do 9. Regional body or committee a. Conducts investigations and decides on most agreements contain the same substantive provision. safeguard duties Moreover, this provision (regarding a mutually accept- b. Reviews or remands final determinations c. Other able solution) involves rather weak language. The other 10. Notification and consultation three hubs often have antidumping rules, but for no hub 11. Special safeguards is a particular rule included in most of the agreements. Thus, while it is fair to say that the Europeans are more 188 Thomas J. Prusa Figure 9.1. Hub-and-Spoke and Cross-Regional Arrangement of PTAs United EU EFTA Mexico Singapore Chile Australia Canada States EU EFTA– NAFTA U.S.– U.S.– U.S.–Chile SPARTECA NAFTA Turkey Singapore Singapore EEA Mexico– Mexico– ANZCERTA Canada– EFTA– Uruguay U.S.– New Chile Israel EU–Turkey Tunisia Morocco Zealand– Australia–U.S. Mexico– Singapore Korea, Rep.– Canada– EU–Tunisia EFTA– Northern U.S.– Chile Australia– Costa Rica Singapore Triangle Jordan Japan– Thailand EU–Syrian Singapore EU–Chile Canada– Arab Republic EFTA– Mexico– U.S.–Israel Australia– Chile Palestinian Nicaragua EFTA– Canada– Singapore Authority Singapore Chile EU– U.S.–Chile Switzerland– Mexico– Liechtenstein EFTA– Japan Australia– Australia– Morocco Singapore U.S. EU–South Mexico– EFTA– Israel AFTA Africa U.S.– Jordan CAFTA-DR EU–Palestinian Mexico– Authority EFTA–Israel EFTA U.S.– Bahrain EU–OCT EFTA– Mexico– Macedonia, Chile NAFTA EU–Morocco FYR Group of EU–Mexico EFTA– Three Croatia EU–Lebanon EFTA–Chile EU–Jordan EFTA EU–Israel EEA EU–Faeroe Islands EU– Macedonia, FYR EU–Egypt, Arab Rep. EU–Croatia EU–Chile EU–Andorra EU–Algeria Source: Author’s compilation. Note: AFTA, ASEAN Free Trade Area; ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Agreement; ASEAN, Association of Southeast Asian Nations; CAFTA-DR, Dominican Republic–Central America Free Trade Agreement; EEA, European Economic Area; EFTA, European Free Trade Association; EU, European Union; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement; OCT, Overseas Countries and Territories; SPARTECA, South Pacific Regional Trade and Economic Cooperation Agreement. open than the United States to incorporating antidump- Ironically, PTAs appear to have a more unified approach ing provisions in PTAs, there is not compelling evidence toward countervailing duties. The commonality seems to that the European PTAs are consistent in the precise rules be that extra provisions are not included in regional agree- they negotiate. ments. With the exception of EFTA and, to a lesser extent, Trade Remedy Provisions 189 Table 9.5. Cross-Tabulation of Contingent Protection Rules, by Hub (percent) Full sample EU EFTA Mexico United States Singapore Chile Australia Canada Provision (74 PTAs) (21 PTAs) (12 PTAs) (9 PTAs) (9 PTAs) (6 PTAs) (5 PTAs) (5 PTAs) (4 PTAs) Antidumping Disallowed 12.2 9.5 33.3 0.0 0.0 16.7 20.0 20.0 25.0 No rules 24.3 9.5 0.0 22.2 88.9 50.0 40.0 20.0 25.0 Rules 63.5 81.0 66.7 77.8 11.1 33.3 40.0 60.0 50.0 Countervailing duties Disallowed 6.8 9.5 16.7 0.0 0.0 0.0 0.0 0.0 0.0 No rules 40.5 52.4 0.0 22.2 66.7 66.7 40.0 40.0 50.0 Rules 52.7 38.1 83.3 77.8 33.3 33.3 60.0 60.0 50.0 Global safeguards Disallowed 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 No rules 60.8 76.2 83.3 11.1 11.1 66.7 0.0 60.0 0.0 Rules 39.2 23.8 16.7 88.9 88.9 33.3 100.0 40.0 100.0 Bilateral safeguards Disallowed 6.8 4.8 0.0 0.0 0.0 33.3 0.0 20.0 25.0 No rules 5.4 4.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Rules 87.8 90.5 100.0 100.0 100.0 66.7 100.0 80.0 75.0 Source: Author’s compilation. Notes: EFTA, European Free Trade Association; EU, European Union; PTA, preferential trade agreement. Table 9.6. Antidumping Template for Selected PTA Hubs Provision EU EFTA Mexico United States Singapore Chile Australia Canada A. Antidumping actions disallowed B. Antidumping actions allowed, but with no specific provisions X X C. Antidumping actions allowed, with specific provisions X X X X X 1. Mutually acceptable solution X X 2. Different de minimis dumping margin 3. Different de minimis dumping volume 4. Lesser-duty rule 5. Different duration of antidumping duty 6. Regional body or committee a. Conducts investigations and decides on antidumping duties b. Reviews or remands final determinations c. Other X X Source: Author’s elaboration. Note: EFTA, European Free Trade Association; EU, European Union; PTA, preferential trade agreement. Regional hubs are checked if at least 50 percent of their PTAs include the specific provision. Mexico, there is little support for prohibiting countervail- countervailing duties may depend on progress on subsi- ing duties or even including CVD provisions in regional dization, and it is likely that substantial progress on agreements. subsidization will only come via a multilateral format. Of the five hubs that tend to have countervailing duty Interestingly, only two hubs (EFTA and Mexico) rules—EFTA, Mexico, Chile, Australia, and Canada—four include the same provision in most of their agreements. have also negotiated restrictions on agricultural subsi- By chance, the one provision again involves rather weak dies (table 9.7). As discussed earlier, developments on language about a “mutually acceptable solution.� This 190 Thomas J. Prusa Table 9.7. Countervailing Duties Template for Selected PTA Hubs Provision EU EFTA Mexico United States Singapore Chile Australia Canada A. Subsidies: Export subsidies on agriculture prohibited X X X X X B. State aid: Incompatible if it distorts competition X X C. Countervailing duties 1. Disallowed 2. Allowed, but with no specific provisions X X X X 3. Allowed, with specific provisions X X X X X a. Mutually acceptable solution X X b. Regional body or committee • Conducts investigations and decides on antidumping duties • Reviews or remands final determinations • Other X Source: Author’s elaboration. Note: EFTA, European Free Trade Association; EU, European Union; PTA, preferential trade agreement. Regional hubs are checked if at least 50 percent of their PTAs include the specific provision. provision does not impinge on any decision criteria once a settlement panels have consistently ruled against excluding case is initiated. PTA partners from safeguard measures, they have done so A survey of global safeguards supports the view that the on quite narrow grounds. Conceivably, under a different hubs strive for a consistent approach across PTAs. Only set of circumstances, exclusion of PTA partners from safe- four hubs (Mexico, the United States, Chile, and Canada) guard measures could pass muster. include additional rules in most of their agreements, but Bilateral safeguards display more consistency across they tend to include similar provisions across PTAs. hubs than do the other three statutes. All eight major hubs All four of these hubs allow PTA members to be have bilateral safeguard rules (table 9.9). Indeed, there excluded from global actions (table 9.8). We also see some appears to be more commonality across the precise rules consistency in how the exclusions are incorporated: three of than for the other three trade remedy statutes combined. the four exclude on the basis of market share and three of Nevertheless, there again appears to be some evidence the four, on the basis of lack of impact. of distinct European and North American approaches As has been mentioned, the exclusion provisions have toward bilateral safeguard rules. Although there is consid- been the subject of repeated WTO disputes (Argentina– erable consistency across EC and EFTA PTAs, the precise Footwear; United States–Wheat Gluten; United States– rules differ from those in the other hubs. Line Pipe; and United States–Steel). In each case, the investigating authority included imports from all sources PTAs That Prohibit Trade Remedies in making the determination that imports were entering in such increased quantities as to cause serious injury to the Thirteen of the surveyed PTAs have abolished the applica- domestic industry, but, instead of applying safeguard tion to intraregional trade of one or more trade remedies. measures to all imports irrespective of their source, the What distinguishes these PTAs? Why have they been able to country invoking the safeguard action excluded its PTA abolish trade remedy measures against members’ trade? partners. In all four cases, the Appellate Body ruled against The depth of market integration incorporated in the the WTO member that included its PTA partners in the PTA is the leading candidate for explaining the abolition of safeguard investigation but excluded them in the applica- trade remedy measures, particularly antidumping. A com- tion of the safeguard measure.14 mon subsidy policy is one example of a policy reflecting The provisions excluding PTA partners from global deeper integration, and, as was previously discussed, there safeguard actions once again raise concerns about is some evidence that a common subsidy policy influences increased discrimination against nonmembers and the countervailing duty provisions. De Araujo, Macario, and welfare impacts of trade diversion. Although WTO dispute Steinfatt (2001) have argued that the implementation of Trade Remedy Provisions 191 Table 9.8. Global Safeguards Template for Selected PTA Hubs Provision EU EFTA Mexico United States Singapore Chile Australia Canada A. Rights and obligations under GATT Article XIX/Safeguards Agreement retained X X X X B. PTA members excluded from global actions under defined conditions X X X X 1. Grounds for exclusion a. Imports from the other party do not account for a substantial share of total imports X X X b. Imports from the other party do not contribute to serious injury or threat thereof X X X 2. Definitions a. Substantial share • Among the top five suppliers during the most recent three-year period X X X • Exports jointly account for 80 percent of the total imports of the importing country b. Contribute importantly to serious injury • Growth rate of imports from a party is lower than the growth rate of imports from all sources X X Source: Author’s elaboration. Note: EFTA, European Free Trade Association; EU, European Union; PTA, preferential trade agreement. Regional hubs are checked if at least 50 percent of their PTAs include the specific provision. common macroeconomic and microeconomic policies in decades, developing countries have become more frequent the EU reduced the social and political cost related to the users of antidumping and safeguard actions, and their removal of antidumping provisions. Wooton and Zanardi embrace of antidumping might make it difficult for them (2002) link the phasing out of antidumping measures to give up such remedies. Indeed, we find that only two with the creation of a single market. Taken together, these developing-economy PTAs, those between China and support the view that PTAs which go beyond the elimina- Hong Kong SAR, China, and between China and Macao tion of border measures or adopt common internal regu- SAR, China, have prohibited antidumping measures. lations are more likely to do away with trade remedy Table 9.10 brings together background data on those measures. PTAs that have abolished trade remedies. On average, such The adoption of a common competition policy might PTAs enjoy greater intra-PTA trade (both in value and also permit the elimination of certain trade remedies. It share) and are more likely to have a competition policy might, for example, make antidumping redundant. Of provision in the PTA and to have achieved deeper integra- course, the two explanations are not mutually exclusive, tion. There does not seem to be any difference with respect since a common competition policy may not make sense to the adoption of a common external tariff. PTAs that until a sufficiently high level of integration is achieved. have disallowed trade remedies and PTAs that retain the Hoekman (1998), however, dismisses the notion of a link instruments appear equally likely to have a common exter- between the adoption of a common competition policy nal tariff. and the abolition of antidumping in a PTA by arguing that the adoption of a common competition policy in a PTA is Trade and Protection Diversion often motivated by the need to manage the consequences of deeper integration.15 A concern about PTAs that prohibit trade remedies or add A third factor might be the development status of the additional rules regarding them is that having such provi- members of the PTA. Development status could proxy for sions does not guarantee that disputes will not occur. The a wide set of political-economy factors that might affect rules may mean that fewer cases will be filed against PTA the ability to prohibit trade remedies. Over the past two members, but that tells us little about what may happen to 192 Thomas J. Prusa Table 9.9. Bilateral Safeguards Template for Selected PTA Hubs Provision EU EFTA Mexico United States Singapore Chile Australia Canada A. Safeguard measures disallowed B. Safeguard measures allowed, but with no specific provisions C. Safeguard measures allowed, with specific provisions X X X X X X X X 1. Conditions for application of safeguard a. Increasing imports cause serious injury to domestic industry X X X X b. During transition period, reductions in tariffs lead to increased imports and to serious injury X X X X X c. Other X X 2. Mutually acceptable solution X X X X X X 3. Investigation X X 4. Application of safeguard measures a. Only to the extent necessary to remedy serious injury and facilitate adjustment X X b. Suspension of concessions, tariff reduction, or reversion to most favored nation rates X X X X X X c. Other 5. Provisional measures X X X 6. Duration and review of safeguard measures a. Less than four years’ duration X X X X X X b. Not allowed beyond transition period X X X 7. Maintenance of equivalent level of concessions (compensation) X X X X X 8. Suspension of equivalent concessions (retaliation) X X X X X 9. Regional body or committee a. Conducts investigations and decides on safeguard duties b. Reviews or remands final determinations c. Other X X X 10. Notification and consultation X X X X X X X X 11. Special safeguards X X X Source: Author’s elaboration. Note: EFTA, European Free Trade Association; EU, European Union; PTA, preferential trade agreement. Regional hubs are checked if at least 50 percent of their PTAs include the specific provision. other countries. The PTA provisions might simply lead to The issue is clearest regarding global safeguards. Fifteen fewer intra-PTA disputes but to just as many (or even PTAs allow members to be excluded from safeguard pro- more) cases against non-PTA members. tection.16 Once safeguard protection is enacted, another Bhagwati (1996) and Bhagwati and Panagariya (1996) form of PTA-induced discrimination is introduced. Even if argue that the elastic and selective nature of administered the PTA discriminatory tariff preferences are modest, protection makes “protection diversion� a particularly per- global safeguard duties often exceed 10 percent. Thus, the nicious and unforeseen consequence of PTAs. Adminis- secondary trade diversion stemming from safeguards may tered protection is elastic because it is arbitrary and the surpass the primary trade diversion resulting from tariff targets can be easily manipulated. So, apart from the dis- preferences.17 crimination introduced by preferential tariffs, PTAs can Protection diversion is also relevant for antidumping lead to more discrimination against nonmembers of the provisions. Unfair trade is poorly measured according to PTA through more frequent trade remedy actions against WTO rules; often, all exporters to a market might be found them: trade diversion begets protection diversion, which guilty of dumping. Over the past decade, it has become begets more trade diversion. increasingly rare for authorities to fail to determine that Trade Remedy Provisions 193 Table 9.10. Characteristics of PTAs That Have Disallowed Trade Remedies Disallowed Intra-PTA imports Value Common (billions Share of external Development Bilateral of U.S. trade tariff Competition PTA level Antidumping CVD safeguards dollars) (percent) (percent) chapter Integration ANZCERTA Developed X 10.1 6.9 X X Australia–Singapore Mixed X 9.9 X Canada–Chile Mixed X 4.7 1.4 X Canada–Israel Mixed X 3.9 1.1 X China–Hong Kong SAR, China Developing X X 202.4 21.1 X China–Macao SAR, China Developing X X 55.4 8.4 X EEA Developed X X 301.4 7.3 X X EFTA Developed X X 1.4 0.8 X X EFTA–Chile Mixed X X 0.3 0.2 X EFTA–Singapore Mixed X 3.1 0.8 X EU Developed X X X 2,419.0 61.1 X X X Mercosur Developing X 22.1 20.1 X X New Zealand– Singapore Mixed X 1.3 0.6 X Group average 233.5 10.2 15.40 11 (of 13) 6 (of 13) Group average, excluding EU 51.3 6.0 Average of other PTAs 28.9 3.1 13.10 43 (of 61) 4 (of 61) Source: Author’s compilation. Note: ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Agreement; CVD, countervailing duty; EEA, European Economic Area; EFTA, European Free Trade Association; EU, European Union; Mercosur, Southern Cone Common Market (Mercado Común del Sur); PTA, preferential trade agreement; SAR, special administrative region. unfair pricing exists.18 Unfair trade may be practiced by The annual number of antidumping disputes initiated suppliers within as well as outside the trade bloc. But, given by PTA members against PTA members (intra-PTA filings) that PTA rules on antidumping measures make it impossi- is calculated for each importing country. Because PTAs are ble (if the measures are abolished by the PTA) or more dif- enacted over a variety of years, I abstract from calendar ficult (if the PTA rules tighten discipline on their use) to time and instead consider time as measured relative to the apply that remedy to intrabloc members, antidumping year the PTA was enacted. For each PTA, year zero is the duties might be applied only to countries outside the bloc. year the PTA was enacted, year t – 1 is the year before estab- Antidumping duties are rarely less than 10 percent, so, as lishment, year t – 2 is two years before, t + 1 is the year after with global safeguards, it is quite possible that the second- establishment, and so on. This view of time allows us to ary trade diversion may surpass the primary trade diver- conveniently aggregate across PTAs. sion. Moreover, as Bhagwati (1996) has argued, the source In figure 9.2, the aggregate number of antidumping of injury might be truly rooted in the PTA preferences, but disputes is plotted in relation to each PTA’s inception. the PTA rules may result in the antidumping duties being The chart is compelling. During the years prior to the imposed on non-PTA sources. creation of the PTA, intra-PTA antidumping activity To get a sense of the extent of the diversion, I aug- increases. In the year the PTA is enacted (time = 0), the mented the PTA database with information on worldwide number of antidumping disputes drops sharply, and it antidumping activity since 1980. The earlier years of data remains much lower than the pre-PTA level. On average, were gathered to allow a better comparison of pre-PTA and during the 10 years prior to the PTA, there were 29.5 post-PTA filing patterns. Altogether, I have information on antidumping cases per year; by contrast, during the 4,805 antidumping cases initiated by WTO countries that 10 years following the PTA, there were just 23.6 cases belong to at least one PTA. per year. 194 Thomas J. Prusa Figure 9.2. Intra-PTA Antidumping Filings, Sample of 74 PTAs 50 29.5 cases per year 23.6 cases per year 45 PTA enactment 40 35 number of cases 30 25 20 15 10 5 0 –10 –9 –8 –7 –6 –5 –4 –3 –2 –1 0 1 2 3 4 5 6 7 8 9 10 time in years, relative to enactment of PTA Source: Author’s calculations. Note: PTA, preferential trade agreement. Although the results are persuasive, the analysis does Table 9.11. Antidumping Activity, by PTA Status not control for the possibility that antidumping activity in Target general—against both members and nonmembers of the PTA—may have fallen coincidentally with the enactment Non-PTA PTA Time member member Total of the PTA. That is, the analysis is not able to distinguish the PTA effect from some other trend. For instance, given Pre-PTA Number of cases 506 370 876 that the Uruguay Round was concluded in 1994 and that Percent 58 42 many PTAs were enacted in the mid-1990s, it is possible Post-PTA that the observed decline in antidumping activity might be Number of cases 3,554 375 3,929 a result of antidumping provisions in the Uruguay Round Percent 90 10 rather than of the PTA provisions. Total To control for this possibility, a difference-in-difference Number of cases 4,060 745 4,805 analysis is needed. The general idea is to identify the effect Percent 84 16 of a specific treatment by comparing the treatment group, Source: Author’s calculations. Note: PTA, preferential trade agreement. after treatment, with the same group before treatment and with some other control group. In this case, the “treat- ment� group is composed of countries that join a PTA, and the “control� group is made up of countries not in a PTA. tries, and only 10 percent were against PTA members. So, The comparison will therefore involve antidumping filings during the “before� period, the difference between PTA against PTA members and nonmembers both before and members and nonmembers is 16 percentage points, but in after each PTA is enacted. the “after� period, the difference is 80 percentage points, In table 9.11, the protection diversion effect is clearly yielding a difference-in-difference result of 64 percentage seen. Of the cases filed during the pre-PTA period, 58 per- points. The implied change in filing patterns is quite large: cent were against non-PTA countries and 42 percent were the results suggest that almost 1,300 antidumping cases, or against PTA members. By contrast, during the post-PTA about one-third of the 3,929 cases filed during the post- period, 90 percent of the cases were against non-PTA coun- PTA period, were diverted away from PTA members. Trade Remedy Provisions 195 The results clearly raise the specter of protection diver- give-and-take balancing of incentives that is at the crux of sion and more subtle forms of trade diversion. It is true the GATT or WTO agreements. that tariff preferences are small and might result in only modest amounts of trade diversion. This does not imply, Notes however, that trade diversion is not a concern; rather, it The author thanks Robert Teh and the staff of the WTO Secretariat for appears that other provisions of the PTA might be a greater their assistance in developing the trade remedy database. source of discrimination. 1. World Bank (2005) contains an excellent discussion of the myriad of effects associated with the proliferation of PTAs. 2. There is growing evidence that a high percentage of PTA tariff Conclusions preferences are never utilized. Brenton and Ikezuki (2005), Amiti and Romalis (2007), and Dean and Wainio (2009) discuss utilization for dif- Overall, the findings highlight the need to be vigilant about ferent countries, products, and time periods. Francois, Hoekman, and the impact of trade remedy provisions in PTAs. These pro- Manchin (2005) find a threshold preference margin of 4 percent below which preference margins are irrelevant, probably because of high com- visions vary greatly across PTAs and increase the overall pliance costs such as the cost of paperwork and red tape. complexity of the world trade environment. Pricing behav- 3. This third rationale does not explain why PTAs simply do not ior that is perfectly fine when the product is exported to prohibit the use of trade remedies against PTA members. After all, from Mastel’s (1988) perspective, the elimination of intraregional tariffs and one country may be sanctioned when the product is other border barriers also means that the raison-d’être for trade remedies exported to another country. is eliminated. Trade remedy provisions in PTAs have a mixed welfare 4. Bhagwati writes, “My belief that [free trade agreements] will lead impact. This ambiguous finding partly reflects the exis- to considerable trade diversion because of modern methods of protection, which are inherently selective and can be captured readily by protectionist tence of trade creation and trade diversion within the PTA. purposes, is one that may have been borne out in the [European Economic In some cases, PTA rules appear mostly to promote trade Community, EEC]. It is well known that the European Community has creation; at other times, the rules seem simply to constitute used antidumping actions and [voluntary export restraints] profusely to erect Fortress Europe against the Far East. Cannot much of this be a trade- trade diversion. diverting policy in response to the intensification of internal competition Some PTA rules make it easier to restrain intra-PTA among member states of the European Community?� (Bhagwati 1996, 37). imports. Such provisions may benefit global welfare by 5. Official statistics on other border measures are not widely pub- lished. From my review of WTO and U.S. reports, I doubt that there have mitigating trade diversion stemming from preferential been more than a few hundred disputes involving all other trade remedies tariffs. More often, PTA rules either prohibit contingent combined. protection against PTA members or make contingent pro- 6. The context should be taken into account in looking at the small tection harder to apply against members. This raises the number of safeguard initiations and actions, compared with the other trade remedy measures, since a safeguard action may involve multiple very real possibility that PTAs induce protection diversion, import sources. which, in turn, produces more trade diversion. 7. In this paper, European Union is regularly used when discussing There are other possible consequences of including the pre-1993 European Community. The European Community enacted many PTAs before 1993. trade remedy provisions in PTAs. For instance, PTAs might 8. The four traditional users now account for only about one-third serve as small-scale experiments that allow countries to of antidumping initiations and less than one-tenth of safeguard initia- better understand the practical effect of certain provisions. tions. Traditional users still account for about three-fourths of all coun- tervailing duty initiations. If parties find certain new rules attractive, those rules 9. The category “developed countries� refers to Australia, Canada, might be incorporated in future WTO negotiations. In this the EU, EFTA members, Japan, New Zealand, and the United States. All sense, PTAs might act as “beta� testing for the larger-scope other countries are classified as “developing.� WTO rounds. By giving members experience with new 10. Free trade areas account for a comparably large share of all noti- fied PTAs. provisions, PTA rules could streamline future WTO nego- 11. As of July 18, 2007, 157 PTAs in force had been notified to the tiations. WTO under either GATT Article XXIV or the Enabling Clause. Conversely, the trade remedy provisions in PTAs may 12. These percentages are very comparable to those for all notified PTAs. Of the 157 PTAs notified to the WTO under either Article XXIV or erode the market access that nonmembers thought they the Enabling Clause, 82 percent were notified under Article XXIV. had secured in prior WTO rounds. The erosion is not lim- 13. In the case of the EEA, the prohibition on antidumping applies ited to trade diversion stemming from preferential tariffs only to intraregional trade in goods that fall under chapters 25 to 97 of the WTO’s Harmonized Tariff System. In other words, antidumping measures but also comes about because of selective use of contingent can still be taken against agricultural and fishery goods. protection rules. As a result, PTAs may make it more diffi- 14. The key concept that underlines all these cases has been called cult for non-PTA members to agree to future WTO liberal- “parallelism.� In brief, parallelism prohibits any asymmetry in the appli- ization out of concern that the requisite quid pro quo by cation of safeguard measures. The Appellate Body’s decisions have been carefully worded; the panel has avoided ruling on whether GATT Article PTA members may not be realized. The complicated pat- XXIV permits the exemption from a safeguard measure of imports origi- tern of inclusion of these provisions threatens the delicate nating in a member of a free trade area. The decisions have all been 196 Thomas J. Prusa crafted so as to address only the specific facts and circumstances of the Bown, Chad P., and Meredith A. Crowley. 2007. “Trade Deflection case at hand. and Trade Depression.� Journal of International Economics 72 (1): 15. Hoekman (1998) defines deep integration as consisting of explicit 176–201. actions by governments to reduce the market-segmenting effect of differ- Brenton, Paul, and Takako Ikezuki. 2005. “The Value of Trade Preferences ences in national regulatory policies that pertain to products, production for Africa.� Trade Note 21, World Bank, Washington, DC. processes, producers, and natural persons. In practice, this will require Dean, Judith M., and John Wainio. 2009. “Quantifying the Value of US Tar- decisions (a) to regard a partner’s policies as equivalent (mutual recogni- iff Preferences for Developing Countries.� In Quantifying the Value of tion) or (b) to adopt a common regulatory stance in specific areas (har- Preferences and Potential Erosion Losses, ed. Bernard M. Hoekman, Will monization). Martin, and Carlos Alberto Primo Braga. Washington, DC: World Bank. 16. The PTAs are United States–Singapore; United States–Jordan; De Araujo, José Tavares, Jr., Carla Macario, and Karsten Steinfatt. 2001. United States–CAFTA-DR (Central America Free Trade Agreement plus “Antidumping in the Americas.� Journal of World Trade 35: 555–74. Dominican Republic); NAFTA; Mexico–Uruguay; Mexico–Northern Tri- Fiorentino, Roberto V., Luis Verdeja, and Christelle Toqueboeuf. 2007. angle; Mexico–Nicaragua; Mexico–Israel; Mexico–Chile; Group of Three; “The Changing Landscape of RTAs: 2006 Update.� WTO Discussion EU–Chile; Canada–Israel; Canada–Chile; Australia–United States; and Paper 12, World Trade Organization (WTO), Geneva. Australia–Thailand. Francois, Joseph, Bernard Hoekman, and Miriam Manchin. 2005. “Prefer- 17. Bown (2004) shows that the exclusions incorporated into the U.S. ence Erosion and Multilateral Trade Liberalization.� World Bank Eco- steel safeguard created significant discrimination and that developed and nomic Review 20 (2): 197–216. developing countries differed in their ability to adjust to the tariffs. Hoekman, Bernard. 1998. “Free Trade and Deep Integration: Antidump- 18. For the United States, the U.S. Department of Commerce finds ing and Antitrust in RTAs.� World Bank Policy Research Working unfair pricing in more than 95 percent of all firms investigated; Blonigen Paper 1950, World Bank, Washington, DC. (2003) finds that the average U.S. dumping margin exceeds 60 percent. Jackson, John H. 1997. The World Trading System: Law and Policy of Inter- national Economic Relations. Cambridge, MA: MIT Press. Jones, Kent. 2000. “Does NAFTA Chapter 19 Make a Difference? Dispute References Settlement and the Incentive Structure of US/Canada Unfair Trade Petitions.� Contemporary Economic Policy 18: 145–58. Amiti, Mary, and John Romalis. 2007. “Will the Doha Round Lead to Pref- Mastel, Greg. 1998. Antidumping Laws and the U.S. Economy. Armonk, erence Erosion?� IMF Staff Papers 54 (2): 338–84. NY: M. E. Sharpe. Bhagwati, Jagdish. 1996. “Regionalism and Multilateralism: An Overview.� Prusa, Thomas J. 2001. “On the Spread and Impact of Antidumping.� New Dimensions in Regional Integration, ed. Jaime de Melo and Arvind Canadian Journal of Economics 34 (3): 591–611. Panagariya. Cambridge, U.K.: Cambridge University Press. ———. 2005. “Antidumping: A Growing Problem in International Bhagwati, Jagdish, and Arvind Panagariya, eds. 1996. The Economics of Trade.� World Economy 28: 683–700. Preferential Trade Agreements. Washington, DC: AEI Press. Tharakan, P. K. M. 1995. “Political Economy and Contingent Protection.� Blonigen, Bruce A. 2003. “Evolving Discretionary Practices of US Economic Journal 105: 1550–64. Antidumping Activity.� NBER Working Paper 9625, National Bureau Viner, Jacob. 1950. The Customs Union Issue. New York: Carnegie Endow- of Economic Research, Cambridge, MA. ment for International Peace. ———. 2005. “The Effects of NAFTA on Antidumping and Countervailing Wooton, Ian, and Maurizio Zanardi. 2002. “Trade and Competition Pol- Duty Activity.� World Bank Economic Review 19 (December): 407–24. icy: Antidumping versus Anti-Trust.� Discussion Paper in Economics Bown, Chad P. 2004. “How Different Are Safeguards from Antidumping? 02-06, University of Glasgow, Glasgow, Scotland. Evidence from U.S. Trade Policies toward Steel.� Working Paper, Bran- World Bank. 2005. Global Economic Prospects 2005: Trade, Regionalism, deis University, Waltham, MA. and Development. Washington, DC: World Bank. 10 PRODUCT STANDARDS Jean-Christophe Maur and Ben Shepherd With the decline of many tariffs in recent decades as a by protectionist considerations. Product standards often result of successive waves of unilateral, regional, and mul- represent a quasi-regulatory means of promoting an tilateral liberalization, nontariff measures have become important public policy objective, such as environmental important barriers to trade. The multilateral trading sys- protection, consumer safety, food quality, or compatibility tem, and an increasing number of preferential trade between different technical norms and standards. agreements (PTAs), has been paying greater attention to This chapter addresses issues arising from the treatment product standards as an important type of potential non- of product standards in PTAs. A unique feature of prefer- tariff barrier. ential liberalization is that it offers an alternative means, The World Trade Organization (WTO) agreements complementary to multilateral efforts, of diminishing set out general rules for the design and implementation through bilateral mutual recognition efforts the costs asso- of product standards, but it has been in regional con- ciated with compliance with standards. This prospect, and texts that the main instruments of liberalization in this the presumption that standards are not necessarily estab- area—harmonization and mutual recognition—have been lished for protectionist purposes, suggest that preferential deployed. These instruments, however, are not free of risk liberalization can be a force for good. The effect on third regarding their compatibility with the broader aim of countries and the lack of international coordination may, multilateral liberalization. Preferential agreements involv- however, carry significant costs. ing both developed and developing countries (North- The next section provides a brief overview of the rela- South PTAs) can lead to specifications that are overly tionship between product standards and trade.1 Following complex or burdensome from the point of view of many that, we consider in greater detail the policy measures developing countries. Indeed, they could be perceived as available for dealing with standards in PTAs, in particular, locking these countries out of vital international markets mutual recognition and harmonization, and look at (Baldwin 2000). It is therefore important for policy mak- broader issues of institutional coordination and regional ers and trade policy practitioners to understand the issues cooperation. Finally, we address the interface between that product standards raise in a regional integration con- regional and multilateral approaches to standards. text and, in particular, the challenges developing countries can face in dealing with foreign standards as they become Product Standards and Trade: An Overview increasingly integrated into the world economy. Although product standards are widely recognized as It is common to distinguish three broad groups of stan- having potential effects on regional and international dards, based on the types of activities to which they apply, as trade flows, their motivation and impact are fundamen- defined by the International Organization for Standardiza- tally distinct from those of traditional trade barriers such tion (ISO): product standards, process standards, and man- as tariffs. Whereas tariffs are mainly protectionist in intent agement systems (see ISO 2006, 2008). Product standards and effect, product standards are not always protectionist have to do with the characteristics of goods or services, in from either point of view. Indeed, even when standards particular with respect to aspects such as quality, safety, and have protectionist effects, they are rarely motivated solely fitness for purpose. A simple example of a product standard 197 198 Jean-Christophe Maur and Ben Shepherd is the limit, set by the U.S. Food and Drug Administration 2000; Ganslandt and Markusen 2001). Two of the most (FDA), of one part per million of methyl mercury (a poten- common reasons why standards might be necessary relate tial toxin) in fish marketed for human consumption. Process to spillover effects (externalities) from certain activities standards apply to the conditions under which goods or and to information asymmetries between buyers and sell- services are produced, packaged, or refined. An example is ers. These effects are clearly relevant for standardization at the production of dairy products without hormones: milk the national level, but they also set the scene for regional from hormone-free cows is indistinguishable from milk and global coordination on standards. from cows treated with hormones, but only farms that meet Externalities. Packaging is an example of an externality. a particular set of production conditions can describe their In the absence of any rules or standards, producers and milk as “hormone free.� Management systems assist organi- consumers do not directly pay the cost of disposing of zations in running their operations and create a framework packaging materials after the product has been bought and within which the requirements of product and process unpacked. These materials must be taken away to a landfill, standards can be consistently met. Management systems where they will break down over time, perhaps releasing are often referred to as meta-standards; an example is the pollutants into the environment. A standard setting out ISO 9000 series of quality standards (see box 10.1). biodegradability requirements for packaging materials can A well-functioning standards system will usually incor- help reduce this problem and limit the negative environ- porate elements of all three groups of standards. This mental spillovers from useful economic transactions. chapter focuses mainly on the first group—the product Social objectives such as environmental protection and standards that apply to everything from primary produce public health are common externality-based rationales for and agricultural products to sophisticated manufactured the development of product standards. goods such as electrical equipment (box 10.2). The Another example of externalities relates to network effects remainder of this section briefly discusses the rationale for and interoperability (see, for example, David and Greenstein product standards before moving to a more detailed dis- 1990; Gandal and Shy 2000; Barrett and Yang 2001). cussion of how they are designed and implemented. Economic Rationale for Product Standards Box 10.2. Proliferation and Growing Importance of Why are product standards necessary at all? Broadly speak- Product Standards ing, standards can be seen as a way of bringing the out- Product standards represent an important and constantly comes of a decentralized market economy more closely growing set of regulatory measures that have potential into line with social objectives that may not otherwise be effects on trade. The available evidence suggests that these achieved (see, for example, Maskus, Otsuki, and Wilson standards cover all sectors, from simple agricultural products (Disdier, Fontagné, and Mimouni 2008) to the most complex electronic goods (Moenius 2007). Many countries produce their own standards but are also increasingly involved in efforts aimed at regional or international Box 10.1. What Are Meta-Standards? standardization. Before looking in more detail at the regional dimension of standards and their trade effects, it is useful to get an idea of the phenomenon’s extent and development Closely associated with standards are the quality assurance over recent years through a few stylized facts: processes that users employ to effect and manage compliance. Indeed, a “new� category of standards has • The International Organization for Standardization evolved that defines and describes such meta-systems. comprises 3,000 working groups and committees. Meta-standards do not concern a specific product or • The ISO has issued about 15,000 international standards production process but, rather, set overall principles and (WTO 2005). rules to be followed in broad areas of activity. Examples • A survey of national product standards in a selection of include the ISO 9000 series of standards for quality member countries of the Organisation for Economic management systems and the ISO 14000 series for Co-operation and Development (OECD) identified a total environmental management systems. In certain industries, of nearly 300,000 documents (Moenius 2005). The compliance with these standards is itself becoming a Perinorm database on which Moenius drew contains commercial necessity, alongside more traditional product about 650,000 standards from 21 countries (WTO 2005). and process standards. For instance, in the area of food • In 1975 there were only 20 Europe-wide standards, but by safety, meta-standards include hazard analysis and critical 1999 the number had grown to 5,500 (Moenius 2005). control point (HACCP) standards, good manufacturing • The number of Europe-wide standards in the agricultural practices (GMP), good agricultural practices (GAP), and and textiles and clothing sectors grew at a rate of nearly ISO 22000 on food safety management systems. 20 percent per year over the period 1995–2003 Source: Henson and Jaffee 2007. (Shepherd 2006). Product Standards 199 The high-definition DVD (HDDVD) market has recently while minimizing costs to the extent possible. Standards seen the emergence of Blu-ray as the de facto dominant may be mandatory or voluntary. Mandatory standards standard over the alternative HDDVD system. DVD play- (also referred to as technical regulations) must be met ers can usually display only one of these two formats, and by firms as a matter of law, and penalties are set for non- the more consumers buy a particular type of player, the conforming products. Compliance with voluntary stan- greater is the incentive for firms to release a wider range of dards remains a matter of commercial choice for individ- movies in the corresponding format. Each consumer’s pur- ual firms. In practice, both types of standards exist side by chase therefore has an externality—a spillover effect—in side, although the bulk of standards-related activity in most the sense that it increases the value for every other cus- countries now consists of voluntary standards. Mandatory tomer of having a player of the same format. At the same standards tend to be mostly confined to core public health time, consumers may be reluctant to purchase a player of and consumer safety areas, such as requirements governing either format while there is uncertainty as to which will food and medicines. This coexistence is also apparent at become dominant in the future. Standardization makes it the regional level—for example, in the European Union’s possible to overcome this reluctance and develop the (EU) “New Approach� to harmonization. (See European spillovers more completely than would be possible if each Commission 2000 and WTO 2005 for a discussion.) equipment manufacturer adopted its own technology Given the coexistence of mandatory and voluntary stan- standard. dards, the responsibility for meeting those standards is Information asymmetries. Information asymmetries refer increasingly shared between the public and private sectors. to the availability of product-related information to buyers In most cases, standards are set up in a complex environ- and to sellers. For example, a consumer wants to buy a car ment characterized by interplay between private and public that is safe and has a certain level of fuel efficiency, but until interests and agents (Casella 2001). In the United States, for she actually acquires it and starts using it, it is very difficult instance, the FDA is a public body (part of the executive to tell to what extent it meets those requirements. The car- branch), and its standards are mandatory and enforce- maker is in a much better position to know the car’s true able through the executive branch and the courts. At the characteristics. An unscrupulous manufacturer might adver- same time, the American National Standards Institute tise a car as being safe and fuel efficient when in fact it is (ANSI), a nonprofit organization, develops and promulgates not.2 How can the consumer go about finding the desired voluntary standards in a wide range of areas. Sometimes the type of car when she knows that advertisements may not boundaries between public and private organizations can always be truthful? Standards can offer a way out of this set become blurred. ANSI, for example, is a private law body, of difficulties: if the consumer sees that an independent test- but its members include government agencies, as well as ing authority has certified that a car meets a particular safety private sector organizations. standard, then she can be confident about its characteristics. In this public-private partnership, the state often finds Objectives such as quality and fitness for purpose are itself at an informational disadvantage with respect to the commonly pursued through these kinds of standards. Alter- private sector when it comes to designing and implement- natively, standards can be seen as a way of reducing the costs ing particular standards. It is therefore important for policy a consumer must bear in searching for the product that best makers to find the right balance between public and private matches her preferences (see, e.g., Jones and Hudson 1996). initiatives and to ensure that they work together. For Producers also need information on consumer tastes to instance, in the European Union’s New Approach to harmo- manufacture profitable products. Gathering such informa- nization, the public sector takes the lead in enacting manda- tion can be costly, especially for overseas firms that may be tory core standards in certain sectors, but private standards unfamiliar with local preferences and practices. Standards organizations such as the European Committee for Stan- can help make this process simpler and less costly, to the dardization (CEN) remain responsible for developing extent that they summarize a set of characteristics consid- detailed voluntary standards setting out particular ways in ered to be valuable in the local market. (See Swann, Tem- which products can be designed and built to meet the ple, and Shurmer 1996 and Moenius 2005 for a discussion mandatory standards (European Commission 2000). of this mechanism.) Of course, drafting the documents containing product standards is only part of the process. National standards systems require supporting infrastructure to ensure the Designing and Implementing Product Standards effective implementation of standards. Figure 10.1 presents Given the role that standards play in the economy, they the infrastructure components of a “farm-to-fork� national should be set up to promote the desired social objectives quality system. All standards systems need a solid basis in 200 Jean-Christophe Maur and Ben Shepherd Figure 10.1. Elements of a Standards Infrastructure National quality system international farm-to-fork system national quality infrastructure system sanitary surveillance ministry accreditation technical regulations ISO/IEC 17021 IAF, ILAC and ISO/IEC guide 65 shrimp farm certification good practices ISO 9000 • products ISO/IEC 17025 processing • processes process HACCP ISO 14000 deep freezing ISO 22000 ISO, CODEX standardization transport ISO/IEC 17025 testing inter comparisons laboratories proficiency tests trace- product ability metrology (shrimp) BIPM • calibration product with certificate • reference materials customer Source: Sanetra and Marbán 2007. Note: The figure represents the infrastructure in developed countries. Accordingly, the term “national� infrastructure is used, but in practice, although all the elements of the standards infrastructure are necessary, functions can sometimes be distributed among countries metrology—the establishment of accurate, reliable, and assessment of their equipment and processes. Accredita- traceable measurements. Without reliable means of meas- tion of testing laboratories serves this purpose. urement, product standards become meaningless, since Standardization involves a complex and technically there is no way of assessing a product’s performance in sophisticated set of organizations and processes. Depend- relation to the set of benchmarks set out in a standard. ing on their level of development, countries, especially A sound basis in metrology makes it possible to demon- developing countries, may need to be selective when they strate whether specified requirements relating to a product adopt standards infrastructure. Regional cooperation have in fact been met. This process, known as conformity offers a way for countries to share the burden and spread assessment, can be completed either through a supplier’s the costs of setting up standards infrastructure. The cre- declaration of conformity, in which the seller states that ation of regional agencies for metrology, testing and con- the product meets certain requirements, or through a formity assessment, accreditation, and standardization is third-party conformity assessment. In the latter case, an an example of a concrete approach to cooperation. Use of independent laboratory tests whether the product meets foreign conformity assessment mechanisms is another pos- certain requirements, and if it does, the laboratory certifies sibility, where domestic capacity is weak. conformity.3 In some cases, conforming products may receive the right to display a particular logo or label, such Regional and International Dimensions as “CE� for certain European standards, or they may be of Standardization included in an official register of conforming goods. For third-party conformity assessment to be reliable, testing The rationales for standards discussed above also apply in a laboratories must themselves be subject to independent cross-border setting. Asymmetries of information between Product Standards 201 consumers and producers are even more serious in a regional bodies deal with standardization. Among them traded-goods context because of the ineffectiveness on the is CEN, which has a regionwide standardization func- international plane of signaling mechanisms that work tion in the EU. Its members are the national standards domestically, such as firm reputations built up over a long bodies of 30 European countries; CEN itself is a private period. In addition, a number of important public policy nonprofit organization. Although the standards CEN goals—concerning environmental protection, for example— issues are voluntary, member organizations are required are now recognized as having regional and even global to issue its standards as national standards and to with- dimensions. In these and other areas, countries are increas- draw any inconsistent national standards. CEN’s stan- ingly keen to cooperate on standards, at least to some extent, dards are therefore often referred to as being “European at the transnational level, through PTAs or the WTO. standards� or “harmonized European standards� because The ISO is just one among many transnational entities they are expected to apply with equal force in all mem- that issue standards. It is a network of national standards ber countries. institutes, and, in keeping with the private-law character of many of these institutes, it is a nongovernmental organiza- Cost Effects of Standards tion. Its standards are, accordingly, private and voluntary. Another example of international standardization is the Policy makers need to be aware of the costs entailed by Codex Alimentarius Commission, a public organization standards and of their potential to distort regional and created as a joint venture between the World Health global trade relations. The implementation of standards Organization (WHO) and the Food and Agriculture policies might lead to discrimination between national and Organization of the United Nations (FAO). Unlike the ISO, foreign suppliers or between different categories of foreign the commission deals with only one area of standardiza- suppliers. From the point of view of foreign exporters, tion, food safety.4 Although its standards do not have direct product standards in an importing country can sometimes legal force, they serve as the basis for legally binding rules represent an additional source of costs (Maskus, Otsuki, in many countries (see box 10.3, below). and Wilson 2005). Moreover, even though national stan- In addition to international structures such as the dards may be legitimate, their multiplicity and diversity ISO and the Codex Alimentarius Commission, some can mean duplication of market access costs that may be inefficient from a global perspective. These effects can put foreign competitors at a disadvantage and generate— intentionally or not—a form of protection for domestic Box 10.3. Do Voluntary Standards Have industries.5 Coordination among countries in implement- Cost Effects, Too? ing their standards policies may yield harmonized policies, reducing the cost of market access while preserving regula- At first glance, it might seem that only mandatory standards would have measurable cost impacts. After all, tory objectives. A potential difficulty with this kind of manufacturers are, in principle, free not to comply with coordination—discussed in more detail below—is its voluntary standards if they so choose. Legally speaking, assumption that it is optimal for the same standard to they are not required to pay additional costs to access a particular market. In practice, the situation is not that apply across a wide range of countries. In fact, however, straightforward. If compliance with voluntary standards is a different economic and social conditions may call for dif- commercial imperative, even if not a legal one, we might still expect to see some evidence of cost effects. Is this, in ferent standards. fact, the case? What kinds of costs can compliance with product stan- As it turns out, there is substantial evidence that voluntary dards impose on manufacturers? It might, for example, be standards do indeed affect trade flows, which is consistent with their having a significant impact on cost. Two studies necessary to redesign a product, in whole or in part. New that look into voluntary standards find significant impacts of machinery may have to be purchased, or a new production standards on trade—negative in the work of Czubala, process may have to be set up. All these adaptations are Shepherd, and Wilson (2009), and a mix of sector-specific negative and positive results in the case of Moenius (2005). In associated with increased fixed costs of production addition, Shepherd (2007) finds evidence that voluntary (including sunk ones), in the sense that they largely involve standards affect export diversification in partner countries, a one-off payment rather than a recurring expense. which may be indicative of an effect on fixed, not just variable, costs of production. These findings, taken together, In some cases, adaptations to deal with product stan- suggest that although compliance with voluntary standards dards can affect the level of variable production costs, may not be necessary as a matter of law, it is still of sufficient commercial importance to produce important links with as well. If the new production process uses more expen- production costs and trade flows. sive inputs, or if the new machinery is more costly to run, the unit cost of production will increase. It might 202 Jean-Christophe Maur and Ben Shepherd also be necessary to formally demonstrate compliance reorganization of firms and of the sector around more with a particular standard, in which case additional test- efficient production methods. (See, for instance, Maertens ing and certification procedures might be needed. These and Swinnen [2009] on how the Senegalese vegetable procedures also increase variable production costs. export chain reacted to the tightening of European stan- Maskus, Otsuki, and Wilson (2005) collected data from dards.) The overall impact on producers is therefore nearly 700 firms in 17 developing countries, as part of an ambiguous and depends on the relative strength of these effort to better understand the cost effects of foreign stan- two effects in particular cases. dards. The authors’ findings are in line with the types of Because of this theoretical ambiguity, empirical work to effects discussed above. In their sample, the fixed costs of assess the trade impacts of standards is difficult to inter- compliance with foreign standards are, on average, nearly pret, since there is no simple way of distinguishing results 5 percent of firm value added. Moreover, increased compli- that are consistent with theory from those that are not. ance investment is associated with a small but significant Although empirical work emphasizes the relative size and effect on variable production costs. significance of these different effects, it remains difficult to In assessing the supply-side effects of standards, the dis- identify them separately in a satisfactory manner. Even tinction between fixed and variable costs is important. apart from this issue, empirical work is hampered by the Recent advances in the theory of international trade need to rely on very rough proxies to measure the costs (Chaney 2008) suggest that higher variable costs primarily associated with standards: only in rare cases is a direct affect trade by reducing exports per firm among the small measure of restrictiveness possible (see box 10.4). subset of firms that already exports to foreign markets. A comprehensive economic assessment of product Higher fixed costs, by contrast, tend to force some firms out standards couched in terms of a measure of aggregate of export markets entirely, thus altering the range of prod- welfare could not, of course, be limited to the supply side. ucts exported, or the set of foreign markets served, or both. It would need to take into account the extent to which a The costs and benefits of standards will also depend on particular standard reduces social costs (e.g., less danger- their dynamic effects in the long run. For instance, Jaffee ous products) or creates social benefits (e.g., compatible (2003) shows how the horticultural industry in Kenya has systems) and thus brings the economy closer to its welfare used changing European regulations as a stimulus to inno- optimum. Such a comprehensive analysis is at the core of vation, competitive repositioning, and industrial upgrading. the concept of a regulatory impact assessment (see, e.g., Diaz Rios and Jaffee (2008) find that developing-country Hahn and Litan 2005) and could easily be extended to firms responded differently to stricter aflatoxin regulations the regional context by explicitly considering costs and in the EU. The new rules only exacerbated the commercial benefits with cross-border dimensions. However, to take difficulties of some exporters, but they offered an opportu- up the issue of internalization and aggregate welfare nity for others to upgrade their production techniques and would be to go beyond the scope of this chapter. Hence, gain additional market share. The reallocation of resources the next sections will deal exclusively with observable over time from small and relatively inefficient firms to larger, trade effects, and largely with the supply-side effects dis- more efficient ones is associated with gains in sectoral cussed above. productivity—a kind of technological upgrading that A final observation, however, points toward what are holds significant development promise. likely to be efficiency costs. The implementation of national standards policies is, by and large, rarely coordi- nated with trading partners, and it would be surprising if Consolidation: Trade Effects of Product Standards this shortcoming did not lead to more complex and costly This brief discussion shows that the overall economic standards systems than are strictly necessary. In particular, impacts of product standards are difficult to assess. From developing countries may not have available to them the the producer’s point of view, there are two opposite forces best standards technology, and they may consequently at play in the short term: possible cost increases stemming implement inefficient policies—policies that are too costly from the need to adapt production processes and demon- or do not achieve their objectives. Furthermore, because strate conformity, and possible cost savings through the national policies, even if they incorporate some form of transmission of market-specific information that would regulatory impact assessment, are decided without consid- otherwise be costly to obtain. Over the longer term, there eration of cross-border effects and externalities, it is likely is also potential for technological progress induced by that duplication and incompatibility of standards and standards. At the industry level, standards may lead to a practices will create systemic costs. Product Standards 203 Second, agreed disciplines on standards in PTAs should Box 10.4. Inventory Methods versus Direct Measures be viewed in the context of the overall bargain leading to of Restrictiveness the agreement, including the mercantilist objectives that may motivate the negotiations. It is not a given that provi- Disdier, Fontagné, and Mimouni (2008) employ the so- called inventory method of measuring standardization, sions on standards in a PTA necessarily improve on exist- using data from the Trade Analysis and Information System ing national policies; countries may negotiate preferential (TRAINS) database to tally the number of standards that market access in exchange for nonoptimal standards poli- potentially affect trade in agricultural products. (The TRAINS database is maintained by the United Nations cies set by their trading partners. (See Fink, ch. 18 in this Conference on Trade and Development [UNCTAD].) From volume, for a discussion in the context of intellectual prop- these tallies, they construct coverage ratios—the percentage of products (tariff lines) in the sample that is subject to erty rights.) product standards. In the aggregate, they find that higher Third, PTAs offer a specific mode of trade liberalization coverage ratios—that is, a larger inventory of standards— when it comes to standards: the recognition (often mutual) tend to be associated with reduced exports from developing countries to member countries of the Organisation for of standards and procedures. This approach to liberaliza- Economic Co-operation and Development. Trade between tion is specific in the sense that it is difficult to achieve it OECD countries appears to be relatively unaffected. This multilaterally. The mutual recognition of standards and result would be consistent with a situation in which adjustment costs are higher or more difficult to cover in the procedures requires a case-by-case approach that only developing world than in rich countries. In line with the seems practical in PTAs consisting of a select number of results of Moenius (2005), Disdier, Fontagné, and Mimouni trading partners with reasonably similar economic and (2008) report that this aggregate impact tends to mask considerable cross-sectoral heterogeneity: in some cases, social characteristics. A case-by-case approach is also often standards have a negative impact, consistent with the required in light of the political sensitivities generated by dominance of cost increases, whereas in others, they have a positive impact, consistent with the dominance of cost- changes in product standards in areas such as environmen- reducing information effects. tal protection and public health. There are very few examples of applied work in which it The importance of these effects is demonstrated by the has proved possible to compare directly the substantive content of standards, rather than just their numbers. Otsuki, prominence product standards have been receiving in Wilson, and Sewadeh (2001) do this for the standards PTAs: Budetta and Piermartini (2009) find that 58 of the governing acceptable levels of aflatoxin in African groundnut 70 PTAs in their sample contain provisions on product exports to the EU. They assess restrictiveness directly in terms of the maximum permissible aflatoxin content of groundnuts, standards. In this section, we briefly examine the econom- measured in parts per billion. Their gravity model results ics of preferential standards liberalization and then discuss suggest that tighter aflatoxin regulations are associated with fewer groundnut exports from African countries to the EU, two of the main approaches that have been adopted within which is consistent with increased compliance costs. As a regional forums and that have also influenced develop- rough order of magnitude, they find that a proposed new EU ments at the multilateral level: harmonization and mutual aflatoxin regulation would result in trade flows 63 percent lower than those that would be observed under a less recognition (Nicolaïdis 2001). restrictive Codex Alimentarius standard. This very strong result needs to be nuanced, however, in light of the conflicting case study evidence presented by Diaz Rios and The Economics of Preferential Standard Liberalization Jaffee (2008). Liberalization of standards in PTAs is not necessarily de jure discriminatory, and when it is, it is not necessarily de facto discriminatory. This point is clearly made, for instance, by Baldwin, Evenett, and Low (2009) for several Dealing with Standards in PTAs deep integration dimensions. Thus, a PTA on standards Having set out the general context, we now turn to look does not translate automatically into a preferential liberal- more closely at the particular issues raised by product stan- ization of standards; in some instances preferential liberal- dards in PTAs. ization is equivalent to multilateral liberalization. We First, it should be clear from the discussion above that return to this issue in the last section of this chapter. regional standards can produce economic effects similar to As we shall see below, there are different routes to liberal- those associated with national standards. The essential ization of standards in PTAs. Some of these—mutual recog- dynamic is one of insiders and outsiders: members of a nition agreements and harmonization with standards of the PTA may benefit from a liberalization of product standards preferential trading partner that differ from international that, at the same time, creates costs for countries outside standards—are distinctly preferential. A parallel can be the agreement. drawn with preferential liberalization of tariffs: standards, 204 Jean-Christophe Maur and Ben Shepherd like tariffs, create compliance costs for business that liberal- is dissipated by costly activities that are required to satisfy ization can reduce. A key difference is that standards do not the standards. As shown in more detail in chapter 3, there is raise revenue (see “A note on nontariff barriers� in Baldwin, no possible negative impact on the importer. Preferential ch. 3 in this volume). Or, rather, standards are not supposed liberalization is unambiguously beneficial in this respect to bring in revenue, but there are instances in which fees because it generates an economy for both exporter and must be paid—for example, for the certification process or importer of not having to pay the costs of meeting the for laboratory examination costs.6 These processes of certi- standards—even though trade diversion may still occur, fication and accreditation may create rents that are captured since a preference is created. Thus, there is a favorable pre- by public or private domestic interests. In this case, the sumption toward discriminatory liberalization of standards analysis of preferential liberalization is similar to that of the in the absence of rent capture. tariff case, with the exception that the tariff revenue is replaced by the rent capture. In strict welfare terms, if Harmonization of Standards domestic interests are benefiting from this implicit tax on importers, it is possible that preferential liberalization may As noted above, the need for producers to comply with generate the negative effects due to trade diversion. multiple, different standards to access different markets Even when undertaken on a discriminatory basis, could act as a barrier to trade. One response to this problem preferential liberalization may nevertheless contribute to is to seek to remove the differences between national kick-starting a process of reform of national standards standards through a process of harmonization. In this con- policies. By introducing more rigorous processes in stan- text, harmonization means the convergence of national stan- dards infrastructure operations, PTAs can make it more dards toward a common set of requirements (box 10.5). The difficult to use standards for purposes other than con- question of whether it is optimal for a particular set of coun- sumer protection, even when the standards are not inter- tries to adopt the same standard depends on the balance nationally harmonized.7 between two effects: the potential for increased trade thanks It is also the case that the gap between the prices of the to reduced cost multiplicity, and the likelihood that differ- good inside the importing nation and the exporting nation ent national preferences and resource endowments will Box 10.5. Facilitating Market Access: Harmonization, Equivalence, and Mutual Recognition Part of the international effort around standards and technical regulations aims at reducing the overall burden on traders. Coordinating policies to make them more “alike,� or more “compatible,� is one way of reducing the costs of compliance. There are three ways of achieving this: harmonization, equivalence, and mutual recognition. Harmonization is straightforward; it simply means replacing two or more rules or procedures with a single one. Nevertheless, the term can be somewhat misleading because there are degrees of harmonization, involving rules alone, procedures alone, both rules and procedures, or even higher-level objectives only (essential requirements), as in the EU’s New Approach. The aim of the other processes, equivalence and mutual recognition, is to preserve diversity of rules and procedures, provided that “equivalent� or “like� objectives are met. Nicolaïdis and Egan (2001, 456) offer the following definition: Recognition refers to the acceptance of equivalence of selected foreign standards or regulations. Mutual recognition establishes the general principle that if a product or a service can be sold lawfully in one jurisdiction, it can be sold freely in any other participating jurisdiction, without having to comply with the regulations of these other jurisdictions. This involves a transfer of regulatory authority from the host jurisdiction to the home jurisdiction from which a product, a person, a service or a firm originates. The “recognition� involved here is of the “equivalence,� or at least “compatibility� of the counterpart’s regulatory system; the “mutual� part indicates that the reallocation of authority is reciprocal and simultaneous. Thus, one difference is that mutual recognition can be broken down into two components: the “recognition� of the equivalence of a partner’s regulatory system, and the “mutual� aspect, which indicates that both parties simultaneously recognize the other (Nicolaïdis and Shaffer, 2005). Equivalence can thus in principle be asymmetric and unilateral as in U.S. PTAs, where the approach is that parties can decide to accept the other party’s regulations as equivalent. Equivalence can be achieved if the outcome of two standards is identical, even though the means of reaching it differ. Veggeland and Elvestad (2004) quote the example of hard cheese, the manufacturing of which in Australia requires the heating of milk, whereas Switzerland achieves the same levels of pathogen destruction through other production methods, using raw, unpasteurized milk. Equivalence is thus the acceptance that a third party’s standards or procedures, in effect, fulfill national requirements. Mutual recognition normally refers to the acceptance of certification of a partner country. It can also be used to refer to agreements on specific sectors or on specific instances of application, or to agreements between specific partners of the “principle of mutual recognition� (Nicolaïdis and Egan 2001). Product Standards 205 interact to produce different optimal regulations in each closed economy. For example, Norway and Zambia might Box 10.6. Trade Effects of Harmonization: Empirical Evidence be able to realize some trade gains by adopting the same standards for environmentally friendly packaging materials, There is substantial empirical evidence to the effect that but their technological and enforcement capacities differ harmonized standards are often associated with increased greatly, and it could be very difficult to ensure that the stan- trade among harmonizing countries. For instance, Henry dard is in fact implemented. Moreover, the differences in de Frahan and Vancauteren (2006) find that harmonization of standards across Europe tends to boost trade among income levels and relative land abundance might suggest European Union (EU) members: bilateral exports in sec- that Norwegians and Zambians could legitimately have dif- tors with harmonized food regulations are, on average, 253 percent higher than in nonharmonized sectors. The ferent preferences in relation to the trade-off between the tariff equivalent of nonharmonization ranges from 73 to cost and the environmental properties of packaging materi- 97 percent, depending on the sector. These findings are als. These issues are extremely complex to resolve but need consistent with the cost-based analysis presented above: a single, harmonized standard avoids the cost multiplicity to be kept in mind when examining harmonization efforts. that arises from multiple standards, making it easier for (See Bhagwati 1996 for a review.) producers to access an expanded market within the Harmonization of standards can take place in two ways. harmonization zone. Of course, this dynamic must be nuanced in certain cases because of the possibility that Unilateral harmonization occurs when one country or information effects will work in the opposite direction: group of countries simply adopts a standard prevailing in Moenius (2005) shows that the effect of harmonization on trade is not always positive, even for the harmonizers, and another country. More common is concerted harmoniza- suggests that the reason could be the dominance of tion, whereby countries work together to identify a set of information effects in some sectors. requirements that is acceptable to all parties. Concerted For countries outside the harmonization zone, the picture is not generally so rosy. Empirical work suggests harmonization can be a lengthy and uncertain process, that there is considerable scope for a kind of trade diversion requiring extensive negotiations among the parties con- effect: the cost reductions implicit in harmonization can cerning every standard in each jurisdiction. The more lead to a switch in demand to a relatively high-cost supplier within the harmonization zone, to the detriment of lower- divergent the parties’ interests and approaches to standardi- cost suppliers elsewhere. In the case of the EU, Baller (2007) zation, the more difficult it is likely to be to negotiate a set of and Chen and Mattoo (2008) show that harmonization harmonized standards. Successful harmonization therefore under the EU’s New Approach directives can sometimes have a negative impact on trade with third countries. It is tends to involve countries at reasonably close levels of devel- plausible that the EU’s harmonization program involves opment, and with some broad similarities in their prefer- a significant degree of harmonizing-up to a higher stan- dard, which would also tend to have negative trade ences and their general approaches toward regulation. consequences for excluded countries. Developing countries Trade effects of harmonization: Insiders versus outsiders. are more likely to be affected than developed ones, With respect to trade effects, the distinction between coun- presumably because adaptation costs are higher in a technology-scarce environment. tries inside the harmonizing region and those outside is crucial. Ordinarily, the cost-reducing effects of harmoniza- tion accrue primarily to firms within the region where standards are harmonized. Foreign exporters still must sat- isfy the importing region’s standards, in addition to what- harmonization but that harmonization with international ever requirements may prevail in their home country. They standards generally leads to dominance of the scale effect. therefore face some level of cost multiplicity (albeit lower In other words, the net effect of harmonization on the than the levels that pertained prior to harmonization). exports of excluded countries tends to be negative for In the case of “harmonizing-up� by the preferential area— regional harmonization but positive for international that is, the adoption of a stricter standard than prevailed harmonization. prior to harmonization— it may even be more difficult to One way of dealing with the difficulties created by this access some markets (box 10.6). insider-outsider dynamic is the hybrid approach adopted At the same time, however, harmonization allows for- by the members of the Asia-Pacific Economic Coopera- eign exporters to realize economies of scale by granting tion (APEC). APEC is a particularly heterogeneous them access to a larger market. Exporters have to meet only regional grouping that includes developed countries such one type of standard for the whole region, and this reduces as Australia, Japan, and the United States; developing the fixed costs of compliance. The balance between cost countries at various income levels (e.g., China, Peru, and and scale effects is an empirical issue that must be resolved Thailand); and transition countries (the Russian Federa- case by case. The currently available evidence suggests that tion and Vietnam). Such a diverse membership would the cost effect sometimes dominates in the case of regional seem to suggest that concerted harmonization could be a 206 Jean-Christophe Maur and Ben Shepherd particularly long and uncertain process. Similarly, the associated with different standards by reducing multiplic- presence of a number of large countries makes it unlikely ity, mutual recognition allows each country to maintain that unilateral harmonization by all members except one potentially different standards but requires each country would be a possibility. to accord equal treatment to goods produced in partner Consistent with its commitment to “concerted unilater- countries, even though standards might be different. If alism� and “open regionalism,� APEC has therefore adopted South Africa and Nigeria decide to harmonize standards, an intermediate approach. APEC member economies com- they adopt a single set of requirements that applies equally mit to increasing harmonization of their own national in both countries. If they agree to mutual recognition of standards with international standards, such as those standards, South African products that conform to local issued by the ISO; they identify priority sectors in which standards can be put on the Nigerian market, even if they harmonization should be pursued first; and they undertake do not comply with Nigerian standards (and vice versa). to participate actively in the work of international stan- Of course, it is possible for recognition to be unilateral dards bodies. As with other APEC commitments, member (equivalence) rather than mutual; for example, Nigeria economies must make public progress reports (individual may decide to treat products conforming to South African action plans) each year. Helble, Shepherd, and Wilson standards as equivalent to those conforming to its own (2007) show that substantial progress appears to have been standards. made on these points. An advantage of mutual recognition of standards over Trade effects of international harmonization. The empiri- harmonization is that once the principle has been agreed cal literature discussed above examined the impact of on, it is not necessary to engage in long and complex nego- regional harmonization on outside countries. In addition, tiations over each individual standard; the rule simply there is now a growing body of empirical evidence suggest- applies across the board. National standards agencies con- ing that harmonization with international standards can tinue to go about their work as usual, the only difference mitigate the costs that foreign exporters might otherwise being that nonconforming products from foreign markets face.8 Czubala, Shepherd, and Wilson (2009) find that EU might now appear on the domestic market. standards that are not harmonized with international In fact, however, mutual recognition can be extremely (ISO) norms have a negative and significant impact on difficult to implement among countries with markedly dif- exports of African clothing to the EU; internationally har- ferent social preferences or with fundamentally different monized EU standards have no statistically significant approaches toward regulation. Although the rule is rela- impact. tively easy to apply in practice, it is usually difficult for These results are confirmed by Shepherd (2007), who countries to reach agreement as to whether it should be focuses on the fixed-cost effects of standards. A 10 percent applied at all. Mutual recognition can be seen as creating a increase in the total number of EU standards leads to a risk that one country’s standards might be undermined by 6 percent reduction in the variety of products exported by another country’s different—perceived as lower—standards. non-EU partner countries. This finding is consistent with In an environment of mobile capital, the fear is that a sharp the idea that standards tend to generate fixed costs that difference in standards might create an incentive for pro- exporters must pay to access foreign markets. By contrast, a duction to relocate from high-standard to low-standard 10.0 percentage point increase in the proportion of those countries. This would, in turn, provide a motive for standards that are harmonized with ISO standards is asso- authorities in the high-standard country to lower stan- ciated with a small (0.2 percent) but significant increase in dards in a “race to the bottom.� partner-country export variety. These results suggest that Although there is considerable debate as to the empiri- convergence of regional standards to international norms cal relevance of this dynamic, there is no doubt as to its can be an effective way of limiting the potential for negative political relevance. (See Drezner 2006 for a review.) As a trade effects in excluded countries, in particular, developing result, mutual recognition is generally only seen among rel- countries. atively similar countries. European countries, for instance, adopted a form of mutual recognition rule as set out in the Cassis de Dijon decision: products that comply with Mutual Recognition of Standards or Conformity mandatory regulations in one European country cannot Assessments usually be prevented from accessing markets in other Euro- Another way of dealing with the cost issues raised by pean countries. But even within a relatively homogeneous divergent national standards is mutual recognition of grouping such as the EU, the idea of adopting a type of standards.9 Whereas harmonization eliminates the costs mutual recognition rule for services trade—the “country of Product Standards 207 origin� principle—proved so controversial that it had to be A Review of Standards in PTAs largely shelved (box 10.7). How commonly is each of the above approaches actually Another form of mutual recognition applies to con- applied in practice? Recent work by Budetta and Piermar- formity assessments. Under such a regime, countries tini (2009) provides some useful first results (see table 10.1). agree to recognize the results of testing and certification The authors analyze the texts of 70 regional and bilateral procedures conducted in other countries, even though PTAs, of which 58 contain some kind of provision related there is no harmonization or mutual recognition of the to product standards. Interestingly, harmonization appears underlying standards themselves. For example, if the EU to be much more common than mutual recognition of and Australia agree to mutual recognition in the area of standards: 29 agreements provide for harmonization of conformity assessments, European exporters can have mandatory standards, and 25 provide for harmonization local laboratories certify their compliance with Australian of voluntary standards. By contrast, only 5 agreements standards. Since the question of recognition is limited to include mutual recognition of voluntary standards, the performance of scientific tests and the certification of and 15 provide for mutual recognition of mandatory results, this procedure should be considerably easier for standards. countries to negotiate than full-scale mutual recognition Mutual recognition is the most frequent approach for of standards. Fundamentally, all that is required is that conformity assessment, perhaps because it is easier to the recognizing countries have a certain level of trust con- achieve mutual recognition of conformity than of stan- cerning the quality of testing and certification authorities dards. Instead of implying equivalence of regulatory objec- overseas. tives, which is a sensitive issue in many cases, recognition of conformity assessment only looks at whether the ways in which tests are performed and certification is granted are equivalent in the countries concerned. Legally, there is an important difference between the Box 10.7. Trade Effects of Mutual Recognition: Empirical Evidence structural patterns of the two types of obligation. Agree- ments to pursue harmonization can sometimes impose a The available empirical evidence on the effects of mutual relatively small number of up-front obligations, and it is recognition is much more limited than in the case of common for the parties to commit to ongoing negotiations harmonization. Chen and Mattoo (2008) look at the effects with a view to harmonization. The devil is thus in the of European mutual recognition agreements (MRAs) with other (non-EU) countries that cover conformity assessments. details, since the extent of harmonization that in fact takes They find that conformity assessment MRAs uniformly place depends on the outcome of a long and complex promote trade between the parties. Baller (2007) confirms that result for a wider range of countries. An and Maskus process. Moreover, Budetta and Piermartini (2009) point (2009) find similar evidence, using firm-level survey data, out that the majority of agreements with harmonization and suggest that their results would be consistent with a obligations include the EU as a party, and most of them stronger beneficial effect on developing-country exporters from MRAs than from international harmonization of require harmonization with EU standards. This dynamic standards. reflects both a long-term dynamic within the EU and the Chen and Mattoo (2008), however, demonstrate that fact that most of these agreements involve much smaller the impact of conformity assessment MRAs on third countries depends crucially on the nature of the rules of and less developed economies. Thus, the figures cited origin that accompany them. Rules of origin set out the above also reflect the influence of different “models� of conditions under which a product is treated as originating in a particular country, for the purposes of applying the treatment of standards and technical regulations in prefer- specific regime under an agreement; in this case, a rule of ential agreements. origin would preclude third countries from accessing the mutual recognition regime. For instance, whereas a product in country A, a member of the agreement, could have local conformity assessment accepted as valid in other Table 10.1. Prevalence of Harmonization and Mutual countries participating in the agreement, the product from Recognition in Preferential Trade Agreements a third country would have to be certified following the practices of the country where the product is being sold. (number, out of sample of 58 PTAs) MRAs with relatively open rules of origin tend to be Voluntary Mandatory Conformity trade promoting for all countries, but the presence of standards standards assessment restrictive rules of origin can reverse that effect. In addition, Amurgo-Pacheco (2006) finds evidence that conformity Harmonization 25 29 25 assessment MRAs can have negative trade impacts on Mutual recognition 5 15 39 excluded developing countries. Source: Adapted from Budetta and Piermartini 2009, table 2. 208 Jean-Christophe Maur and Ben Shepherd Mutual recognition, by contrast, tends to impose Some PTAs go even further in their treatment of prod- stronger obligations up front: countries sometimes imme- uct standards and incorporate institutions designed to diately commit to give full force to each other’s standards, make the process of standard setting and administration although this may in some cases be limited to particular work more smoothly between trading partners. Of the sectors. The extent to which individual agreements result agreements reviewed by Budetta and Piermartini (2009), in the removal of multiple standards-related barriers is 34 provide for some kind of regional administrative body therefore an empirical question that needs to be examined to deal with the administration of standards systems, and in detail in each case. 24 include a dispute settlement mechanism. Interestingly, Many regional and bilateral PTAs contain additional 22 agreements have provisions relating to technical assis- provisions related to the design and management of tance. This last point is consistent with the increasing regional standards systems (see table 10.2). Usually, a clus- trend toward North-South integration agreements and ter of obligations concerning the transparency of standards suggests that the parties are aware of the asymmetric and their administration forms a key component. For challenges that can arise when trading partners at differ- instance, Budetta and Piermartini (2009) find that 21 of ent levels of development pursue integration bilaterally. It the 58 PTAs dealing with product standards impose a is impossible, however, to draw any general conclusion as requirement of prior notification on the parties. This to the effectiveness of these provisions because the way means that new standards, or modifications of existing they are implemented is so important. Legal provisions of ones, must be notified prior to their entry into force; in agreements tell us about the intentions of their drafters many cases, there is also an obligation to allow time for but not about the actual implementation. comments. Another common example of a transparency Still, there is evidence in Europe, in the Southern Cone obligation is the creation of a national contact point or the Common Market (Mercosur, Mercado Común del Sur), establishment of a consultation system. Twenty of the and in the Andean countries that PTAs generate actual agreements surveyed by Budetta and Piermartini (2009) changes in standards policies. Aldaz-Carroll (2006) reports contain such obligations. that by 2004, Mercosur had developed about 370 regional voluntary standards and 407 regional technical regulations and sanitary and phytosanitary (SPS) measures. The Table 10.2. Content of Preferential Trade Agreements Andean Community has harmonized technical regulations Relating to Provisions on Standards (percent) for 31 agricultural products representing about 60 percent of intraregional trade. An additional piece of indirect evi- Lesser, based Budetta and on 24 case Piermartini, dence of the attention paid to implementation issues is Provision studies based on 58 PTAs provided by Budetta and Piermartini (2009), who examine WTO disputes concerning technical regulations and find Reference to WTO TBT Agreement 86 52 that a number of them involve PTA partners, suggesting Harmonization of standards, that such disputes are linked to the agreements they have technical regulations, signed. and conformity assessment 47 46 What do existing regional experiences tell us about the Harmonization to regional ingredients of a successful approach to standards? Aldaz- technical regulations Carroll (2006) reviews the evidence from Asia and Latin and standards 34 45 America, and concludes that the following aspects are Equivalence of technical regulation and standards 33 26 crucial: Recognition of conformity assessment 77 67 • Building trust among parties to the PTA Transparency 80 52 • Building regulatory capacity among parties Joint committee or regional body 80 62 • Focusing on simplification, transparency, and dynamism Dispute settlement 80 50 in the standards-upgrading process Technical assistance 47 38 • Allowing for gradual reform where appropriate Metrology 14 29 • Promoting mutual recognition of conformity assess- Source: Lesser 2007; Budetta and Piermartini 2009. ments as a first step Note: TBT, technical barriers to trade; WTO, World Trade Organization. • Identifying priority sectors for harmonization. Product Standards 209 Lessons from the Standards Provisions in PTAs recognition. Some U.S. agreements use the concept of equivalence, which can be pursued unilaterally by each The examination of specific provisions in agreements partner without reciprocity. The EU on the other hand, confirms that countries seek to use PTAs to help with focuses on mutual recognition or harmonization. It is also market access and implementation of compliance with unclear which model is superior. Where the advantage lies foreign standards. Chapter 11, which takes up the prac- is most likely to vary depending on the partner country tice of product standards in PTAs, provides additional (and the similarity of regulatory preferences). Second, empirical information. The establishment of institutions agreements between partners at similar development levels to help with implementation is widespread and is often (North-North or South-South) are more likely to lead to coupled with transparency requirements (box 10.8). deep integration measures such as mutual recognition; These features suggest that regional institutions can con- Lesser (2007) cites the example of the Trans-Pacific Strategic tribute to making standards less burdensome to trade. Economic Partnership Agreement between Brunei Darus- Access to information, trust, and capacity building all salam, Chile, New Zealand, and Singapore. Third, technical matter. The need to coordinate national, regional, and assistance dimensions are more prevalent in North-South multilateral efforts on standards might, however, place a agreements, where upgrading of capacity may be needed significant strain on resources and capacity in some before regulatory reform takes place. poorer countries. Finally, the agreements contain examples of good prac- The treatment of standards in PTAs is highly dependent tice. Among them are, obviously, support for the multilat- on the type of agreement. First, agreements involving the eral WTO framework and, to the extent that it is consistent EU and the United States each propose different models, with the adoption of optimal standards, international har- and it is still unclear which advantages each partnership monization efforts. Advance notification to trading part- may or may not yield. Take, for instance, the question of ners is another. The approach favored by the United States Box 10.8. How Small ASEAN Countries Manage to Access Certification and Accreditation Services The absence of internationally recognized public laboratories need not act as a binding constraint on the implementation of mutual recognition agreements, provided that the private firm can either use a private company or obtain access to the testing infrastructure of neighboring countries. Permitting the efficient operation of private testing service providers (local and foreign) can enable export-ready firms to access testing services at low cost. In the presence of internationally recognized third-party certifiers, the absence of a national accreditation agency or office need not be a serious constraint. The major export markets will accept certification from these third-party certifiers. Although the small size of the market in the Lao People’s Democratic Republic and in Cambodia might discourage foreign testing companies from establishing local branches and offering services across a wide range of sectors, at least one foreign third- party certification company has begun operating in Cambodia. Intertek, an internationally recognized testing company, has opened an office in Phnom Penh and is offering testing and certification services to exporting companies. The service conducts tests for companies that export garments to the U.S. and EU markets. Since the advent of Intertek, inspection costs have declined. Intertek has no contact with the government of Cambodia, but it works closely with foreign buyers. This example illustrates the importance of private sector third-party certifiers in enabling exporters to obtain the necessary documentation to prove they meet international standards. East Asia is developing a network of calibration laboratories with traceability to physical measurement standards at the national level or to the internationally recognized national physical standards of another country. Most members of the Association of Southeast Asian Nations (ASEAN) have both privately and publicly owned laboratories that are accredited by a government accreditation service. The original six ASEAN members (Brunei Darussalam, Indonesia, Malaysia, the Philippines, Singapore, and Thailand) have entered into MRAs on laboratory accreditation with other ASEAN members. In the newer members of ASEAN, the CLMV countries (Cambodia, Lao PDR, Myanmar, and Vietnam), individual laboratories engage in MRAs with foreign counterparts; this is the case for Intertek in Cambodia. In countries without a national accreditation agency, the government may contract a foreign accreditation body to carry out national accreditation activities on its behalf. Within ASEAN, the Brunei Ministry of Development has an agreement with the Singapore Accreditation Council (SAC) that includes the use of SAC accreditation of laboratories and of certification and inspection bodies and that, in addition, provides for training to build up capacity in Brunei Darussalam. A government can also allow foreign accreditation bodies to provide services directly to laboratories in a foreign country without any formal arrangement with the government. Although this would work for many markets, it would not assist with improved market access to the EU under any of the EU’s MRAs because, under those agreements, the exporting country is required to endorse the accreditation service. Source: World Bank 2008. 210 Jean-Christophe Maur and Ben Shepherd to request justification for decisions to refuse equivalence • Article 4 of the SPS Agreement suggests that members also stems from the right kind of principles, as does the seek mutual recognition agreements on SPS measures.11 promotion of suppliers’ declarations of conformity. Finally, the establishment, in the context of agreements, of institu- The fact that regional aspects are so explicitly men- tions such as specialized committees that meet regularly tioned in the TBT and SPS Agreements reflects the nature seems to offer a good venue for dialogue and exchange of of standards, which must meet regulatory objectives such information, enabling learning by doing, monitoring of as protecting the environment. Meeting these regulatory implementation, and defusing of possible disputes. objectives, by definition, creates barriers to trade. The role of the WTO is to help minimize any excessive burden on trade created by such regulations and to ensure that no dis- Regional Standards in a Multilateral World crimination arises from them. This requires that the need Standards in the WTO are disciplined by the Agreement on for regulation, as well as the principle of nondiscrimina- the Application of Sanitary and Phytosanitary Measures tion, be taken into account (Trachtman 2003).12 (SPS Agreement) and the Agreement on Technical Barriers In some instances, necessity may be compatible with to Trade (TBT Agreement). The WTO agreements do not regional interventions. At the same time, it has to be acknowl- force countries to adopt standards, but they do provide dis- edged that there is an immediate tension between the risk of ciplines to be adopted when applying standards. For exam- discrimination created by any agreement between a select ple, the SPS Agreement states a specific preference for the few and the pursuit of legitimate objectives of protection Codex Alimentarius. One very specific dimension of the through regional standards interventions. two agreements is, therefore, to aim for a balance (some There is a certain lack of clarity as to how WTO disci- will see it as a tension) between countries’ autonomy to plines apply to regional TBT and SPS measures because of pursue domestic regulatory objectives and the objective of the need to interpret the relation between the provisions of nondiscrimination. the General Agreement on Tariffs and Trade (GATT)—in particular, Article XXIV, on preferential trade agreements, and Article I, on the most favored nation (MFN) obliga- WTO Disciplines on Regional Standards tion—and the provisions of the SPS and TBT Agreements Unlike other trade policies in which regional and bilateral themselves. Trachtman (2003) is of the opinion that the PTAs may be seen as an exception to multilateralism, the WTO language does not require harmonization or mutual SPS and TBT Agreements incorporate the regional dimen- recognition within PTAs. He notes that a specific area of sion into their provisions: uncertainty relates to mutual recognition agreements, in particular, the potential that they create for discrimination • Articles 4.1, 9.2, and 9.3 of the TBT Agreement address toward nonparticipating trading partners; unlike the General PTA issues. Agreement on Trade in Services (GATS), the TBT and SPS • Article 2.4 of the TBT Agreement recognizes that in Agreements do not suggest that recognition be offered on some instances, international standards may not be an open basis (i.e., that third-party countries be allowed to appropriate means of fulfilling certain objectives obtain recognition). A too strict application of the MFN because of “fundamental climatic, geographical, or fun- principle, however, could prevent legitimate liberalization damental technological problems.� of trade in PTA through harmonization and recognition. • Article 2.7 of that agreement seeks to promote recognition of other members’ equivalence of technical regulations. Regional Standards Systems and Multilateralism • The TBT Agreement suggests that members seek mutual recognition agreements on conformity (Article 6.3).10 As we saw earlier, regional standards and bodies are an • The TBT Agreement refers to international and regional important layer of the international trade system and are standards-setting bodies (Article 4.1), as well as regional recognized as such in the WTO texts. In what way can certification bodies (Articles 9.2 and 9.3), although not regional initiatives be compatible with multilateral, to regional standards. nondiscriminatory, and open-trade objectives? • Article 13 of the SPS Agreement refers to the applicabil- The first contribution of PTAs might be their role in ity of the agreement to regional bodies. enforcing multilateral disciplines, providing, in a way, • The SPS Agreement recognizes that national boundaries some redundancy in enforcement. Multilateral and prefer- are not necessarily relevant for the application of SPS ential agreements have different enforcement mechanisms measures and refers to regional conditions (Article 6). that may strengthen each other. Concessions in PTAs may Product Standards 211 also be perceived as more valuable to trading partners than beyond deeper harmonization, to the definition of common multilateral concessions, and infringement of commit- procedures (e.g., risk management, testing protocols). ments as thus more costly. PTAs offer more possibilities for Mutual recognition, or, in WTO language, equivalence of “soft� dispute resolution through dialogue and informa- measures, in standards or in the testing and certification of tion sharing at an expert level that can help defuse many trade partners, is a facilitating practice that is implemented disputes. In some cases, too, PTAs offer more stringent through specific ad hoc agreements (e.g., the EU–U.S. agree- arbitration rules than the WTO, foreshadowing, for ment on mutual recognition of conformity assessment) or instance, the repeal of offending standards. In some agree- as part of PTAs.14 This process is essentially bilateral or, in ments, such as the North American Free Trade Agreement some rarer cases (the EU, Mercosur) regional, since relatively (NAFTA), the possibility of resorting to both the WTO and intensive cooperation among the parties is required if it is to PTA dispute settlement mechanisms is explicitly men- be acknowledged that foreign standards or testing systems tioned (Budetta and Piermartini 2009). are equivalent to national ones and go toward meeting the Second, PTAs can offer scope for further autonomous same regulatory objectives. liberalization in the area of standards by promoting harmo- Finally, regional cooperation can also involve the provi- nization in areas not explicitly covered by the WTO, sion of technical assistance and transfer of knowledge (see contributing to the elimination of national standards, or box 10.8, above). Such a level of cooperation can be easier promoting provisions that are stricter than WTO language. to attain and can be more flexible than in international In the first instance, given that standards are generally agreements. designed to be MFN—that is, the standards specification is the same for products from all origins—regional standards Regional Standards: Stumbling Blocks, design or discipline can complement multilateral disci- or Building Blocks? plines (Lesser 2007; see also the further discussion below).13 Mercosur provides an example of harmonization of Like two sides of the same coin, the characteristics of pref- regional standards. In addition, as Trachtman (2003) notes, erential cooperation that could favor multilateral liberal- agreements among countries with more homogeneous reg- ization are often the same that could hinder it. At the heart ulatory preferences may render the reduction of standards of this paradox is the fact that cooperation creates a “spe- barriers easier. In this second instance, principles similar to cial relationship� to which other trading partners lack those professed by the WTO are adopted, but in a more access. Two cases can arise. The first is that this preferential binding way. relationship is used to raise the standards applied to the In some instances, national systems may not be adapted rest of the world (without a justifiable change in the regu- to guarantee the correct application of a given standard. latory objectives). This can occur when a regional standard The economic reason for such a situation is the existence of that is used as the basis of harmonization is protectionist in cross-border externalities or cross-border economies of intent. An example of a protectionist regional standard is scale. In such circumstances, transnational cooperation Mercosur’s prohibition of imports in wine in barrels may be called for. The SPS Agreement mentions regional (Nofal 2004). conditions with respect to diseases and pests that may The second, more common, occurrence is when the require cross-country coordination to ensure control or preferential relationship, while leaving standards outside eradication. (This is an example of an externality.) Small the PTA unchanged or improved and trade within the par- countries may also lack the resources to develop adequate ties to the agreement liberalized, provides preferential institutions to manage standards. In particular, accredita- access to products from within the PTA but not to third tion and metrology bodies may not be available in some parties, leading to trade diversion. This is what happens countries, or it might not make economic sense to have with “nonopen� MRAs that are confined to members of the such services in small markets, and access to regional facil- agreement only and cannot be extended to third countries. ities in a partner country is therefore needed (an example The economic incentives so created are the same as for of economies of scale). Thus, international cooperation other forms of preferential access; they generate market might help implement a “division of labor� among coun- access rents that may act in the future as stumbling blocks tries according to their specific comparative capacities in to further liberalization. certification (Aldaz-Carroll 2006). There is also the related question of whether regional A related point is the capacity of PTAs or regional institu- standards systems produced by preferential agreements tions to help with the implementation of standards and, may, as they grow in size, create disincentives for progress more generally, with the sharing of experience. This may go toward greater international harmonization, as the cost of 212 Jean-Christophe Maur and Ben Shepherd changing standards rises in relation to the marginal benefit Similarly, transparency measures often echo those con- of increased market access. tained in the WTO, such as advance notification of new The cost of switching to more open international regulations. standards and certification can be increased in two ways. Although PTAs seem to promote aspects of multilateral First, problems of compatibility and complexity can arise standards disciplines, these same agreements are pushed by when a country belongs to many standards systems. Sec- large economies with relatively idiosyncratic and distinct ond, regional systems may alter incentives for further lib- approaches. Budetta and Piermartini (2009) conclude their eralization. Once a country joins a regional standards investigation on a note of caution, observing that harmo- system, changing to an international system may not be nization seems generally favored over mutual recognition, attractive enough because the costs would be too high especially for standards setting. (Mutual recognition of and the marginal benefit of extra market access too certification is more widespread.) This opens the possibil- small. There is also the possibility that a regional stan- ity of the strong party’s imposing its vision of standards dards group may become large enough to exert market policy—formally, as part of the agreement, or perhaps power and affect the terms of trade, thus providing more subtly, through technical assistance, for example. incentives to exclude nonmembers. Finally and perhaps Harmonization to EU or international standards is, for closer to reality, large economic areas that are producing instance, a distinct feature of EU agreements. Moreover, standards have an incentive to export their own policy the fact that regional standards are promoted alongside or models. The EU and the United States, for example, are instead of international ones (again, the EU has promoted known to be pushing to spread their trade policy norms the use of its standards in several agreements) adds to the (Maur 2005; Horn, Mavroidis, and Sapir 2010), as is dis- risk that different hubs will pull partner countries in sepa- cussed next. rate directions incompatible with a multilateral vision.15 What the Texts of Agreements Tell Us Beyond the WTO Provisions Lesser (2007) and Budetta and Piermartini (2009) examine PTAs also predictably focus on areas in which multilateral the legal provisions relating to TBTs in PTAs notified to initiatives cannot or do not yet offer disciplines. Two main the WTO with a view to assessing, in particular, whether these models of recognition prevail. One, promoted in some U.S. provisions promote convergence toward the multilateral sys- PTAs, is the recognition of conformity assessment con- tem. Some dimensions reviewed by the two studies convey ducted in the territories of partner countries as equivalent a positive picture, in that PTAs do not overtly conflict with to one’s own. As noted by Lesser (2007), the U.S. agree- the WTO disciplines: ments often go beyond WTO rules by stating that any refusal to authorize certification performed in the part- • Of 70 surveyed PTAs, 58 have TBT provisions, and 30 ner country has to be justified on request. This contrasts (51 percent) make explicit reference to the WTO TBT with the softer language in the WTO TBT agreement, Agreement (Budetta and Piermartini 2009). which merely encourages equivalence. Another feature of • Lesser (2007) finds that 86 percent of the PTAs in a sam- the U.S. agreements is their promotion of private sector ple of 24 refer to the WTO TBT Agreement. self-certification through supplier declarations of con- formity assessment. The second model is that of the A mere reference to the WTO TBT Agreement would European Union, which insists more on MRAs on con- not be sufficient to justify a conclusion that PTAs consti- formity. Generally, this translates into separate bilateral tute a building block for a liberal multilateral system. The MRAs. To date, these agreements involve partners at sim- researchers therefore look at several other dimensions of ilar development levels; apparently PTAs between devel- PTAs that may contribute to greater or less convergence. oped and developing partners, even when calling for On the whole, a picture of agreements that, by and large, such MRAs, have not led to any being signed yet. act as potential complements to international initiatives A desire to go beyond the WTO agreements is one emerges. For instance, when PTAs seek harmonization of motivation of PTA provisions on TBTs. Transparency standards, technical regulations, and certification proce- provisions in PTAs are in some instances more onerous dures among partners (about half of PTAs reviewed do than those required by the WTO, with longer notification so), in about 60 to 70 percent of these cases, the use of times (90 days). In the case of harmonization to regional international standards is encouraged (Lesser 2007). standards and certification, the rationale is often to fill Product Standards 213 gaps where international standards do not exist or are inappropriate; this is the case with the Common Market Box 10.9. The Codex Alimentarius and Preferential Trade Agreements for Eastern and Southern Africa (COMESA), APEC, and the Andean Community (Lesser 2007). The North American Free Trade Agreement (NAFTA) The second motivation in PTAs is to promote imple- between Canada, Mexico, and the United States; the Treaty mentation and enforcement. Such agreements often create of Asunción, which established the Southern Cone Common enforcement and implementation institutions for stan- Market (Mercosur, Mercado Común del Sur) between Argentina, Brazil, Paraguay, and Uruguay; and the Asia- dards. Most PTAs call for the establishment of a committee, Pacific Economic Cooperation (APEC), with 21 members, body, or network in charge of TBTs (Lesser 2007). The have all adopted measures consistent with principles embraced by the Uruguay Round agreements and related to functions of such bodies can be varied: exchange of infor- Codex Alimentarius standards. mation, monitoring of implementation, consultation, and NAFTA includes two ancillary agreements dealing with in the more advanced cases, harmonization and legal sanitary and phytosanitary (SPS) measures and technical barriers to trade (TBT). In connection with SPS measures, enforcement. Examples of regional organizations for stan- Codex standards are cited as basic requirements to be met dards include the Caribbean Regional Organization for by the three member countries with respect to the health Standards and Quality (CROSQ, established by the and safety aspects of food products. Mercosur’s Food Commission has recommended a range of Codex Caribbean Community, CARICOM); the African Regional standards for adoption by member countries and is using Organization for Standardization (ARSO, set up by other Codex standards as points of reference in continuing deliberations. APEC has drafted a mutual recognition COMESA); and the Mercosur Standardization Association arrangement on conformity assessment of foods and food (AMN). An initiative worth noting is the Standardisation, products. It calls for consistency with SPS and TBT Quality Assurance and Metrology (SQAM) initiative by the requirements and with Codex standards, including the recommendations of the Codex Committee on Food Southern African Development Community (SADC); the Import and Export Certification Systems. EU directives also initiative encompasses cooperation on standards manage- frequently refer to the Codex Alimentarius as the basis for ment, accreditation, and metrology.16 their requirements. Dispute settlement is another dimension in which some Source: Web site of the Food and Agriculture Organization of the United Nations (FAO). PTAs create mechanisms supplementing or going beyond what is foreseen in the WTO. Numerous PTAs contain some form of dispute resolution arrangement (as does over half the sample reviewed by Lesser 2007). The Andean agreement is an example of one in which the revocation of economic analysis because of spillover effects or informa- infringing TBT measures can be ordered and in which dis- tion asymmetries. Since these effects are not limited by pute settlement measures provide for alternative mecha- national borders, they can serve as a basis for regional and nisms of redress to those offered by the WTO. global cooperation on product standards. Policy makers, We cannot really draw conclusions here as to whether however, need to be aware that product standards also these added disciplines have a measurable positive impact impose costs and that multiple or conflicting standards can on developing countries' ability to meet standards and create an overly burdensome cost environment for business their standard systems. More detailed sector-level studies and international trade. Like any regulatory instruments, would be required to determine whether these disciplines standards are open to capture by vested interests and may in promote more effective and less costly use of standards. some cases act as a form of protectionist measure. We have so far presented evidence about TBTs only. We There is a clear tension between, on the one hand, the can presume that similar regional provisions relating to legitimate protection of important social goals, and the SPS standards and measures are strongly linked to interna- promotion of economic efficiency in some areas, and, on tional harmonization efforts. Casual evidence is provided the other, the costs that standards can entail both inside by references in PTAs to the Codex Alimentarius (see box and outside national borders. These costs can be particu- 10.9), which is mentioned in 20 agreements in the World larly severe for firms in countries that are excluded from Trade Law database. common approaches to standards, as well as in developing countries where technical and financial constraints can make it difficult to comply with burdensome standards Conclusions abroad. The economic costs and benefits of product stan- Product standards are an important fact of commercial life dards need to be carefully assessed in each case, paying par- and in many instances are justified—at least in part—by ticular attention to the possibility of cross-border impacts 214 Jean-Christophe Maur and Ben Shepherd and to whether individual countries are in a position to roles to play in supporting the development of standards build a satisfactory national standards infrastructure or infrastructure within PTAs. whether cross-border cooperation is needed. The next chapter reviews in greater depth actual prac- PTAs can provide an answer to regional cooperation tice with respect to product standards in a selected number problems. Moreover, they are at the heart of liberalization of representative PTAs. efforts regarding standards and technical regulations. This is partly because core measures such as harmonization and Notes mutual recognition of standards or of conformity assess- ment are much easier to negotiate and implement among The authors are grateful to Juliana Salles Almeida, Silja Baller, Michael Friis Jensen, and Roberta Piermartini for suggestions and sharing of groups of countries with relatively similar development results. levels and institutional settings. As with other types of pref- 1. For an overview of policy issues in this area and a review of erential liberalization, however, it is important for policy empirical work, see WTO (2005). makers to ensure that such efforts work in tandem with 2. This is a variant of the “lemons� problem discussed by Akerlof (1970). and do not undermine the broader multilateral agenda 3. In some cases the declaration of conformity can be made by the expressed in the WTO SPS and TBT Agreements. purchaser, as well. A supplier’s declaration regime can potentially lead to Indeed, most PTAs do not seem to be designed with significant cost savings for business and government. Fleiss, Gonzales, and Schonfeld (2008) find some empirical evidence in Europe to support the the objective of taking a different path from the one view that a shift from third-party certification to a supplier-declaration agreed in the WTO. WTO disciplines, however, remain system can be trade promoting. relatively general, so PTAs should ensure that they adopt 4. The WTO agreements maintain a distinction between standards broadly related to food safety and those related to more general issues in disciplines that reduce the risks of exclusion of third- the area of manufactured goods. The former are largely dealt with under country traders. This chapter has reviewed a number of the Agreement on the Application of Sanitary and Phytosanitary Meas- possible approaches, such as a focus on international har- ures (the SPS Agreement); the latter come under the Agreement on Tech- monization and inclusion of open rules of origin regard- nical Barriers to Trade (the TBT Agreement). 5. The empirical literature offers little evidence that protectionism ing certification. Both instruments can be seen as ways of may be a motivation behind some standards (see Kono 2006). The finan- multilateralizing PTAs. Of course, it is also important for cial crisis and measures taken by some countries to protect domestic PTA groupings to retain sufficient flexibility to allow industries have, however, raised again the specter that standards may be used as protectionist devices. An example is India’s recent decision to ban standards to evolve in line with international develop- Chinese toys on safety grounds (“Downturn Heightens China-India Ten- ments and not to lock participants into a particular set of sion on Trade.� Wall Street Journal, March 20, 2009, http://online.wsj norms that is difficult to modify. .com/article/SB123749113639187441.html). 6. WTO Article VIII states, “All fees and charges of whatever charac- For parties to a PTA, harmonization and mutual recog- ter (other than import and export duties and other than taxes within the nition seem to have positive effects on the volume of trade. purview of Article III) imposed by contracting parties on or in connection Choosing between harmonization and mutual recognition with importation or exportation shall be limited in amount to the is, however, important because the countries’ level of devel- approximate cost of services rendered and shall not represent an indirect protection to domestic products or a taxation of imports or exports for opment and their regulatory objectives will influence which fiscal purposes.� approach is better suited to their particular circumstances. 7. This happens when agreements are signed with the EU, which In general, harmonization seems suitable for a minority of requires strict standards. Countries have to meet the safety expecta- tions set by EU legislation if they want to be able to export foods to the cases. Whether to pursue commonality in standards defini- EU. In some African, Caribbean, and Pacific (ACP) countries, this tion or simply in conformity assessment is another consid- requirement has been met with resistance by policy makers, who see a eration. Cooperation on conformity assessment is arguably traditional sphere of influence being undermined by the new regime (Doherty 2008). a more accessible first step for many countries. 8. As noted above, harmonization is not always optimal in a welfare Product standards raise a number of important issues in sense because countries (and regions) can differ in their preferences and the context of North-South PTAs, which are becoming resource endowments. Thus, while adoption of internationally harmo- steadily more common. Adoption of identical standards by nized standards can often be beneficial from a trade point of view, policy makers need to ensure that the overall welfare effect is positive. This ques- countries at very different development levels raises serious tion is particularly important for developing countries and regions, given questions as to whether either country, and in particular the technical and financial burdens that international standards can the less developed partner, will achieve a socially optimal impose. 9. Harmonization and mutual recognition should generally be level of regulation. In addition, developing countries can viewed as complements, not substitutes. In the EU’s New Approach, for face particularly severe technical and financial obstacles instance, both instruments work together. when it comes to undertaking the investments necessary to 10. TBT Article 6.3 states, “Members are encouraged, at the request of other Members, to be willing to enter into negotiations for the conclu- bring about compliance with some developed-country sion of agreements for the mutual recognition of results of each other’s standards. This analysis suggests that technical assistance, conformity assessment procedures. Members may require that such capacity building, and aid for trade might have important agreements fulfill the criteria of paragraph 1 and give mutual satisfaction Product Standards 215 regarding their potential for facilitating trade in the products Bhagwati, Jagdish. 1996. “The Demands to Reduce Domestic Diversity concerned.� among Trading Nations.� In Fair Trade and Harmonization: Prerequi- 11. SPS Article 4 states, “Members shall, upon request, enter into con- sites for Free Trade?, ed. Jagdish Bhagwati and Robert E. Hudec. sultations with the aim of achieving bilateral and multilateral agreements Cambridge, MA: MIT Press. on recognition of the equivalence of specified sanitary or phytosanitary Budetta, Michele, and Roberta Piermartini. 2009. “A Mapping of Regional measures.� Rules on Technical Barriers to Trade.� In Regional Rules in the Global 12. It is worthwhile at this stage to mention another area in which the Trading System, ed. Antoni Estevadeordal, Kati Suominen, and Robert question of (regional) standards applies: that of services. Authorization, cer- Teh. Cambridge, U.K.: Cambridge University Press. tification, licensing of services suppliers, and recognition of qualifications Casella, Alessandra. 2001. “Product Standards and International Trade: are areas in which the question of standards arises and regional integration Harmonization through Private Coalitions?� Kyklos 54 (2/3): prospects may have a role to play. The WTO General Agreement on Trade in 243–64. Services resembles the TBT and SPS Agreements in recognizing this, Chaney, Thomas. 2008. “Distorted Gravity: The Intensive and Extensive although in a quite different way. Trachtman (2003) notes that GATS Article Margins of International Trade.� American Economic Review 98 (4): VII provides specifically for the autonomous or mutual recognition of stan- 1707–21. dards of other trade partners. It thus clearly views mutual recognition as an Chen, Maggie X., and Aaditya Mattoo. 2008. “Regionalism in Standards: acceptable exception to MFN treatment. Good or Bad for Trade?� Canadian Journal of Economics 41 (3): 13. Note, however, that the existence of a uniform standard does not 838–63. guarantee nondiscrimination: the standard may be so designed as to tar- Czubala, Witold, Ben Shepherd, and John S. Wilson. 2009. “Help or Hin- get a particular source of imports (by including a characteristic specific to drance? The Impact of Harmonised Standards on African Exports.� that source only) or may be implemented in a discriminatory fashion Journal of African Economies 18 (5): 711–44. through testing and certification procedures. David, Paul A., and Shane Greenstein. 1990. “The Economics of Compat- 14. The EU–U.S. agreement covers telecommunication equipment, ibility Standards: An Introduction to Recent Research.� Economics of electromagnetic compatibility (EMC), electrical safety, recreational Innovations and New Technologies 1: 3–41. craft, manufacturing practices for pharmaceutical goods, and medical Diaz Rios, Luz B., and Steven Jaffee. 2008. “Barrier, Catalyst, or Distrac- devices. MRAs have also been concluded with Australia, Canada, Japan, tion? Standards, Competitiveness, and Africa’s Groundnut Exports to New Zealand, and Switzerland; see http://ec.europa.eu/enterprise/ Europe.� Agriculture and Rural Development Discussion Paper 39, international/index_en.htm. World Bank, Washington, DC. 15. See table 10.2, above. The diffusion of EU standards is actually Disdier, Anne-Célia, Lionel Fontagné, and Mondher Mimouni. 2008. clearly advocated in the European Commission (2001), which states as an “The Impact of Regulations on Agricultural Trade: Evidence from the aim, “to promote where possible, the adoption of overseas standards, and SPS and TBT Agreements.� American Journal of Agricultural Economics regulatory approaches based on, or compatible with, international and 90 (2): 336–50. European practices, in order to improve the market access and competi- Doherty, Martin. 2008. “The Importance of Sanitary and Phytosanitary tiveness of European products.� Measures to Fisheries Negotiations in Economic Partnership Agree- 16. A recent review of standard provisions in African PTAs can be ments.� ICTSD Issues Paper 7, International Centre for Trade and Sus- found in Meyer et al. (2010). tainable Development, Geneva. Drezner, Daniel W. 2006. “The Race to the Bottom Hypothesis: An Empirical and Theoretical Review.� Fletcher School, Tufts Univer- sity, Medford, MA. http://www.danieldrezner.com/policy/RTBreview References .doc. Akerlof, George A. 1970. “The Market for Lemons: Quality Uncertainty European Commission. 2000. Guide to the Implementation of Directives and the Market Mechanism.� Quarterly Journal of Economics 84 (3): Based on the New Approach and the Global Approach. Luxembourg: 488–500. Office for Official Publications of the European Communities. http:// Aldaz-Carroll, Enrique. 2006. “Regional Approaches to Better Standards ec.europa.eu/enterprise/newapproach/legislation/guide/index.htm. Systems.� Policy Research Working Paper 3948, World Bank, Washington, ———. 2001. “Implementing Policy for External Trade in the Fields of DC. Standards and Conformity Assessment: A Tool Box of Instruments.� Amurgo-Pacheco, Alberto. 2006. “Mutual Recognition Agreements and Commission Staff Working Paper SEC(2001)1570, European Com- Trade Diversion: Consequences for Developing Nations.� Working mission, Brussels. Paper 20/2006, Graduate Institute of International and Development Fleiss, Barbara, Frederic Gonzales, and Raymond Schonfeld. 2008. “Tech- Studies, Geneva. nical Barriers to Trade: Evaluating the Effects of Supplier’s Declara- An, Galina, and Keith E. Maskus. 2009. “The Impacts of Alignment with tion of Conformity.� Trade Policy Working Paper 78, Organisation for Global Product Standards on Exports of Firms in Developing Coun- Economic Co-operation and Development, Paris. tries.� World Economy 32 (4): 552–74. Gandal, Neil, and Oz Shy. 2000. “Standardization Policy and International Baldwin, Richard E. 2000. “Regulatory Protectionism, Developing Trade.� Journal of International Economics 53 (2): 363–83. Nations, and a Two-Tier World Trading System.� In Brookings Trade Ganslandt, Mattias, and James R. Markusen. 2001. “Standards and Related Forum, 2000, ed. Susan Collins and Dani Rodrik. Washington, DC: Regulations in International Trade: A Modeling Approach.� NBER Brookings Institution. Working Paper 8346, National Bureau of Economic Research, Baldwin, Richard E., Simon Evenett, and Patrick Low. 2009. “Beyond Tar- Cambridge, MA. iffs: Multilateralizing Non-Tariff RTA Commitments.� In Multilateral- Hahn, Robert W., and Robert E. Litan. 2005. “Counting Regulatory Bene- izing Regionalism: Challenges for the Global Trading System, ed. fits and Costs: Lessons for the US and Europe.� Journal of International Richard E. Baldwin and Patrick Low. Cambridge, U.K.: Cambridge Economic Law 8 (2): 473–508. University Press. Helble, Matthias, Ben Shepherd, and John S. Wilson. 2007. Transparency Baller, Silja. 2007. “Trade Effects of Regional Standards Liberalization: A and Trade Facilitation in the Asia-Pacific: Estimating the Gains from Heterogeneous Firms Approach.� Policy Research Working Paper Reform. Canberra: Department of Foreign Affairs and Trade. 4124, World Bank, Washington, DC. Henry de Frahan, Bruno, and Mark Vancauteren. 2006. “Harmonization Barrett, Christopher B., and Yi-Nung Yang. 2001. “Rational Incompatibil- of Food Regulations and Trade in the Single Market: Evidence from ity with International Product Standards.� Journal of International Disaggregated Data.� European Review of Agricultural Economics Economics 54 (1): 171–91. 33 (3): 337–60. 216 Jean-Christophe Maur and Ben Shepherd Henson, Spencer, and Steve Jaffee. 2007. “Public and Private Food Safety Nicolaïdis, Kalypso. 2001. “Harmonisation and Recognition: What Have Standards: Compliance Costs and Strategic Opportunities.� World We Learned?� In Trade and Regulatory Reform: Insights from Country Bank Institute, Washington, DC. Experience. Paris: Organisation for Economic Co-operation and Horn, Henrik, Petros C. Mavroidis, and André Sapir. 2010. “Beyond the Development. WTO? An Anatomy of EU and US Preferential Trade Agreements.� Nicolaïdis, Kalypso, and Michelle Egan. 2001. “Transnational Market World Economy 33 (11): 1565–88. Governance and Regional Policy Externality: Why Recognize Foreign ISO (International Organization for Standardization). 2006. Metrology, Standards?� Journal of European Public Policy 8 (3): 454–73. Standardization, and Conformity Assessment: Building an Infrastruc- Nicolaïdis, Kalypso, and Gregory Shaffer. 2005. “Managed Mutual Recog- ture for Sustainable Development. Geneva: ISO. nition Regimes: Governance without Global Government.� Law and ———. 2008. Fast Forward: National Standards Bodies in Developing Contemporary Problems 68: 263–318. Countries. Geneva: ISO. Nofal, Beatriz. 2004. “Constructing a Deeper Integration in MERCOSUR.� Jaffee, Steven. 2003. “From Challenge to Opportunity: Transforming Background paper for Global Economic Prospects 2005, World Bank, Kenya’s Fresh Vegetable Trade in the Context of Emerging Food Safety Washington, DC. and Other Standards in Europe.� Agriculture and Rural Development Otsuki, Tsunehiro, John S. Wilson, and Mirvat Sewadeh. 2001. “What Discussion Paper 1, World Bank, Washington, DC. Price Precaution? European Harmonization of Aflatoxin Regulations Jones, Philip, and John Hudson. 1996. “Standardization and the Costs of and African Food Exports.� European Review of Agricultural Economics Assessing Quality.� European Journal of Political Economy 12 (2): 355–61. 28 (3): 263–84. Kono, Daniel. 2006. “Optimal Obfuscation: Democracy and Trade Pol- Sanetra, Clemens, and Rocío M. Marbán. 2007. “The Answer to the Global icy Transparency.� American Political Science Review 100 (3): 369–84. Quality Challenge: A National Quality Infrastructure.� Organization of Lesser, Caroline. 2007. “Do Bilateral and Regional Approaches for Reduc- American States, Washington, DC; International Trade Centre, Geneva; ing Technical Barriers to Trade Converge towards the Multilateral Physikalisch-Technische Bundesanstalt, Braunschweig. http://www.ptb Trade System?� OECD Trade Working Paper 58, TAD/TC/WP(2007) .de/en/org/q/q5/docs/OAS_EN07.pdf. 12/FINAL, Organisation for Economic Co-operation and Develop- Shepherd, Ben. 2006. “The EU Standards Database: Overview and User ment, Paris. Guide.� World Bank, Washington, DC. http://siteresources.worldbank Maertens, Miet, and Johan F. Swinnen. 2009. “Trade Standards and .org/INTRES/Resources/469232-1107449512766/EUSDB_Overview_ Poverty: Evidence from Senegal.� World Development 37 (1): 161–78. UserGuide_021306.pdf. Maskus, Keith, Tsunehiro Otsuki, and John S. Wilson. 2000. “Quantifying ———. 2007. “Product Standards, Harmonization, and Trade: Evidence the Impact of Technical Barriers to Trade: A Framework for Analysis.� from the Extensive Margin.� Policy Research Working Paper 4390, Policy Research Working Paper 2512, World Bank, Washington, DC. World Bank, Washington, DC. http://go.worldbank.org/FS4JCCSV10. Swann, Peter, Paul Temple, and Mark Shurmer. 1996. “Standards and ———. 2005. “The Cost of Compliance with Product Standards for Firms Trade Performance: The UK Experience.� Economic Journal 106 (438): in Developing Countries: An Econometric Study.� Policy Research 1297–1313. Working Paper 3590, World Bank, Washington, DC. http://www-wds Trachtman, Joel P. 2003. “Towards Open Recognition? Standardization .worldbank.org/servlet/WDSContentServer/WDSP/IB/2005/05/15/00 and Regional Integration under Article XXIV of GATT.� Journal of 0090341_20050515133900/Rendered/PDF/wps3590.pdf. International Economic Law 6 (2): 459–92. http://jiel.oxfordjournals Maur, Jean-Christophe. 2005. “Exporting Europe’s Trade Policy.� World .org/cgi/reprint/6/2/459. Economy 28 (11): 1565–90. Veggeland, Frode, and Christel Elvestad. 2004. “Equivalence and Mutual Meyer, Nico, Tamas Fenyes, Martin Breitenbach, and Ernst Idsardi. 2010. Recognition in Trade Arrangements: Relevance for the WTO and the “Bilateral and Regional Trade Agreements and Technical Barriers to Codex Alimentarius Commission.� NILF-Report 2004-9, Norwegian Trade: An African Perspective.� OECD Trade Policy Working Papers 96, Agricultural Economics Research Institute, Oslo. Organisation for International Co-operation and Development, Paris. World Bank. 2008. Overcoming Trade Barriers from Standards and Tech- Moenius, Johannes. 2005. “Information versus Product Adaptation: The nical Regulations. Policy Research Report, East Asia and Pacific Role of Standards in Trade.� University of Redlands, Redlands, CA. Region Poverty and Economic Management. Washington, DC: http://ssrn.com/abstract=608022. World Bank. ———. 2007. “Do National Standards Hinder or Promote Trade in Elec- WTO (World Trade Organization). 2005. World Trade Report 2005: Trade, trical Products?� University of Redlands, Redlands, CA. http://www Standards and the WTO. Pt. 2. Geneva: WTO. http://www.wto.org/ .iecchallenge.org/papers/pdf_iecchallenge/moenius.pdf. english/res_e/publications_e/wtr05_e.htm. 11 TBT AND SPS MEASURES, IN PRACTICE Andrew L. Stoler In recent years, the number of technical regulations and assessment procedures. There is, however, a risk that they standards adopted by countries has grown significantly. can introduce de facto discrimination in global markets, More stringent regulatory policy can be seen as impelled by particularly against developing countries, because achiev- higher standards of living worldwide, which have boosted ing conformity in technical standards requires capacity and consumers’ demand for safe, high-quality goods, and by resources. growing problems of water, air, and soil pollution that have This chapter looks at the experience of representative encouraged modern societies to explore environmentally PTAs with TBT and SPS provisions, with a view to identify- friendly products. ing common characteristics of, and differences between, Measures related to technical barriers to trade (TBTs) their basic approaches to standards. and to sanitary and phytosanitary (SPS) standards and reg- ulation have become important dimensions of preferential Standards and International Trade trade agreements (PTAs) (see Maur and Shepherd, ch. 10 in this volume). Governments seek to act through their PTAs, The aims of SPS regulations and standards are to protect as well as through the World Trade Organization (WTO), to human beings or animals from risks arising from additives, protect human, animal, or plant life or health. Such efforts contaminants, toxins, and disease-causing organisms in are within WTO guidelines, provided that they are not dis- their food; to protect human life from plant- or animal- criminatory and that regulations and standards are not used carried diseases; to protect animal or plant life from pests, as disguised protectionism. In a number of instances, PTA diseases, or disease-causing organisms; and to prevent or members seek to go beyond the broad rules-based approach limit other damage to a country from the entry, establish- followed in the WTO and to reduce differences in national ment, or spread of pests.1 standards and certification processes that impede trade. TBT technical regulations and standards set out spe- There are, broadly, two models for dealing with stan- cific characteristics of a product, such as its size, shape, dards measures in PTAs. Where the European Union (EU) design, functions, and performance, or the way it is labeled is a party to a PTA, the agreement often calls for the partner or packaged before it is put on sale. In certain cases, how a country to harmonize its national standards and conform- product is made can affect these characteristics, and it ity assessment procedures with those of the EU. PTAs in the may then prove more appropriate to draft technical regu- Asia-Pacific region and those in which the United States is a lations and standards on the basis of process and produc- partner typically seek to address problems resulting from tion methods rather than of the product’s characteristics different national standards and conformity procedures per se. through a preference for international standards or through In all the PTAs reviewed in this chapter, members use the use of mutual recognition mechanisms. the PTA to go beyond what is achievable through the Both approaches can be successful in reducing the nega- multilateral instruments of the WTO (box 11.1). They tive impact of a multiplicity of standards and conformity appear to recognize that only by avoiding a situation in 217 218 Andrew L. Stoler Box 11.1. WTO Standards and Guidelines on TBT and SPS Measures Article XX of the General Agreement on Tariffs and Trade (GATT) allows governments to enact trade measures to protect human, animal, or plant life or health, provided that the provisions do not discriminate and are not used as disguised protectionism. In addition, two specific World Trade Organization agreements deal with food safety, animal and plant health and safety, and product standards in general. Both seek to identify how to meet the need for standards and at the same time avoid protectionism in disguise. These issues are becoming more important as tariff barriers fall. The WTO Agreement on the Application of Sanitary and Phytosanitary Measures (the SPS Agreement) lays out the basic rules on food safety and on animal and plant health standards. It allows countries to set their own standards, but it stipulates that regulations must be based on science and should be applied only to the extent necessary to protect human, animal, or plant life or health. Furthermore, such regulations should not arbitrarily or unjustifiably discriminate between countries where identical or similar conditions prevail. Member countries are encouraged to use international standards, guidelines, and recommendations where these exist. When this practice is followed, the measures are unlikely to be challenged legally in a WTO dispute. Members, however, may impose measures that result in higher standards if there is scientific justification; they may set higher standards on the basis of appropriate assessment of risks, so long as the approach is consistent and not arbitrary; and they can, to some extent, apply the “precautionary principle�—a kind of safety-first approach—to deal with scientific uncertainty. The agreement allows countries to use different standards and different methods of inspecting products. If an exporting country can demonstrate that the measures it applies to its exports achieve the same level of health protection as in the importing country, the importing country is expected to accept the exporting country’s standards and methods. The SPS Agreement includes provisions on control, inspection, and approval procedures. Governments must provide advance notice of new or changed sanitary and phytosanitary regulations and establish a national enquiry point to provide information. The agreement complements the WTO Agreement on Technical Barriers to Trade (the TBT Agreement). The TBT Agreement seeks to ensure that technical regulations, standards, and testing and certification procedures do not create unnecessary obstacles. The agreement does recognize countries’ rights to adopt the standards they consider appropriate—for example, to protect human, animal, or plant life or health; to safeguard the environment; or to meet other consumer interests. Furthermore, members are not prevented from taking measures necessary to ensure that their standards are met. But the agreement also lays down disciplines. A myriad of regulations can be a nightmare for manufacturers and exporters. Life can be simpler if governments apply international standards, and the agreement encourages them to do so. In any case, whatever regulations countries use should not discriminate. The agreement also sets out a code of good practice for both governments and nongovernmental or industry bodies in preparing, adopting, and applying voluntary standards, and more than 200 standards- setting bodies use this code. Under the agreement, the procedures used to decide whether a product conforms with relevant standards have to be fair and equitable, and any methods that would give domestically produced goods an unfair advantage are discouraged. The agreement also encourages countries to recognize each other’s procedures for assessing whether a product conforms. Without recognition, products might have to be tested twice, first by the exporting country and then by the importing country. Manufacturers and exporters need to know what the latest standards are in their prospective markets, and, to ensure that this information is made available conveniently, all WTO member governments are required to establish national enquiry points and to keep each other informed through the WTO. About 900 new or changed regulations are notified each year. The Technical Barriers to Trade Committee is the main clearinghouse for sharing information among members and the primary forum for discussing concerns about the regulations and their implementation. Source: World Trade Organization documents, http://www.wto.org. which producers have to manufacture to different stan- agreements among developing countries; these include the dards in different national markets, or have to test the Association of Southeast Asian Nations (ASEAN) and the same product repeatedly, will the parties be able to foster Southern Cone Common Market (Mercosur, or Mercado more deep-seated economic integration. Común del Sur). PTAs that include TBT and SPS provisions normally There are different models for dealing with the elimina- incorporate an active work program of cooperation on tion of the employment of TBT and SPS measures as trade standards, certification, and conformity assessment issues. barriers in PTAs. In PTAs involving the EU, there is a strong This produces a stronger economic development focus, in preference for harmonization of standards and for con- contrast to the way these issues are treated in WTO agree- formity assessment procedures. As a trade-off, the EU typi- ments. The latter are designed to set standards and guide- cally supports, both technically and financially, significant lines, whereas the PTAs examined here go beyond that technical assistance programs to assist developing-country objective to contribute to economic integration through partners with the harmonization effort. When the EU con- the phased elimination of standards-related barriers. Some cludes an agreement with more distant countries, such as of the most ambitious PTAs in this regard are South-South Chile, there is not normally an obligation to harmonize TBT and SPS Measures, in Practice 219 completely with EU standards and procedures; instead, the this section briefly reviews the Australia–United States agreement calls for the promotion and use of both EU and Free Trade Agreement (AUSFTA) as illustrative of North- international standards. North PTAs. Agreements concluded among Asian countries and After relatively brief negotiations, the AUSFTA entered those involving the participation of the United States take a into force in 2005. The agreement was not the outcome of different approach from that of the EU. Rather than an attempt to solve any significant trade issues between the require harmonization, these PTAs typically seek to facili- two countries; rather, it was seen by both Canberra and tate mutual recognition agreements and approaches based Washington as a demonstration of how two like-minded on equivalence of different approaches in different coun- developed countries could move ahead with significant tries. The U.S. PTAs normally include the establishment of trade liberalization in a period when multilateral progress a committee charged with addressing TBT measures or SPS in the Doha Round was stalled. That said, both U.S. and rules that are seen to be creating trade problems. Australian exporters did face some difficulties with SPS In chapter 10 in this volume, Maur and Shepherd measures in their bilateral export markets, and it was natu- observe that it is unclear which advantages each model ral to build into the agreement provisions addressed to may or may not yield and that any judgment as to which both SPS and TBTs. model is superior may well depend on the particular part- Both Australia and the United States accepted as the ner country and the similarity of regulatory preferences. basis for their obligations the provisions of the WTO SPS In the sections that follow, the main features of TBT and and TBT Agreements. The chapters of the agreement build SPS provisions in 11 representative PTAs are reviewed, on these basic obligations by establishing bilateral institu- taking first North-North PTAs and then North-South and tions and procedures designed to further facilitate trade South-South agreements. Experience with the cost of between the parties. implementing TBT and SPS measures and the role of tech- Chapter 7 of the AUSFTA establishes a bilateral com- nical assistance are then discussed. Table 11.1 provides a mittee with a mandate of facilitating bilateral information summary of the main features of the 11 agreements with exchange and bilateral consultations on SPS measures that respect to TBT and SPS measures. could affect bilateral trade; on technical cooperation activ- ities and issues; and on positions and agendas in multilat- eral SPS forums, including the WTO’s SPS Committee. The North-North PTAs: The Example of AUSFTA committee is supplemented by a Standing Technical Work- As a benchmark for a discussion of North-South and ing Group on Animal and Plant Health Measures. The South-South PTAs in the area of TBT and SPS measures, mandate of the working group includes resolving specific Table 11.1. Comparison of Main Features Relating to TBT and SPS Measures, 11 PTAs Use of “Living Provisions Provisions Technical Binding with international agreement� on TBT on SPS WTO+ assistance dispute PTA standards institutions issues issues transparency provisions settlement ASEAN Trade in Goods Agreement (ATIGA) Yes Yes Yes Yes Yes Unclear Yes Australia–United States FTA (AUSFTA) Yes Yes Yes Yes Yes No Not for SPS Chile–China Unclear Yes Yes Yes Unclear No Yes Chile–United States Yes Yes Yes Yes Yes No Not for SPS China–New Zealand Yes Yes Yes Yes Yes Yes Yes EU–Chile Yes Yes Yes Yes No Yes No EU–Morocco Yes Yes Yes Yes No Yes No EU–South Africa TDCA Yes Yes Yes Yes No Yes No Mercosur Unclear Yes Yes Yes Unclear No Yes Singapore–Australia FTA (SAFTA) Yes Yes Yes Yes Yes No Yes Thailand–Australia FTA (TAFTA) Yes Yes Yes Yes No Yes Not for SPS Source: Author’s compilation. Note: ASEAN, Association of Southeast Asian Nations; FTA, free trade agreement; PTA, preferential trade agreement; SPS, sanitary and phytosanitary; TBT, technical barriers to trade; TDCA, Trade and Development Cooperation Agreement. 220 Andrew L. Stoler SPS issues, engaging in bilateral scientific and technical dominant. Meeting TBT and SPS regulations and stan- exchange on risk assessment and regulatory processes, and dards in those circumstances could involve significant costs considering measures relating to SPS that affect or are for producers and exporters located in the less developed likely to affect bilateral trade. The working group is also to country. These costs typically arise from the translation of establish specific work plans to address any SPS issues in foreign regulations, the hiring of technical experts to explain the bilateral relationship. foreign regulations, and the adjustment of production facil- The TBT chapter of the PTA differs from the SPS chap- ities to comply with the requirements. In addition, there is ter in that it creates TBT chapter coordinators that are the need to prove that the exported product meets the for- responsible for coordinating TBT matters with the other eign regulations. The high costs involved may discourage party and for consulting with the other side should a TBT manufacturers from trying to sell abroad. In the absence of issue arise in the bilateral relationship. Where an issue can- international disciplines, a risk exists that technical regula- not be resolved through the efforts of the coordinators, the tions and standards could be adopted and applied solely to PTA provides for the establishment of ad hoc technical protect domestic industries. working groups charged with identifying workable and The remainder of this section reviews the treatment of practical solutions that would facilitate trade. TBT and SPS provisions in seven PTAs representative of the The parties to the agreement obligate themselves to use variety of North-South agreements: EU–Morocco, EU–Chile, relevant international standards to the extent provided in EU–South Africa, Singapore–Australia, Thailand–Australia, the WTO Agreement and to give positive consideration to Chile–United States, and China–New Zealand. accepting as equivalent the technical regulations and con- formity assessment procedures of the other party. The EU–Morocco: A Euro-Mediterranean Agreement transparency provisions of the WTO Agreement are backed up by bilateral TBT notification obligations. The EU–Morocco PTA is representative of the approach According to Article 7.2.2 of the agreement, the entire to TBT and SPS provisions in the Euro-Mediterranean SPS chapter is off-limits with respect to dispute settlement agreements. The agreement action program is aimed at action. Thus, although the SPS provisions impose a kind of eliminating trade barriers associated with standards and “best-efforts� obligation on the parties, they are not bind- conformity assessment. It obliges the parties to the PTA ing over the longer term. A different approach is taken in to take appropriate steps to promote the use by Morocco the TBT chapter; only one article addressed to technical of EU technical rules and EU standards for industrial regulations (Article 8.5) is exempt from recourse to dis- and agricultural products and certification procedures. pute settlement. This implies that the remainder of the A system of accreditation of conformity assessment pro- chapter—in particular, its obligations regarding conform- cedures based on international and EU standards is also ity assessment and transparency—are legally binding on foreseen for Moroccan adoption. A separate mutual the parties and that this binding nature applies in both the recognition agreement (MRA), the Agreement on Con- short and long terms. formity Assessment and Acceptance of Industrial Prod- The provisions of the AUSFTA have been fully imple- ucts, is intended to implement this aspect of the PTA mented with respect to TBT and SPS measures. Implemen- (Lesser 2007). tation, however, has not been problem free. U.S. exporters In 2003, at the Palermo Euro-Med ministerial meeting, now enjoy relatively uninhibited (in terms of quarantine Euro-Med participants agreed to address the approxima- restrictions) access to the Australian market for stone tion of legislation in the field of standards, technical regu- fruit, table grapes, and citrus fruit, but this access has not lations, and conformity assessment procedures, through a put an end to U.S. complaints about the overly restrictive six-point program.3 That program calls for nature of Australian quarantine measures at the WTO SPS Committee.2 For its part, Australia still complains about • Identifying priority sectors the effect of certain American SPS measures on its exports. • Building acquaintance with applicable EU legislation and conducting a gap analysis • Transposing necessary framework legislation and sec- North-South PTAs toral legislation The impact on international trade of the need to comply • Creating or reforming institutions with different foreign technical regulations and standards • Setting up necessary certification and conformity may be more significant in the case of “asymmetrical� assessment bodies North-South agreements where one party is economically • Identifying technical assistance needs. TBT and SPS Measures, in Practice 221 Perhaps because of the gap between the two countries’ EU–Chile Free Trade Agreement relative levels of economic development, it has been diffi- The EU–Chile Free Trade Agreement (FTA) is representa- cult for Morocco to harmonize its standards with those of tive of a looser EU approach to TBT and SPS standards the EU or even to develop acceptable approaches to the and regulations. Here, the focus is on the use of technical development of mutual recognition agreements. A 2007 regulations and conformity assessment procedures based document of the Moroccan government comprehensively on international standards, unless those standards are lays out the steps that Morocco needs to take to realize the judged to be ineffective or inappropriate for fulfilling objectives of the 2005 EU–Morocco Plan of Action, which legitimate objectives. The parties also agree to work toward covers a period of three to five years.4 Among the actions compatibility or equivalence of their respective technical deemed necessary in the short term are the reinforcement regulations, standards, and conformity assessment proce- of local institutions charged with standardization, con- dures, and the agreement establishes a special committee formity assessment, metrology, and market surveillance. on these matters. The committee has identified the need The document notes that it will be particularly important for a technical assistance program and has launched such a for Morocco to reinforce its conformity assessment struc- program, funded by the EU. tures and guarantee their competence in areas covered by The special committee is cochaired by Chile and the EU. regulation. Although it is clear from both the 2007 govern- Its work program aims at ment paper and the action plan that harmonizing national standards with EU and international standards remains an • Monitoring and reviewing the implementation and important medium-term goal, there is little evidence that administration of the TBT chapter of the agreement Morocco has as yet been successful in these efforts. • Providing a forum for discussing and exchanging infor- Working with representatives of Moroccan industry and mation on any matter related to the TBT chapter, in the Commerce and Industry Ministry, the Moroccan Office particular as it relates to the parties’ systems for techni- of Standardisation and Quality Promotion has identified a cal regulations, standards, and conformity assessment number of priority sectors for the conclusion of conformity procedures, as well as developments in related interna- assessment recognition agreements with the EU. The choice tional organizations of priority sectors (electrical appliances, construction • Providing a forum for consultation and prompt resolu- machinery, and building products) stemmed partly from a tion of issues that act or can act as unnecessary barriers desire to ensure that certain products were safe to use in the to trade Moroccan market and partly from knowledge of where • Encouraging, promoting, and otherwise facilitating national testing capacity is already in place in the country.5 cooperation between the parties’ organizations, whether The EU has worked to support Morocco through tech- public or private, for metrology, standardization, test- nical assistance. A document of the Directorate General for ing, certification, inspection, and accreditation Trade lists 53 instances of SPS-related technical assistance • Exploring any means for improving access to the par- provided to Morocco during the period 2001–05.6 But ties’ respective markets and enhancing the functioning although this assistance has undoubtedly been helpful, of the TBT chapter. Morocco faces other important problems in selling its products on the EU market. The committee has met regularly. According to a joint A major problem relates to the increasing need for pro- communiqué issued following the seventh meeting, held in ducers and exporters to meet private quality control stan- 2009, “the overall working of the committee was consid- dards in the EU, many of which differ greatly from one ered highly satisfactory by the parties.�7 member state market to another. For instance, a 2005 Information about the details of the agendas and the World Bank case study notes that producers and exporters results of the meetings of the committee is not readily of citrus and tomatoes have to meet six different quality available. A specific working group on animal welfare was control standards: of these, two are recognized worldwide, established in 2003 by the Joint Management Committee three are European retailers’ individual standards, and one for SPS, with the aim of achieving the objectives of SPS relates to organic and biodynamic standards (Aloui and provisions in the EU–Chile PTA. According to the infor- Kenny 2005). Just how much the machinery of the PTA can mation available, this working group has developed its help Moroccan exporters in areas like this is unclear. Fur- activities on the basis of an annual action plan that has thermore, compliance has proved not only technically dif- been agreed jointly. At the beginning, it was focused on ficult, but also costly (see “Implementation Costs and practices relating to the stunning and slaughter of ani- Technical Assistance Needs,� below). mals. In 2006 both parties decided to incorporate animal 222 Andrew L. Stoler transport as a new issue to work on in the context of the pursued in harnessing the potential created by the Agree- agreement (Benavides and Jerez 2007). The European Com- ment for expanding EU interests into the surrounding mission approved a financial contribution to a maximum countries. Full use should be made of the action agreed amount of 35,000 euros for the organization of an interna- and cooperation tools envisaged, particularly in the fol- lowing areas: . . . .11 tional seminar on animal welfare in the framework of the PTA. A series of seminars grew out of conclusions reached It appears that Chile has, to some extent, been caught in in the working group on how better to reach the objectives a bind resulting from differing obligations in the PTAs it has of the agreement through exchange of information on sci- concluded with the EU and the United States. The passage entific expertise and the establishment of active contacts continues, concerning standards and technical regulations: between scientists from both parties.8 The EU observed with concern a marked tendency for the An operational fund was established to facilitate the Chilean standardisation process to incorporate solely a ref- implementation of the EU–Chile Association Agreement. erence to the US standards, particularly when no agreed Five of 13 projects covered by these funds involved TBT international standards exist. The immediate effect of such and SPS issues. These were related to behaviour is to divert trade to imports of non-EU origin or to give rise to additional costs to adapt products made in • International traceability and comparability of chemical the EU. The EU will focus on increased cooperation and measurements and supervision of market risk for the pay political attention to the promotion of international main exported food products standards, or in their absence, to double recognition of • Creation of a Web site on Chilean and European techni- both US and EU norms. Such an approach should be fol- cal regulation, including the EU’s environmental and lowed, in particular, for new technologies where the EU quality requirements local value added is still prominent. • Requirements of standards and conformity assessment On SPS, the annex adds: in the EU market • Harmonization of SPS regulations In the sanitary and phytosanitary area the objectives are to • A project to establish mutual recognition of testing implement the EC/Chile SPS Agreement fully and effec- tively. The EU’s potential to export food products and standards by Chile’s farming and stockbreeding service semi-processed agricultural products in order to complete (Servicio Agrícola y Ganadero) and by agricultural lab- the national product range on offer should be supported oratories in the EU. by smooth SPS administration. The benefits of these projects have been recognized by Chilean and EU authorities. For instance, Chile created a EU–South Africa Trade and Development Cooperation Web site, “Chile–UE: Requerimientos de Mercado, Más Agreement Información Mejores Negocios,� that was expected to facil- The 1999 EU–South Africa Trade and Development Coop- itate trade through improved transparency relating to tech- eration Agreement (TDCA) is another illustration of the nical and environmental regulations in both Chile and the EU’s approach to TBT and SPS standardization and con- EU and through the reduction of the costs of access to the formity assessment, this time in a low-income country. information.9 Under the agreement, the parties pledge to cooperate on Another good example of the progress achieved in the context of this operational fund is related to the interna- • Measures, in accordance with the provisions of the WTO tional traceability and comparability of chemical meas- TBT Agreement, to promote greater use of international urements and supervision of market risk for the most technical regulations, standards, and conformity assess- important exported food products. Advances here have ment procedures, including sector-specific measures contributed to the international recognition of Chile’s • Development of agreements on mutual recognition of Center of Chemical Metrology and are expected to benefit conformity assessment in sectors of mutual economic Chilean export sectors.10 interest Finally, it is important to note an observation in Annex 2, • Cooperation in the area of quality management and on policy coherence analysis, of the EU–Chile country assurance in selected sectors of importance to South strategy paper for 2007–13: Africa The EU/Chile Association Agreement represents the • Facilitation of technical assistance for Southern African most ambitious FTA that the EU has signed with a third capacity-building initiatives in the fields of accredita- party up to now. A similar level of ambition should be tion, metrology, and standardization TBT and SPS Measures, in Practice 223 • Development of practical links between South African From the start of the negotiations it was clear that all trade and European standardization, accreditation, and certi- and trade-related issues would be discussed in the context fication organizations.12 of the talks on the future Economic Partnership Agree- ment (EPA) with the countries of Southern Africa. The TDCA negotiations on trade and trade-related matters In comparison with other agreements signed by the were therefore immediately suspended, pending the out- EU—for instance, the EU–Chile Association Agreement— come of the EPA talks.16 the TDCA is less ambitious in the TBT and SPS areas. In fact, there is just one provision on TBTs that appears in Trade and trade-related issues seem to be handled more the section related to other trade-related matters, while effectively under the EPA between the EU and the South- SPS is referred to in the general provisions on agriculture ern African Development Community (SADC). According (Article 61) within the section on economic cooperation. to a trade cooperation report, “the revision of the trade There is no ready evidence on the extent to which chapter of the TDCA continued to be dealt with within the implementation of TBT and SPS provisions has been framework of the negotiations for an EU–SADC [EPA].�17 achieved. According to an official communication from the The EU’s interest seems to be to try to achieve meaning- European Commission, it became clear to the Commission ful coexistence between the TDCA and the EU–SADC EPA. that TBT and SPS provisions were not being implemented That, however, may take longer than expected because the fully when it recognized that bilateral cooperation on some regional context in southern Africa is particularly complex. topics, including SPS, was mostly limited and ad hoc. In In general terms, it is known that the EPA’s objectives are to addition, the Commission pointed out, develop a more predictable and rules-based regional mar- ket for goods and services with a view toward fostering As regards the provisions that have not been imple- regional integration; to harmonize trade rules within the mented yet, there seems to be a strong interest in deep- region; and to create a simple trading framework between ening cooperation in the following areas: trade and the countries of southern Africa and the EU. related-trade areas, intellectual property rights, customs, In the relationship between the SADC, the Southern competition policy, regional policy, sanitary and phy- African Customs Union (SACU), and the TDCA, a number tosanitary measures, technical barriers to trade, . . . (italics added)13 of issues appear to require urgent attention. At the Second South Africa–EU Summit, it was noted, This situation seems to explain the initial interest in The experience of the EU is witness to the benefit of strong including both TBT and SPS issues in the TDCA review.14 regional integration. We recognise that the Regional Eco- According to another European Commission communica- nomic Communities (RECs) are key pillars for deeper tion, the parties should consider a new orientation and integration in Africa and therefore support greater politi- possible revision of the TDCA.15 The Commission identi- cal cohesion and stronger economic integration in [SACU fied some provisions in the PTA that might need to be and the SADC]. amended. In effect, it suggested new commitments and In this context, we engaged in a frank and open discus- enhanced cooperation in trade-related areas and men- sion on the EU–SADC Economic Partnership Agreement tioned, among other articles, Article 47, which is related to and on the implications that these negotiations, at all its standardization and conformity assessment. stages, have on current processes of regional integration in The communication suggested, concerning a revision of Southern Africa. We agreed to urgently pursue the negotia- the economic cooperation text: “updated wording must be tion and resolution of all outstanding issues with a view to considered for cooperation in the fields of energy, trans- a prompt and mutually satisfactory conclusion that sup- ports regional integration and development in Southern port, agriculture and sanitary and phyto-sanitary matters� Africa.18 (italics added). On SPS, said the communication, a more substantial change was recommended because SPS meas- Cooperation on SPS measures is one of the areas explic- ures had become “an important impediment to trade and itly identified as a priority for treatment in the joint action therefore required enhanced cooperation.� plan of the EU–South Africa Strategic Partnership, estab- The negotiations on an agreement amending the lished May 14, 2007.19 This inclusion is important because TDCA, which started formally in 2007, did not in the end the joint action plan and the TDCA constitute the ground- include negotiations on trade and trade-related matters. In work for enhanced and deepened political dialogue and a European Commission communication concerning a cooperation. proposal for an EU–South African agreement amending The trade cooperation report mentioned above con- the TDCA, it was acknowledged that tains a section on SPS matters. Within it, both parties 224 Andrew L. Stoler express their commitment to conduct “open and trans- the most active negotiators of PTAs at the time.) The bilat- parent collaboration and communication� and “to avoid eral relationship underlying this PTA is significant both or minimise trade obstacles in the future.� This commit- economically and politically, and each party had some- ment followed an earlier South African ban on imports of thing to offer the other through additional liberalization. EU meat. During early 2009, close contacts between both In particular, through a process of annual reviews, the parties continued, with a view to ensuring SPS protection agreement has helped considerably to facilitate trade and doing so in “a way, which prevents any undue trade between Australia and Singapore, including trade affected restrictions between the EU and SA.� The report points by TBT and SPS measures, as exemplified by the Sectoral out that Annex on Horticultural Goods (box 11.2). The TBT and SPS chapter of SAFTA builds on an earlier the European Commission continued to extend SPS train- ing opportunities to the competent South African authori- bilateral mutual recognition agreement on conformity ties. For example, on 25–27 November 2008, a regional assessment. The chapter includes best-efforts obligations training workshop was held in Johannesburg on EU legis- on harmonization of mandatory requirements, acceptance lation for Bovine Spongiform Encephalopathy (BSE). The of the equivalence of each other’s mandatory require- South African authorities were also invited to participate ments, and cooperation on SPS questions that might arise in a training course in Austria on 2–5 March 2009 for vet- in the bilateral relationship. The core purposive policy in erinarians at border inspection posts at airports. An invi- this PTA is found in its provisions related to the negotia- tation has also been extended to a forthcoming training tion of sectoral annexes, which are, in effect, the imple- programme for laboratory experts in the area of aflatoxins/ menting arrangements for the chapter. For SAFTA, the sec- ochratoxins. Another meeting on food analysis will take toral annexes are intended to resolve specific issues in the place in Munich on 3–14 August 2009.20 bilateral relationship over time. The EU also funded a South Africa Pesticide Initiative The conformity assessment procedure has had effects in Program (SA PIP), which, although related to SPS matters, other market sectors. This includes the restriction on was developed in the context of the private sector develop- imports of cars that are more than three years old, a way of ment program. It was reported that the program has had ensuring that Singaporean road safety standards are met. It good results: is further illustrated by the removal of technical barriers to Australian electrical exports to Singapore. In the past, such The South Africa Pesticide Initiative Program (SA PIP) has goods had to be inspected and approved by two separate successfully hosted training workshops throughout the years, disseminating information to emerging farmers; Singaporean agencies before they could be approved for grower associations and various stakeholders. . . . The Pro- sale. Since the implementation of SAFTA, Singapore recog- gramme’s two main focus areas are to assist exporting pro- nizes Australian conformity assessment procedures.23 ducers to comply with the EU’s legal food safety require- ments; and to support and develop emerging farmers through training and empowerment initiatives, and so to grow the South African export volume destined for Box 11.2. Success Story: Orchids to Australia Europe.21 Within the context of the TDCA framework, South The Sectoral Annex on Horticultural Goods of the Singapore– Australia Free Trade Agreement establishes the concept of Africa and the EU jointly developed a country strategy “accredited exporter,� defined as “an exporter of the paper (CSP) and a European Commission multiannual scheduled horticultural goods who has demonstrated to its indicative program (MIP) for the period 2007–13. At least regulatory authority that it possesses the necessary technical capabilities, management competence, facilities, equipment 5 percent of this 980 million euro funding allocation will and production systems required to meet the mandatory go toward supporting a TDCA facility program. A develop- requirements of the importing Party.� In an early success story, Australia and Singapore agreed that Australia will ment cooperation report indicates that the EU will con- minimize import control and inspection and approval tribute 5 million euros to the project, to help with the procedures when orchids shipped by accredited Singaporean implementation of the TDCA and the joint action plan.22 exporters are accompanied by the required certificates and reports. Other horticultural goods are similarly subject to reduced import control, and goods can be added to the Singapore–Australia Free Trade Agreement coverage of the annex by mutual agreement of the SAFTA parties. The Singapore–Australia Free Trade Agreement (SAFTA), Sources: “Singapore-Australia Free Trade Agreement (SAFTA)—A signed in 2003, was Australia’s first PTA since the conclu- Business Guide� (http://www.dfat.gov.au/trade/negotiations/safta/ ch3_safta_guide.html), and SAFTA Annex 5-B, text (http://www sion of an agreement with neighboring New Zealand .dfat.gov.au/ta/safta). 20 years earlier. (Singapore, by contrast, was already one of TBT and SPS Measures, in Practice 225 Although SAFTA’s provisions are not very binding, as In most important areas, TAFTA obligations do not take the obligations are nearly all of a best-efforts nature, the the form of binding hard law. Generally, the obligation is PTA has produced some worthwhile short-term results. In for best efforts. In addition, the fact that the SPS chapter is the longer term, the agreement is likely to progressively specifically excluded from recourse to the PTA’s dispute take on a more binding character, as sectoral annexes are settlement mechanism also loosens the binding nature of negotiated and implemented to deal with specific issues the PTA. Mandated, more binding, obligations exist in and problems in bilateral trade. It is unlikely that the spe- areas in which compliance are easiest, such as information cific best-efforts obligations of SAFTA Chapter 05 could sharing and cooperation on procedural questions. be the basis for a successful action under the PTA’s own By virtue of its Article 610:2, TAFTA’s SPS provisions dispute settlement procedures. However, Article 10.1 of cannot give rise to invocation of the bilateral PTA’s dispute Chapter 05 makes it clear that implementing arrangements settlement provisions. Part of the mandate of the SPS for the chapter are intended to be reflected in sectoral Expert Group, however, is “progressing resolution of dis- annexes with more operational legal content. Failure to putes that arise in connection with the matters covered by abide by the terms of a negotiated sectoral annex would this Chapter.� The TBT chapter is silent on the question of appear to give cause for invocation of the PTA’s dispute set- dispute settlement, but its best-efforts obligations would be tlement provisions. By their very nature, obligations under difficult to litigate successfully under most dispute resolu- a sectoral annex could not be the cause for action under the tion systems. In all likelihood, serious disputes between WTO’s Dispute Settlement Understanding (DSU) because Australia and Thailand would need to be adjudicated they do not fit the definition of a “covered agreement.� under the WTO DSU. Cooperation under the PTA has been unable to resolve Thailand’s long-standing complaints about lack of access Thailand–Australia Free Trade Agreement to the Australian poultry market. Australia cites concerns The Thailand Australia Free Trade Agreement (TAFTA), over infectious bursal disease virus (IBDV) as necessitating which entered into force in 2005, resulted from a negotiation SPS measures that exclude the importation of fresh poul- that was much more problematic than the Singapore– try. An Australian risk assessment procedure in 2006 has Australia negotiations. Given Thai resistance to hard-law had the effect of making the Australian SPS measures more obligations in many areas of the agreement, Australia was severe. Market access has improved with the implementa- often willing to settle for looser language and weaker obliga- tion of TAFTA because of the utilization of TBT-related tions. In a number of areas in which it proved too difficult to mechanisms designed to promote world best practices in reach agreement, the two sides pledged to return to the nego- transparency, quarantine, and industrial standards.24 tiating table in 2008. Political issues have since interfered with that second stage of negotiations. Chile–United States Free Trade Agreement Against this background, it is not surprising that TAFTA incorporates a number of best-efforts obligations on the Chile and the United States have long had one of the best parties in respect of industrial standards, including obliga- economic and political bilateral relationships in the West- tions to endeavor to harmonize technical regulations and to ern Hemisphere. Originally, the plan was for Chile to join give positive consideration to accepting as equivalent the the North American Free Trade Agreement (NAFTA), and technical regulations of the other party. For TBT matters, there was strong support from both Mexico and Canada the parties agree to work together to resolve problems for this idea. Opposition from several quarters in the resulting from differing conformity assessment procedures United States sidetracked that plan, and both Canada and and to share information through contact points estab- Mexico negotiated separate bilateral PTAs with Chile lished for the purpose. In the SPS area, obligations are before the United States and Chile sat down at the negoti- more significant, and the parties have created a standing ating table. The PTA entered into force in 2004. Agricul- Expert Group on Sanitary and Phytosanitary Measures and tural trade issues and quarantine questions are important Food Standards. This body is charged with consulting on to both parties in this agreement, and it is altogether natu- requests for recognition of equivalence of SPS measures ral that forums would be established through the PTA to and promoting resolution of disputes that arise in connec- address these issues. tion with SPS measures. A unique and apparently valuable The Chile–United States PTA relies on a living agree- feature of the SPS chapter is its provision, in Article 607.4, ment approach to resolving SPS and TBT issues, in both for cooperation on a product trace-back system for notifi- cases establishing committees mandated to enhance mutual cation of noncompliance of imported consignments for understanding, consult on matters related to the develop- commodities subject to SPS measures. ment and implementation of TBT and SPS matters, and 226 Andrew L. Stoler coordinate technical cooperation programs. The SPS com- it is not surprising that the chapters dealing with SPS and mittee is also charged with reviewing progress on address- TBT matters in this PTA are more detailed than in some ing SPS matters arising in the bilateral relationship. The other agreements and that the PTA has an important proac- TBT committee’s responsibilities include consulting on tive character. matters that come up under the agreement, and where such The SPS chapter establishes a joint management com- consultations relate to a dispute between the parties, they mittee charged with overseeing implementation of the constitute consultations for the purpose of the operation of chapter, with responsibilities that include drawing up for the PTA dispute settlement chapter. each party a priority order for consideration of market The PTA’s TBT chapter also contains provisions designed access requests by the other party, including the under- to facilitate negotiations on recognition of conformity taking of risk analyses. Areas addressed in considerable assessment procedures, and it enhances transparency detail include adaptation to regional conditions; the over and above that provided through the multilateral acceptance of the other party’s measures as equivalent; TBT Agreement in the WTO. verification, certification, and import check procedures; The PTA affirms that the basic binding obligations of and a well-developed cooperation plan. Because the PTA the parties are those of the WTO TBT and SPS Agreements. entered into force relatively recently, China and New Instead of hard-core binding obligations, the PTA’s provi- Zealand are still negotiating certain aspects of the imple- sions in these areas establish cooperative frameworks for menting arrangements, but concrete results have already discussion and resolution of problems that might arise in been achieved (box 11.3). bilateral trade. The PTA created a joint TBT committee to oversee Recourse to the PTA dispute settlement provisions is implementation, identify priority sectors for enhanced not permitted for any matter that might arise under the cooperation, monitor the progress of work programs, and SPS chapter, meaning that all serious SPS disputes would facilitate technical consultations. Other provisions in the need to be dealt with only under the WTO DSU. Obliga- TBT chapter favor the use of international standards, tions under the TBT chapter could presumably be taken to where possible, and the acceptance of each others’ technical the PTA-specific dispute settlement provisions, but hard regulations and conformity assessment procedures. Provi- obligations exist only with respect to transparency and sion is also made for regulatory cooperation in a number national treatment in accreditation of conformity assess- ment bodies. Discussions in the context of the PTA’s SPS Committee Box 11.3. Impact of SPS Measures in the China–New reportedly led to the easing of Chilean SPS restrictions on Zealand PTA citrus fruits, creating a new market opportunity for Sun Pacific Shippers Sales, a company that is now able to access In the Joint Report prepared by China and New Zealand on the occasion of the Two Year Review of the Free Trade the Chilean market in the off-season for citrus.25 Similarly, Agreement, the Parties note that the following are among the agreement allowed Chile to raise issues concerning the arrangements concluded on new or improved market access: plant registration in exportation, where a certification sys- tem has been established to address SPS matters.26 • Arrangement on New Zealand Product Process Hygiene Requirements for Processing Edible Tripe Products for Export from New Zealand to the People’s Republic of China; China–New Zealand Free Trade Agreement • Cooperation Arrangement on Management of Sanitary Measures Regulating the Import of Dairy Products from In the Asia-Pacific region, China has been actively pursuing a New Zealand; program of negotiating PTAs with its main trading partners. • Official Assurance Programme (OAP) for the Export of Pears from China to New Zealand; Part of China’s motivation for these agreements is to gain • Agreed electronic certificate for live seafood exports from formal recognition from its partners of its market economy New Zealand to China; status, an acknowledgment that it sets as a precondition for • Official Assurance Programme for the Export of Table Grapes (Vitis vinifera) from the People's Republic of China the commencement of PTA negotiations. For China’s part- to New Zealand; and, ners, of course, the motivation is access to the huge • Official Assurance Programme for the Export of Fresh Chinese market. In its negotiations with China, New Processed Onions (Allium cepa) from the People’s Republic of China to New Zealand. Zealand was keenly aware that Australia was also negotiating Source: “China–New Zealand Free Trade Agreement, Two-Year a PTA with the Chinese. New Zealand won the race, and its Review: Joint Report,� p. 15, at http://www.chinafta.govt.nz/ PTA entered into force in 2008. SPS measures are a critically 3-Progressing-the-FTA/3-Moving-forward/index.php. important trade concern in both China and New Zealand, so TBT and SPS Measures, in Practice 227 of important areas, for enhanced transparency, and for for Mercosur-specific work plans and decisions affecting technical assistance. intra-Mercosur trade. The PTA contains an agreement on mutual recognition Before the end of the Uruguay Round, for example, of conformity assessment for electrical and electronic Mercosur’s member countries adopted, through Mercosur equipment (EEEMRA). Before the implementation of that Decision 6/93, their own agreement on SPS measures agreement, Chinese exports to New Zealand had to be (Acuerdo Sanitario y Fitosanitario entre los Estados Partes tested against New Zealand standards, and New Zealand del Mercosur, ACSAFIM). The agreement contained provi- exports to China had to be tested, inspected, and certified sions on harmonization with international standards for by Chinese conformity assessment bodies. The EEEMRA SPS protection, a provision facilitating acceptance of other gives suppliers in both countries an alternative way of countries’ measures as equivalent, rules on risk assessment demonstrating compliance with electrical safety and elec- and regionalization, and other provisions that later tromagnetic compatibility regulatory requirements.27 appeared in the WTO SPS Agreement. When the WTO SPS The SPS and TBT provisions of this PTA are considered Agreement was adopted by Mercosur member countries, binding by the parties and are subject to dispute settlement. the Mercosur Council approved the WTO Agreement as Implementation of plans and work programs still under the operative agreement in Mercosur and denounced discussion are likely to make the PTA provisions even more ACSAFIM. binding over time, as the parties progressively adopt meas- For SPS measures, Mercosur Resolution 60/99 sets out ures to further facilitate trade between them. principles, directives, criteria, and parameters for SPS Although the TBT and SPS chapters lack dispute settle- equivalence agreements among Mercosur member coun- ment provisions, the obligations of those chapters appear tries. The resolution starts with an indication that the to be fully subject to recourse under the dispute settlement basic principles to be applied in the Mercosur environ- provisions in Chapter 16 of the PTA. There appears to be a ment are those of the WTO and the WTO SPS Agreement. soft-law option for both SPS and TBT issues. One function Key principles specifically referred to are the desirability of the Joint Management Committee for SPS is consulta- of acceptance of others’ measures as equivalent where the tion with a view to resolving SPS issues arising in bilateral protection level is as high as that in the importing coun- trade; a similar provision exists with respect to the Joint try; the need to facilitate reasonable access for inspectors TBT Committee. For TBTs, there is also an option for spe- from the importing country; nondiscrimination; and cialized technical consultations, which are explicitly stated proportionality in measures, which should in all cases not to prejudice rights under Chapter 16. The dispute set- be science based. Referenced international norms for the tlement language of the PTA indicates that the complain- purpose of equivalence agreements include those of the ing party can choose to turn to either the PTA or the WTO Codex Alimentarius, the World Organisation for Animal forum for dispute resolution. Health (OIE), and the Plant Health Committee of the Southern Cone (COSAVE, Comité de Sanidad Vegetal del Cono Sur).28 South-South PTAs In the TBT area, the Mercosur countries decided to When TBT and SPS provisions are present in South-South build on the basics of the WTO Agreement through an PTAs, they tend to refer mainly to the WTO Agreements. internal work program aimed at reducing intraregional This is shown here by the review of three representative barriers to trade by harmonizing technical regulations South-South PTAs: Mercosur, ASEAN, and the Chile– and setting up procedures to govern the recognition of China FTA. different conformity assessment processes as equivalent (Decision 56/02). Key terms in this decision are taken directly from the WTO TBT Agreement. (A Mercosur Mercosur technical regulation has the same meaning as a technical Mercosur, which links Argentina, Brazil, Paraguay, and regulation as defined in Annex 1 of the TBT Agreement.) Uruguay, dates from the December 1991 Treaty of Asun- A basic principle of the Mercosur decision is that in the ción. The economic aspects of the arrangements in the elaboration or revision of Mercosur technical regulations group have evolved over time, and significant amendments and conformity assessment procedures, the PTA partners were made to Mercosur’s regional rules after the conclusion must take as the basis of their work the general principles of the Uruguay Round of multilateral trade negotiations. and rules of the TBT Agreement. The decision also Mercosur’s basic approach to TBT and SPS matters has obliges Mercosur countries to take international stan- been to use the WTO Agreements in these areas as the basis dards into account where applicable and to apply adopted 228 Andrew L. Stoler technical regulations in the same way to commerce plastic chairs of Brazil’s adoption of legislation requir- among Mercosur states and to imports from third coun- ing mandatory quality certification.31 tries (the national treatment principle). The binding provisions in Mercosur appear to be more ASEAN Trade in Goods Agreement or less those of the WTO TBT and SPS Agreements. The decisions and resolutions specific to Mercosur are more The countries of the Association of Southeast Asian Nations process oriented and focus on practical steps, such as how (ASEAN) have not had generally applicable mutual obliga- to develop a Mercosur technical regulation. tions with respect to SPS and TBT matters, apart from the Mercosur’s basic approach to dispute settlement is gov- WTO Agreements. They have relied instead on the negotia- erned by Decision 37/03, which sets out the rules and regu- tion and implementation of sectoral MRAs. This changed lations of the Olivos protocol for the settlement of dis- with the implementation of the new ASEAN Trade in putes. Interestingly, Article 1 of the annex to that decision Goods Agreement (ATIGA), which contains new obliga- indicates that Mercosur members have the option of tions in both the TBT and SPS areas. choosing the forum for settlement of their disputes. Evi- In the TBT area, for example, ATIGA obligates ASEAN dently, they have the option of pursuing dispute settlement member governments to follow the TBT Agreement’s Code under Mercosur rules or taking the matter up under the of Good Practice, to use international standards where WTO DSU. possible, and to ensure that technical regulations are not When a dispute is submitted to resolution under Mer- adopted in ways that frustrate trade within ASEAN. Tech- cosur rules, a number of options are open to the parties: nical regulations, where applicable, must be applied in direct negotiations between the disputing parties; the ways that facilitate the implementation of any ASEAN sec- intervention of the Grupo Mercado Común (which is toral MRAs, and conformity assessment procedures are optional); a WTO-like ad hoc arbitration procedure; and, expected to be consistent with international standards and finally, a review procedure by the Tribunal Permanente de practices. A unique feature of ATIGA is the establishment Revisión del Mercosur. If judgments under the arbitration of a postmarket surveillance system that is supported by procedure selected are not adhered to by the losing party, alert systems designed to ensure ongoing compliance on the Mercosur dispute settlement procedure provides for the part of producers. compensatory measures, which must be proportional to In the SPS area, ATIGA obligates members of ASEAN to the harm suffered by the winning party. be guided by international norms and standards in their Research undertaken in 2003–04 for Mercosur identi- SPS-related activities and encourages member govern- fied more than 110 TBT and SPS practices in the region ments to develop equivalence agreements and explore that acted as barriers to trade (Zago de Azevedo 2004). To additional opportunities for intra-ASEAN cooperation. address these issues, the Mercosur countries embarked on a The new agreement is generally considered to be bind- harmonization and mutual recognition project. Hundreds ing in nature. Its provisions make frequent use of the word of Mercosur technical regulations exist today, and guide- “shall,� which is interpreted within ASEAN as a mandatory lines and guides have been developed for the mutual recog- obligation that must be complied with. Where it was felt nition of conformity assessment procedures. Considerable that commitments could be less rigid, other expressions, progress has been made in harmonizing SPS regulations in such as “endeavor to,� were used in place of “shall.� Exam- the region. ples of such wording are, “Member States shall develop and In the Mercosur Trade Commission, consultations have implement a Marking Scheme, where appropriate, for been initiated between and among members with a view to products covered under the ASEAN Harmonised Regula- exchanging information in an effort to resolve standards tory Regimes or Directives,� but “Member States are disputes.29 Two recent cases illustrate the process: encouraged to actively participate in the development of international standards.� • Argentina initiated a complaint concerning unilateral With the entry into force of ATIGA, its obligations are modifications by Brazil and Uruguay to a Mercosur tech- subject to potential dispute settlement action under the nical regulation dealing with toys.30 In Argentina’s view, ASEAN Protocol on Enhanced Dispute Settlement Mecha- the actions by its Mercosur partners negatively affected nism, signed in 2004. Although this is considered a hard- the harmonization process in Mercosur and prejudiced law agreement, backed by dispute settlement, it is also the interests of Argentine manufacturers. the case that ASEAN traditionally favors resolution of • In another case, Paraguay requested consultations with problems through consultation—an approach explicitly Brazil regarding the negative impact on its exports of referred to in the description of the mandate of the TBT and SPS Measures, in Practice 229 ASEAN Committee on SPS Measures established by Arti- Implementation Costs and Technical cle 82 of ATIGA. Assistance Needs Implementing TBT and SPS provisions could be costly, Chile–China FTA especially for exporters. If a firm has to adjust its produc- tion facilities to comply with diverse technical require- Chile is among the most active countries in the world when ments in individual markets, production costs per unit are it comes to pursuing PTAs. The Chileans saw a PTA with likely to increase. This imposes handicaps, particularly on China as helping to lock in Chile’s important position in small and medium-size enterprises. Conformity assess- China as a copper supplier and as an opportunity for China ments are also costly. Compliance with technical regula- to use the PTA with Chile as a gateway for Chinese business tions generally needs to be confirmed, and this may be in South America. The PTA between Chile and China, in done through testing, certification, or inspection by labo- operation since October 2006, is a modern agreement in ratories or certification bodies, usually at the company’s which the SPS and TBT provisions are tied closely to those expense. Information needs impose other costs, including in the WTO Agreements. This extends to the adoption of the costs of evaluating the technical impact of foreign reg- WTO definitions, phrases, and principles. In both the SPS ulations, translating and disseminating product informa- and TBT sections, the PTA addresses such issues as the tion, training experts, and so on. Finally, exporters may be recognition of equivalence in technical regulations and subject to surprise costs because they are usually at a disad- conformity assessment, as well as notification and trans- vantage vis-à-vis domestic firms, in terms of adjustment parency provisions. costs, when confronted with new regulations. The parties to the PTA consider that it is legally binding The major costs to governments and exporters in the on both countries’ central governments and regional TBT and SPS areas do not arise out of the technical imple- authorities. Should a dispute arise under the agreement, mentation of the WTO or PTA agreements themselves but the parties may have recourse to the PTA dispute settle- from compliance with trading partners’ SPS and TBT ment provisions, ensuring the legally binding nature of the measures. Although data on the actual cost of compliance TBT and SPS provisions. with PTA partner measures are not readily available, exam- As with China’s PTA with New Zealand, the SPS and ples of the cost of compliance with specific TBT and SPS TBT obligations can give rise to a complaint under the measures, and of the cost of technical assistance provided PTA’s dispute settlement chapter even if the SPS and TBT pursuant to specific PTAs, could be cited. sections themselves make no provision for dispute resolu- For instance, earlier in this chapter, mention was made tion. The complaining party has the option of choosing of the need of Moroccan citrus and tomato exporters to whether to pursue a complaint under the WTO DSU or meet various nongovernment standards in the EU market. under the PTA dispute settlement provisions. One of these standards is the private standard GLOBAL- The implementation of the SPS chapter in the GAP, relating to good agricultural practices; its certificates Chile–China FTA seems to have helped facilitate bilateral are increasingly required by European retailers as a condi- trade in certain products through expanded trade oppor- tion for accepting Moroccan products.34 Aloui and Kenny tunities. Chile and China have successfully conducted their (2005) estimated the cost to a 10-hectare tomato farm of first bilateral meeting on the implementation of the SPS meeting the standard at US$71,000, or 3 percent of the free chapter. Three protocols on quarantine were signed during on board (FOB) value of the farm’s exports; of this amount, the official visit of the Chilean president to China in 2008; US$20,000 represents annually recurring costs. they deal with quarantine for exports of cherries and plums The WTO Secretariat has attempted to measure more from Chile to China and for exports of citrus and shallots systematically the cost and need for technical assistance from China to Chile, and with inspection and quarantine for related to compliance with standards-related measures. The imports and exports of pork, milk, and dairy products data collected show that these costs can be substantial but between China and Chile.32 also that the costs to developing countries can be mitigated Chile considers the conclusion of these protocols very to a significant degree by technical and capacity-building positive developments in the relationship, since they for- assistance provided by developed countries (box 11.4). mally open the Chinese market to Chilean exports of cher- Many of the PTAs reviewed in this chapter are living ries, plums, and milk and dairy products and define the agreements in the sense that they create bilateral or sanitary requirements for pork exports.33 For their part, regional institutions, such as the committees and work- the Chinese expect to see a growth in their exports to Chile ing groups established in the U.S.–Australia FTA, the of shallots and citrus. 230 Andrew L. Stoler Box 11.4. WTO Assessment of TBT Box 11.5. Using the PTA’s Living Agreement Implementation Costs Institutions for Capacity Building: An Example In 2002, the World Trade Organization (WTO) surveyed 45 The United States has entered into a PTA with the Central developing countries’ priorities for technical assistance and America Free Trade Agreement plus the Dominican Republic capacity building in the area of technical barriers to trade (CAFTA-DR). Article 7.8 of this PTA establishes a Committee (TBTs). Although the survey was conducted with reference on Technical Barriers to Trade. Among the explicitly stated to the implementation of the multilateral TBT Agreement, functions of the committee are “enhancing cooperation in in many cases the survey results would apply as well to the the development and improvement of standards, technical implementation of TBT provisions in PTAs. regulations, and conformity assessment procedures and, as The responses enabled the WTO Secretariat to identify appropriate, designing and proposing mechanisms for seven main areas of needs and specific technical assistance technical assistance of the type described in Article 11 of the activities to help meet these needs. In the order of frequency TBT Agreement, in coordination with the Committee on of response, the perceived needs are: Trade Capacity Building, as appropriate.� The scope of assistance referred to in Article 11 of the • Assistance in infrastructure and capacity building: 43 out TBT Agreement is very broad and includes assistance with of 45 responses (96 percent) the preparation of technical regulations, support for the • Improved knowledge of the TBT Agreement, including establishment of national standardizing bodies, and Annex 3 (the Code of Good Practice for the Preparation, assistance with conformity assessment procedures. Adoption and Application of Standards), and dissemination The committee’s discussions and procedure are and increased awareness of the agreement: 33 responses supplemented by regularly updated national action plans (73 percent) for trade capacity building prepared by the developing- • Exchange of experience among members, and bilateral country parties to the PTA. As TBT and SPS issues or contact and cooperation: 33 responses (73 percent) implementation problems arise, they can be added to the • Assistance with the effective implementation of the TBT updated action plan. Agreement: 32 responses (71 percent) • National and regional coordination and strategy: Sources: Lesser 2007, 24–25; http://www.ustr.gov/trade- 29 responses (64 percent) agreements/free-trade-agreements/cafta-dr-dominican-republic- • Assistance in participating in the work of the WTO central-america-fta/cafta-dr-tcb; http://www.ustr.gov/trade- TBT Committee and other organizations: 20 responses agreements/free-trade-agreements/cafta-dr-dominican-republic- central-america-fta/final-text. (44 percent) • Help with market access questions arising out of TBT measures: 16 responses (36 percent). Source: For the survey, G/TBT/W/178, July 18, 2002; for the results, G/TBT/W/193, February 10, 2003. In nearly all of the PTA provisions examined in this chapter, frequent reference is made to the WTO SPS and TBT Agreements and to the desirability of relying on international standards and procedures. In a very real New Zealand–China FTA, and others. Another good exam- sense, the many years of work and cooperation at the mul- ple is described in box 11.5. Often, the PTA committees are tilateral level in these areas have made it easier to manage tasked with drawing up work plans or prioritizing issues implementation of PTA-level provisions. Of course, fun- for resolution. damentally different climatic, geographic, or technological Effective participation in the bilateral process implies factors in different countries will always limit the appro- the need for a certain amount of capacity building, at least priateness of international standards and harmonization in those developing countries that have less experience to some degree, particularly when the PTA partners are with TBT and SPS measures. For example, before ATIGA not neighbors. entered into force on 17 May 2010, according to the In only about half of the PTAs surveyed for this chapter ASEAN Secretariat, a significant amount of technical assis- did the parties make the agreements’ TBT and SPS provi- tance was already taking place within ASEAN to prepare sions legally binding and enforceable through dispute set- for the implementation phase. tlement, and in a number of cases, provisions relating to Over time, as government officials gain greater experi- SPS were specifically excluded from the possibility of dis- ence with implementation of the TBT and SPS provisions pute settlement. Where dispute settlement does apply to in PTAs, the incremental effect of participation in a new or TBT or SPS provisions, the parties to the agreement expanded PTA should decrease considerably. For example, usually have the choice of bringing a dispute before the as Chinese officials become familiar with the operation of WTO DSU or the PTA-specific dispute settlement mecha- the agreement with New Zealand and their responsibilities nism. Parties to the Andean Community, for instance, under that PTA, they will be that much more prepared to have relied heavily on the Community settlement mecha- take on whatever provisions are eventually agreed in their nism for their TBT and SPS disputes (box 11.6). Yet, in PTA negotiations with Australia. general, PTAs favor dispute avoidance (i.e., working out TBT and SPS Measures, in Practice 231 3. If one partner is less developed than the other, the PTA Box 11.6. Dispute Settlement of TBT and SPS should incorporate technical assistance and capacity- Measures in the WTO and within the Andean building measures to assist the institutions and exporters Community of the developing-country partner. In negotiating a PTA, governments should recognize that deeper integration Over the period 1995–2004, just 12 cases involving TBT or SPS measures were brought under the WTO’s Dispute and the resolution of standards-related problems will Settlement Understanding. In a roughly comparable period take time and will require considerable bilateral work. (1997–2004), 24 cases out of a total of 104 legal cases brought to dispute settlement in the Andean Community A PTA that aims to be effective should incorporate bilat- involved TBT or SPS measures, with 88 percent of these eral institutions (committees and the like) that have a disputes related to SPS matters. Commonly noted mandate to deal with standards-related questions over procedural issues at dispute included time through harmonization, equivalence, or mutual • Delays of more than five months in granting SPS permissions, whereas the maximum time frame for recognition techniques. Ideally, the institutions estab- granting permission is 10 days lished in the PTA should also be capable of helping to • Validity limited to 60 days on SPS permissions, although resolve trade-related problems arising out of exporters’ the minimum validity period established by Andean Community regulations is 90 days need to comply with private standards in an importing • Establishment of complementary requirements for country’s market. granting SPS permission, beyond Andean Community 4. If technical regulations and conformity assessment pro- legislation • Grants of permissions for only a small portion of the cedures cannot be harmonized, it is important for the products, with other products subject to indefinitely purposes of the PTA that the parties work to eliminate pending deliberation without any stated objections on SPS duplicate or multiple measures or mandatory tests for grounds. the same product. This is particularly crucial for small In some instances, it was indicated that the complainant perceived the procedural problem in granting SPS approval and medium-size enterprises that cannot afford the as intentional or as a hidden restriction. high cost of meeting differing regulations and testing Source: Fliess and Lejarraga 2005, 246–47. regimes. Mutual recognition agreements are important tools in this respect. 5. Transparency regarding SPS standards in international trade is very important for businesses and consumers. differences through consultation or within a technical PTA partners should consider enacting WTO+ notifica- group) over litigation, even if litigation is an option. In tion obligations and a commitment not to implement some cases, implementing committees are charged with any technical regulation or SPS measure until it has pursuing resolution of disputes—implicitly, through coop- been published and comments from the PTA partners erative consultation. have been taken into account. 6. The PTA should be a living agreement with a commit- ment to a work plan or to prioritization of problem res- Conclusions olution through harmonization, mutual recognition, Research into the practice of addressing TBT and SPS equivalence measures, and other policy tools that enable measures in PTAs suggest that such agreements converge elimination or mitigation of trade-related problems with, and support, the multilateral trading system. To over time. Ideally, the work program should also be ensure that this does happen, PTAs should include, where capable of addressing problems relating to compliance feasible, a number of important best-practice provisions. with private standards. 7. PTA provisions on TBT and SPS matters should be 1. The parties to the PTA should undertake to use inter- legally binding through a judicious combination of soft national standards whenever possible, as doing so and hard law. The agreement should provide a pathway guarantees a high level of protection in the integrated that permits an evolution and deepening of integration market and makes it easier for third parties to trade in over time by allowing the gradual resolution of TBT and that market. SPS issues in the bilateral relationship. Such a pathway 2. If the parties to the PTA decide on an approach of har- should be considered an integral part of any PTA that monizing their standards and conformity assessment aims to deal effectively with standards, certification, and procedures, they should accept that it might be neces- conformity assessment problems. Eventual recourse to sary to limit harmonization to essential health and the PTA dispute settlement provisions should be an safety standards and rely on mutual recognition and option, in addition to recourse to the WTO Dispute Set- equivalence techniques for other areas. tlement Understanding. 232 Andrew L. Stoler 8. PTA parties should agree to an overall commitment 16. European Commission, “Proposal for a Council Decision on the Signing of an Agreement between the European Community and its whereby technical regulations and conformity assess- Member States, of the one part, and the Republic of South Africa, of the ment procedures are always applied on a national other part, amending the Agreement on Trade, Cooperation and Develop- treatment basis. Third parties whose technical regula- ment,� COM(2008)50 final, Brussels, February 4, 2008, 2. tions and conformity assessment procedures can be 17. European Commission, “Cooperation between the European Union and South Africa,� Brussels, July 23, 2009. UE-ZA 4903/09, demonstrated as being equivalent to the level agreed to http://register.consilium.europa. by the PTA partners should be permitted to benefit from 18. Council of the European Union, “Second South Africa-European the arrangements between the partners. A commitment Union Summit,� Kleinmond, South Africa, September 11, 2009, 13231/09 (Presse 266), http://register.consilium.europa. to open regionalism would help to ensure that PTAs sup- 19. Council of the European Union, “South Africa–European Union port the multilateral system. Strategic Partnership Joint Action Plan,� Brussels, May 15, 2007, http://register.consilium.europa. 20. Ibid., 5. Notes 21. “Europe and South Africa Development Partners, 2007/8,� http://www.eusa.org.za/en/eu_and_country/EU20078.pdf. 1. The terms “standard� and “technical regulation� are used through- 22. Council of the European Union, “Cooperation between the Euro- out this chapter. The difference is that whereas conformity with standards pean Union and South Africa,� Brussels, November 24, 2008, UE-ZA is voluntary, technical regulations are by nature mandatory. The two types 4904/09, http://register.consilium.europa. of provision have different implications for international trade. If an 23. Ibid. imported product does not fulfill the requirements of a technical regula- 24. Australia, Department of Foreign Affairs and Trade, “Thailand tion, it will not be allowed to be put on sale. In the case of standards, non- Country Brief,� December 2008, Canberra, http://www.dfat.gov.au/geo/ complying imported products will be allowed on the market, but their thailand_thailand_brief.html, accessed February 23, 2009. market share may be affected if consumers’ prefer products that meet local 25. Office of the U.S. Trade Representative, “The U.S.–Chile Free Trade standards, such as quality or color standards for textiles and clothing. The Agreement: An Early Record of Success,� http://www.ustr.gov/Document WTO Web site pages offer a detailed presentation of technical barriers to _Library/Fact_Sheets/2004/The_US-Chile_Free_Trade_Agreement_An_ trade (http://www.wto.org/english/tratop_e/tbt_e/tbt_info_e.htm) and of Early_Record_of_Success.html, accessed February 20, 2009. sanitary and phytosanitary measures (http://www.wto.org/english/tratop_ 26. Office of the U.S. Trade Representative, “The U.S.–Chile Free e/sps_e/spsund_e.htm). Trade Agreement: Excerpts from Trade Advisory Committee Reports,� 2. Office of the U.S. Trade Representative, “U.S.-Australia Free Trade http://www.ustr.gov/assets/Trade_Agreements/Bilateral/Chile_FTA/Reports/ Agreement, Report of the Agricultural Technical Advisory Committee in asset_upload_files85_4956.pdf, accessed February 22, 2009. Fruits and Vegetables,� http://ustraderep.gov/Trade_Agreements/Bilateral/ 27. Ministry of Foreign Affairs and Trade, “Guide to the New Australia_FTA/Reports/Section_Index.html, accessed January 1, 2009. Zealand-China Free Trade Agreement,� Wellington, 2008, 34. 3. Euromed Report 64, July 9, 2003, 3-4. 28. The Codex Alimentarius, which comes under the auspices of the 4. Morocco, “Fiche signalétique de projet de jumelage: appui a la Food and Agriculture Organization of the United Nations (FAO) and direction de la normalisation et de la promotion de la qualite dans l’har- the World Health Organization (WHO), is a collection of internation- monisation et la mise en oeuvre de la legislation technique,� final version, ally recognized standards, codes of practice, guidelines, and other rec- May 2, 2007, Rabat. ommendations relating to foods, food production, and food safety. OIE 5. Ibid, 6. is the intergovernmental organization responsible for improving animal 6. European Commission, Directorate General for Trade, “Inventarisa- health worldwide. COSAVE is a regional organization established by tion of Technical Assistance Provided to Third Countries in the Area of SPS agreement between the governments of the Mercosur parties within the by the European Union and Its Member States,� October 15, 2009, Brussels. framework of the International Plant Protection Convention. 7. European Union, “Joint Communiqué, VII EU-Chile Association 29. The Trade Commission assists the Mercosur executive body in Committee, 2009,� http://www.delchl.ec.europa.eu/en/whatsnew/NEWS applying the instruments of common trade policy agreed to by the mem- _2009_15_10_association_comittee.htm. ber states for the operation of customs unification. 8. EUR-Lex, Access to European Union Law, http://eur-lex.europa.eu/ 30. For Brazil, CCM XCVII, Acta No 06/08, Consultation No 08/07; LexUriServ/LexUriServ.do?uri=CELEX:32004D0907:EN:NOT. for Uruguay, CCM XCVII, Acta No 08/07; http://www.mercosur.int/ 9. Chile, Dirección General de Relaciones Económicas Interna- msweb/portal%20intermediario/es/index.htm. cionales (DIRECON), “Relaciones Multilaterales: Noticias,� http://www 31. CCM CV, Acta No 08/08, Consultation No 06/08, http://www .direcon.cl/index.php?accion=prensaArticulo&id_prensa=153. .mercosur.int/msweb/portal%20intermediario/es/index.htm. 10. Fundación Chile, “Presidenta Bachelet Inauguró Moderno Centro 32. “Communiqué between the People’s Republic of China and the de Medición de Residuos de Fundación Chile,� http://www.fundacionchile Republic of Chile,� http://hr.china-embassy.org/eng/zxxx/t427229.htm. .cl/portal/page?_pageid=113,232271&_dad=portal&_schema=PORTAL&p_ 33. “Visita de Estado de S.E. la Presidenta de la República Michelle item_id=5166352&p_area_id=206271. Bachelet a la República Popular China,� http://www.echilecn.com/eg/ 11. European Commission, “Chile Country Strategy Paper html/31/2008-5/78.html. 2007–2013,� ec.europa.eu/external_relations/chile/csp/07_13_en.pdf. 34. The standard had been called EUREPGAP; the name was changed 12. Official Journal of the European Communities, L 311/18, December in 2007. 12, 1999. 13. European Commission, “Toward an EU-South Africa Strategic Partnership,� communication to the European Council and European Bibliography Parliament, COM(2006) 347 final, Brussels, June 28, 2006, http://eur- lex.europa.eu. Aloui, Omar, and Lahcen Kenny. 2005. “The Cost of Compliance with SPS 14. Article 103 provided for a review of the TDCA within five years of Standards for Moroccan Exports: A Case Study.� Agriculture and the agreement’s entering into force. Rural Development Discussion Paper, World Bank, Washington, DC. 15. European Commission, communication to the European Council, APEC (Asia-Pacific Economic Cooperation) Secretariat. 2000. “Information COM(2006) 348 final, Brussels, June 28, 2006, http://eur-lex.europa.eu. Notes on Good Practice for Technical Regulation.� APEC, Singapore. TBT and SPS Measures, in Practice 233 Benavides, Daniela, and Sandra Jerez. 2007. “Bienestar animal (BA): WP(2007)12/FINAL, Organisation for Economic Co-operation and responsabilidad ética, exigencia commercial y atributo de valor.� Development, Paris. Boletin Veterinario Oficial 8 (1 semestre). http://www2.sag.gob.cl/ Maskus, Keith, Tsunehiro Otsuki, and John S. Wilson. 2000. “Quantifying Pecuaria/bvo/BVO _8_I_semestre_2007/articulos/bienestar_animal the Impact of Technical Barriers to Trade: A Framework for Analysis.� .pdf. Policy Research Working Paper 2512, World Bank, Washington, DC. Budetta, Michelle, and Roberta Piermartini. 2009. “A Mapping of Nidhiprabha, Bhanupong. 2002. “SPS and Thailand’s Exports of Regional Rules on Technical Barriers to Trade.� In Regional Rules in the Processed Food.� Revised version of a paper presented at the Project Global Trading System, ed. Antoni Estevadeordal, Kati Suominen, and Launching Workshop, Royal Princess Hpoteh, Bangkok, October 1–3. Robert Teh. Cambridge, U.K.: Cambridge University Press. Turner, J., F. Maplesden, B.Walford; and S. Jacob. 2005. “Tariff and Non- Fliess, Barbara, and I. Lejarraga. 2005. “Looking beyond Tariffs: Nontariff tariff Barriers to New Zealand’s Exports of Wood-Based Products to Barriers of Concern to Developing Countries,� ch. 7. OECD Trade China.� New Zealand Journal of Forestry 50 (1). Policy Studies, Organisation for Economic Co-operation and Devel- WTO (World Trade Organization). 2005. World Trade Report 2005: Trade, opment, Paris. Standards and the WTO. Pt. 2, Geneva: WTO. Lesser, Caroline. 2007. “Do Bilateral and Regional Approaches for Reducing Zago de Azevedo, André Filipe. 2004. “Mercosur: Ambitious Politics, Poor Technical Barriers to Trade Converge towards the Multilateral Trading Practices.� Brazilian Journal of Political Economy 24 (4, October– System?� OECD Trade Committee Working Paper 58, TAD/TC/ December): 584–601. 12 Services Aaditya Mattoo and Pierre Sauvé An extensive literature exists on the costs and benefits of relationship between preferential and multilateral regimes integration agreements concerning trade in goods, but for services trade. This chapter considers a number of hardly any analysis of the implications of such agreements questions—both theoretical and policy related—arising on trade in services has been carried out. Such a gap is sur- from the study of the PTA-WTO divide in services trade. prising, given the strong growth witnessed since the mid- Do the tools of economic analysis developed for studying 1990s in the number of preferential trade agreements the effects of preferences in goods trade yield meaningful (PTAs) featuring detailed disciplines on trade and invest- insights in the services field? To what extent, and in what ment in services. The recent proliferation of PTAs covering form, can developments in preferential agreements inform services is evidence of heightened policy interest in the approaches to rule making and market opening under the contribution of efficient services sectors to economic General Agreement on Trade in Services (GATS)? Do the development and of a growing appreciation of the gains observed differences in negotiating architectures across likely to flow from the progressive dismantling of impedi- services-related PTAs matter? What is the likelihood that ments to trade and investment in services. such differences may inform the evolution of the WTO’s To date, 76 PTAs featuring provisions on trade and architecture of services rules following the completion of investment in services have been notified to the World the Doha Development Round of trade negotiations? Does Trade Organization (WTO), as shown in annex tables the issue of preference erosion arise in services trade, and if 12A.1 and 12A.2. This number represents 28 percent of so, what is its sectoral or modal incidence? Does the prac- all notified PTAs, a proportion slightly higher than the tice of liberalization of services markets suggest that PTAs share of services in world trade (figure 12.1). As figure 12.2 can be likened to “optimal regulatory convergence areas�— illustrates, developed countries are party to just under two- groups of countries whose aggregate welfare is maximized thirds (62 percent) of all PTAs featuring services provi- by regulatory convergence? sions, commensurate with their aggregate share in world Regional and bilateral attempts at developing trade rules services trade. North-South agreements, with both devel- for services continue to parallel efforts at framing similar oped and developing countries as members, account for disciplines in the WTO, under the aegis of GATS. For this 49 percent of such PTAs, and North-North agreements reason, regional and multilateral efforts at services rule among developed countries, for 13 percent. Services-related making are closely—indeed, increasingly—intertwined South-South PTAs have grown noticeably in number and processes, with much iterative learning by doing, imitation, today account for 38 percent of notified agreements. From and reverse engineering. Experience gained in developing a political-economy perspective, the predominance of the services provisions of PTAs has expanded negotiating services-related PTAs involving developed-country partners capacity in participating countries and has built up is consistent with the aggregate pattern of specialization expertise that is available for deployment in a multilat- and the direction of trade and investment in services eral setting. GATS itself remains incomplete, with nego- trade today. tiations pending or stalled in such important areas as If all trade agreements can, in essence, be likened to emergency safeguards, subsidies, government procure- “incomplete contracts,� the incipient multilateral regime ment, and domestic regulation, but regional and bilateral for services is arguably the most incomplete of all. That experimentation has generated useful policy lessons in consideration greatly heightens the salience of studying the comparative negotiating and rule-making dynamics. In 235 236 Aaditya Mattoo and Pierre Sauvé Figure 12.1. Services-Related PTAs as a Share of The efforts that countries devote to developing rules Total PTA Notifications to the WTO, 2010 governing the process of services trade liberalization at the regional level typically follow in the wake of the far- reaching changes in services and investment policy frameworks that have taken hold in many countries in the post–Uruguay Round period. PTA negotiations offer the opportunity to pursue, deepen, or lock in some (or many) 28% of the policy reforms put in place domestically in recent years and to reap the benefits likely to flow from such pol- icy consolidation—notably, in the form of improved 72% investment climates. This chapter takes stock of the more recent wave of PTAs with a view to informing some of the policy choices developing countries will typically confront in negotiat- ing regional regimes for services trade and investment. Although a country’s choice of integration strategy will in goods services most instances be dictated by political considerations, there remains a need for a careful assessment of the eco- Source: World Trade Organization, Regional Trade Agreement (RTA) nomic benefits and costs of alternative approaches to database, http://rtais.wto.org/UI/PublicMaintainRTAHome.aspx. services liberalization. The chapter focuses on two core issues. The first is the economics of regional integration in services. Does services trade differ sufficiently from trade in goods as to require Figure 12.2. Services-Related PTAs Notified to the WTO, by Country Group different policy instruments and approaches in the context of preferential liberalization? Do PTAs allow deeper forms of regulatory cooperation to occur, and in what way? The discussion highlights the importance for third countries of 13% multilateral disciplines on PTAs and the criteria for pre- venting PTAs from being detrimental to nonmembers. The second concern is the political economy of region- alism in services trade. Here, the discussion highlights lessons arising from the practice of preferential liberal- 49% ization in services, drawing on a sample of 55 of the 38% 76 services-related PTAs notified to date in the WTO. Within the sample, 3 PTAs are North-North agreements, 27 are classified as North-South, and 25 as South-South. The Economics of Services Trade in PTAs North-North North-South The economic effects of preferential tariff arrangements South-South are generally well understood (see Baldwin, ch. 3 in this Source: World Trade Organization, Regional Trade Agreement (RTA) volume) and form the core of conventional trade theory. database, http://rtais.wto.org/UI/PublicMaintainRTAHome.aspx. Such, however, is hardly the case with services. The analy- sis of preferential agreements in services trade requires an extension of conventional trade theory in two ways, both of which relate to core distinguishing features of addition, developments in WTO jurisprudence in the services: first, the manner in which trade in services services field have begun to be reflected in patterns of occurs and, second, the form that trade protection takes market-opening commitments found in PTAs—a trend in the sector. that can be expected to deepen as judicial activism under Since services trade often requires proximity between the GATS increases.1 supplier and the consumer, the analysis has to take account Services 237 of preferences extended not just to cross-border trade but Countries can and do impose on foreign providers also to foreign direct investment (FDI) and to individual qualification and licensing requirements that may be more foreign services providers. Another point is that preferential burdensome than is necessary to satisfy otherwise legiti- treatment in services is granted not through tariffs but mate public policy objectives. When these are waived through discriminatory restrictions on the movement of selectively in favor of members of a PTA and denied to labor and capital (e.g., on the quantity or share of foreign nonmembers that would otherwise qualify for the bene- ownership) and through a variety of domestic regulations fits, de facto preferences result. Regulatory preferences such as technical standards and licensing and qualification may arise in all sectors, but they are especially prevalent in requirements. professional and financial services, where domestic regu- Given such differences, can one say that trade in serv- latory requirements and licensing regimes respond to ices differs enough from trade in goods as to modify the information asymmetries. accepted conclusions regarding the economic effects of Measures affecting variable costs. A common effect of preferential liberalization? In particular, what would hap- many restrictive measures in services trade is to increase the pen if a country liberalized services trade faster in a variable costs of operation faced by foreign providers with- regional or bilateral context than at the multilateral level? out necessarily generating equivalent rents. In such cases, To answer such questions, we next review the costs and the analysis of discriminatory regulation can proceed in a benefits of trade preferences arising in services trade and manner analogous to that for tariffs. When tariffs are the examine the scope for regional or bilateral regulatory instruments of protection, the costs of trade diversion can cooperation in this sector. be an important disincentive to concluding preferential liberalization agreements. Despite the increase in consumer surplus from any liberalization, there may still be an aver- Costs and Benefits of Preferential Treatment sion to such agreements because the displacement of high- in Services Trade tariff imports from third countries by low- or zero-tariff The manner in which privileged access is granted in serv- imports from preferential sources implies lost revenue. ices markets depends on the instruments of protection that The situation may differ when the protectionist instru- are in use. By imposing quantitative restrictions on services ment is a regulatory barrier that imposes a cost on the output or on the number of service providers, a country exporter without necessarily yielding corresponding rev- can allocate a larger proportion of the quota to a preferred enues for the government or any other domestic entity. source. For example, countries often preferentially allocate Under such circumstances, which characterize much of freight and passenger quotas in air, land, and maritime services trade (given the regulatory nature of impedi- transport, limit the airtime allocated to foreign broadcasts, ments), there may be little or no cost to granting preferen- and restrict the numbers of foreign telecommunications tial access because there is little or no revenue to lose. In operators, banks, or professionals. such circumstances, preferential liberalization will neces- Another common means of restricting access to service sarily be welfare enhancing. markets that lends itself to preferential treatment is to place Countries outside the preferential arrangement may, conditions on foreign ownership, the type of legal entity per- however, lose. Exemption from a needlessly burdensome mitted, and branching rights. Most host-country govern- regulation implies reduced costs for a class of suppliers and ments accord national treatment to foreign investors after hence a decline in prices in the importing countries. This their establishment, but national treatment rarely applies decline may hurt third-country suppliers, who may suffer before establishment. Host countries are therefore able reduced sales and a decreased producer surplus. to impose performance requirements on prospective for- The analysis of discriminatory regulation is also rele- eign services providers with respect to such things as vant to quantitative restrictions on the sale of services. In training, or employment in managerial-level positions. the case of goods, the quota rents can be appropriated by These requirements can easily be waived for members of a domestic intermediaries such as the importer, rather than preferential arrangement. by the foreign exporter. For many services, intermediation Preferences can also be granted through taxes and subsi- is difficult because the service is not always storable and is dies. Foreign providers may be subject to different taxes directly supplied by producers to consumers. Rents are from nationals and may be denied access to certain subsidy therefore usually appropriated by exporters rather than by programs. These forms of discrimination, too, can be domestic importers. As in the case of frictional measures waived selectively, as is the case, for example, with copro- that increase variable costs, there is typically no cost of duction agreements in audiovisual services. trade diversion to the preference-granting country. 238 Aaditya Mattoo and Pierre Sauvé The main policy implication that emerges from the in the context of progressive liberalization. These consider- above discussion is that when a country maintains regula- ations may affect the preferred mode of entry: acquisition tions that impose a cost on foreign providers without gen- implies less competition than greenfield entry, but it allows erating any benefit (such as improved quality) or revenue domestic firms to extract some rents through the disposal for the government or other domestic entities, welfare is of their assets. necessarily enhanced by preferential liberalization. Non- Liberalization tends to generate gains when all barriers preferential liberalization, however, would yield an even to entry are removed. If only limited entry is allowed, then greater increase in welfare, both nationally and globally, open, nondiscriminatory access—through, for example, the because the service would then be supplied by the most global auctioning of licenses—would have an edge over efficient providers. preferential access, which cannot guarantee that preferential Measures affecting fixed costs. Some regulatory measures (i.e., insider) investors will be the most efficient ones. can have the effect of increasing the fixed costs of entry or Absent liberal rules of origin for investment, the establish- establishment in services markets. Examples include ment of preferences may indeed result in entry by inferior mandatory establishment of a local presence, license fees suppliers. Because the most efficient suppliers (in terms of for entry into the market, and the need to requalify to costs, quality, or both) may generate the greatest positive provide professional services. As with measures affecting externalities, including the dynamic learning properties variable costs, a country is likely to benefit from eliminat- associated with knowledge flows and the associated rise of ing, even on a preferential basis, excessive fixed costs of total factor productivity, the downside risks of preferential entry by removing unnecessary burdensome qualification, liberalization are magnified (Mattoo and Fink 2002). The licensing, and local establishment requirements for profes- ability of nonpreferential liberalization to more readily sional and financial services. secure access to the most efficient suppliers of services is a Regardless of the chosen partners, the presumption that matter of some importance, given the crucial infrastruc- a country will benefit from such initiatives is greater if tural role many services perform and the strong influence of agreements are not exclusionary but are open to all parties their intermediate inputs on economywide performance. able to satisfy the regulatory requirements maintained Preferential liberalization of entry barriers may also lead within the integrating area. The greatest benefits arise if to higher prices for consumers, lower takeover prices for agreements to recognize professional qualifications include domestic assets, or lower license fees for the government all countries that have comparable regulations. The bene- (because the pool of potential buyers is limited). These fits in such instances stem both from increased competi- concerns are likely to be compounded in concentrated tion and from greater diversity of services. markets, which are common in many services industries in Measures restricting the number of service providers. The the developing world. norm in many service industries is for the level of competi- Sunk costs and the sequence of liberalization. Sunk costs tion to be restricted by government regulation. There may are important in goods and services industries alike, but be legitimate reasons to do so: significant economies of location-specific sunk costs—those incurred in supplying a scale may be possible, and some industry segments, such particular market—are arguably higher in a number of serv- as network-based energy, water distribution, and trans- ices sectors, insofar as their provision requires proximity port services, have natural monopoly features. In such between suppliers and consumers. One consequence (which circumstances, the manner in which entry is allowed—by is closely related to the above discussion on barriers to new mergers and acquisitions, or through greenfield (de novo) entrants) is that preferential liberalization may have more investments—can assume considerable significance. durable effects on the nature of competition than in the Interestingly, allowing limited new entry by foreign case of goods. For instance, under an agreement that allows firms, whether this is done preferentially or on a most inferior providers to establish, a country could be stuck favored nation (MFN) basis, may not be welfare enhanc- with such providers even if it subsequently liberalizes on an ing. The main reason is that even though consumers may MFN basis. benefit from the increased competition, the gain may be Sunk costs matter because they have commitment value offset by the transfer of rents from domestic to foreign and can be used strategically by first movers to deter new oligopolists. entrants (Tirole 1998). A firm that establishes a telecom- Restrictions on de novo entry are often imposed with a munications or transport network signals that it will be view to channeling new foreign capital into weak or under- around tomorrow, since it cannot easily dispose of its capitalized domestic institutions (as is common in finan- assets. The commitment value is stronger, the more slowly cial services, for example) and so assist with restructuring capital depreciates and the more firm specific it is. Services 239 Firms allowed early entry into such markets may accu- Regional liberalization can also act as an inducement mulate a quantity of capital sufficient to limit the entry of to FDI. Apart from changing the organization of local new rivals. These incumbency effects may be stronger in industry, if PTAs create large markets and do not impose services with network externalities, such as telecommuni- stringent ownership-related rules of origin, they may help cations, where new entrants must match the technical stan- attract foreign investment when economies of scale mat- dards of the incumbent—standards that the latter may ter. For example, a foreign transport service provider have played a large part in defining. The incumbent may might not find it worthwhile to establish in Latin America also be able to assure itself of the services of the best fran- if each country market were segmented, but it might find chisees by selecting them early on and imposing exclusivity a continentwide integrated market attractive. arrangements on them. Each of these forms of capital One rationale for coverage of services in PTAs is a vari- accumulation enhances first-mover advantages and allows ant of dynamic economies of scale, or of the infant- established firms to prevent, restrict, or retard competition. industry argument. South-South PTAs, in particular, are Because of the importance of sunk costs in many serv- seen as a means of gradual liberalization. Exposure to ices industries, sequential entry (which preferential lib- competition at first in the more sheltered confines of a eralization with restrictive rules of origin may entail) can regional market may help firms prepare for global com- produce very different results from simultaneous entry. petition. This approach improves on traditional protec- If entry is costly, an incumbent may be able to completely tion of infant industries because the integration process deter entry, leading to greater market concentration. Fur- promotes some degree of international competition. Fur- thermore, the first-mover advantage may be conferred on thermore, firms that have become more competitive at the an inferior supplier, which may naturally exploit such regional level are less likely to resist broader liberalization. advantages to establish a position of market dominance, They may even champion subsequent MFN liberalization insulated from more efficient third-country competitors. as they begin to reap the benefits of open markets and How durable such a position may be in practice will encounter the constraints of a regional market. In this depend on the importance of sunk costs relative to differ- sense, as noted by Baldwin and Freund (ch. 6 in this vol- ences in price and quality. ume), PTAs can be seen as building blocks toward multilat- There are two important qualifications to the above eral liberalization (Bhagwati 1990; Lawrence 1991). There reasoning. First, subsequent entry by a more efficient firm is, however, a risk that regional liberalization might create can take place by acquisition, circumventing some of the a new constellation of vested interests that could resist problems linked to first-mover advantages. This has further market opening, raising the concern that region- notably happened in a number of countries in the finan- alism could prove a stumbling block to further multilat- cial sector, especially where first movers may have overbid eral liberalization. GATS offers a way out of this dilemma or sunk excessive costs into setting up their operations in by allowing member countries to precommit to future the early stages of liberalization. Second, in certain serv- multilateral liberalization, signaling a time frame over ices sectors, firms may learn by doing: the experience which regional preferences may be progressively eroded acquired by established operators may reduce their cur- or eliminated. rent costs, enhancing their profitability and discouraging others from entering. Caveats aside, a country needs to Regionalism and Regulatory Cooperation carefully evaluate not just the static costs of granting pref- erential access to a particular partner country but also The gains from PTAs are likely to be significant in areas how the eventual benefits from multilateral liberalization where there is scope for attaining economies of scale, as in are likely to be affected. certain international transport and financial services, and Static and dynamic economies of scale. Combining serv- for promoting increased competition, as in business or ices markets through a regional integration agreement professional services. In principle, these gains can also be can lead to gains arising from a combination of scale realized through MFN liberalization, but in practice, the effects and changes in the intensity of competition. In a integration of markets often requires a convergence of market of a given size, there is a trade-off between scale regulatory regimes. Such convergence might well be more economies and competition: if firms are larger, there are feasible in a bilateral or regional context—for instance, fewer of them, and the market is less competitive. Enlarg- where proximity, whether geographic or in terms of ing the market shifts this trade-off, as it becomes possible income levels or legal traditions, implies closer institu- to have both larger firms and more competition (World tional and regulatory ties. The regulatory intensity of serv- Bank 2000). ices trade makes it necessary to consider whether and how 240 Aaditya Mattoo and Pierre Sauvé PTAs can be conduits for trade- and investment-facilitating tends to be either a precondition or a result of the latter. convergence in domestic regulatory practices. Simply put, Where differences in mandatory quality standards matter, under what circumstances is a country more likely to bene- mutual recognition may be feasible only when there is fit from cooperation in a plurilateral or regional forum a certain degree of prior harmonization of mutually than in a multilateral one? acceptable minimum standards. A similar logic applies to Addressing the regulatory intensity of services trade. The compatibility standards—although there may be no alter- economic case for regulation in services arises essentially native to full harmonization if differences matter, as in from market failure attributable to three kinds of problems: the cases of road-safety standards, railway gauges, and (a) asymmetric information, especially in knowledge- legal procedures. intensive industries such as financial or professional Regulatory cooperation may be more desirable, and is services; (b) externalities, as in tourism, transport, and likely to be more feasible, among a subset of countries water supply; and (c) natural monopolies or oligopolies, than if pursued on a global scale. There is, however, little, especially in network-based services for which access to if any, empirical guidance on the payoffs to regulatory essential facilities is a critical ingredient. cooperation—on the costs and benefits of mutual recog- In the first two cases, national remedies can themselves nition agreements or the deeper harmonization of regula- become impediments to trade if domestic regulatory tory standards. The lack of empirical evidence complicates requirements are needlessly burdensome or are framed so the task of deciding on the scope and depth, as well as the as to tilt competitive conditions in favor of domestic sup- geographic reach and the optimal institutional forms, of pliers. The institution of some variant of a necessity test regulatory cooperation. in services agreements (the purpose of which, as in goods If national standards are not optimal or are insuffi- trade, is to ensure a broad measure of proportionality ciently developed, regional or international harmonization between regulatory objectives and the means of pursuing or standardization can be a way of improving such stan- them), together with strengthened disciplines on trans- dards, as has happened in the financial services field with parency, would enable exporters to challenge the appro- the Basel Accord on capital adequacy. In such situations, priateness of regulatory regimes abroad. Doing so would the best partners for regulatory cooperation are likely to help ensure that domestic regulations serve legitimate be those with the soundest regulatory frameworks. Such objectives, rather than mask protectionist interests, and partners may not always be found within regional com- would thus create benefits for domestic consumers and pacts. Moreover, the standard-setting process can at times users of services. be captured by protectionist interests, in which case con- In the third case, that of natural monopolies and oligop- vergence around “best� regulatory practice can serve a use- olies, it is the absence of regulation (typically, procompeti- ful liberalizing purpose. tive regulation) that can lead to trade problems and Another consideration is that there are gains from regu- directly inhibit or nullify negotiated market access. As latory cooperation, but also costs. The former will domi- negotiations on basic telecommunications services have nate where national regulation can be improved. The shown, international rules on access to essential facilities, aggregate adjustment cost of regulatory convergence and on means of ensuring that dominant suppliers do not depends on the degree of differences between the policy- abuse their market advantages to deter entry and stifle related standards of the countries involved in an integra- competition, can provide significant benefits to consumers tion area. The costs are likely to be smallest when foreign and users of telecommunications services. regulatory preferences are similar and regulatory institu- To ensure that domestic regulations at home and tions are broadly compatible. The benefits of eliminating abroad support trade, a country must decide on the appro- policy differences through harmonization depend on the priate level of coordination (multilateral, regional, or prospects of creating a truly integrated market, and that bilateral), the appropriate mechanism (such as interna- depends on the “natural distance� between countries and, tional rules or standards), and the appropriate approach ultimately, on factors such as levels of development, physi- (mutual recognition or harmonization) to be pursued in cal distance, legal systems, and language. individual services sectors. International rules can do little If national standards optimally serve national objec- to address impediments to trade arising from fundamen- tives, there is a trade-off between the gains from integrated tal differences among countries in regulatory standards. In markets and the costs of transition and of departing from such circumstances, two approaches can be envisaged: optimal domestic standards. For instance, a poor country harmonization and mutual recognition. These approaches may prefer to maintain a low mandatory standard for cer- are often presented as alternatives, but in fact, the former tain services because that reflects the socially optimal Services 241 trade-off between price, quality, and implementation exchanging information on experience with regulatory capacities, whereas the socially optimal trade-off in a rich reform and identifying good regulatory practices. This country may lead to a preference for and the adoption of a form of cooperation can be especially useful for regulating higher standard. Under such circumstances, harmoniza- new services in sectors characterized by continuous techni- tion of standards could create benefits in the shape of cal change. Developing countries may have a particular increased competition in integrated markets, but it would interest in cooperating with advanced industrial countries necessarily impose a social cost in at least one country. that have the longest experience with regulatory reform This matter may be nontrivial in the growing number of and in which the newest technologies and their regulatory integration agreements concluded along North-South implications are often first introduced. lines (see box 12.1). Whether or not an individual country benefits from PTAs as optimal regulatory areas for services. An optimal regulatory convergence or harmonization, its willingness regulatory area can be thought of as defining the set of to participate in this effort may hinge on where the stan- countries whose aggregate welfare would be maximized by dard is set, the level at which it is set, and the regulatory regulatory convergence. Such an area would balance the environment to which the standard responds. The latter benefits and costs of participation. The gains from eliminat- factors will in turn determine who will bear the costs of ing policy differences through harmonization depend on transition toward the adoption of the standard. The the prospects for creating truly integrated markets, which incentive to make regulations converge may depend on are conditioned by natural ties between countries and on the relative size of markets, and small countries often have factors such as geographic and linguistic proximity. The more to gain. This may explain why small countries acced- costs depend on the ex ante similarity of regulatory prefer- ing to the European Union (EU) accept that they will bear ences and the compatibility of regulatory institutions. the full cost of transition. In the definition of an optimal regulatory area, it must It should be noted that the process of regulatory con- be recognized that cooperation can be a vehicle for vergence can itself involve sunk costs of transition. The Box 12.1. WTO+ and WTO-Extra Provisions in U.S. and EU PTAs Horn, Mavroidis, and Sapir (2010) compare the substantive differences in the preferential trade agreements (PTAs) of the United States and the European Union (EU). Their work draws attention to the distinction between “WTO+� provisions, referring to PTA- induced outcomes that build on existing World Trade Organization (WTO) disciplines and commitments, and “WTO-extra� provisions, which involve disciplines or commitments that have yet not been agreed at the WTO level. The authors further highlight the issue of “legal inflation,� distinguishing those provisions that are legally binding and enforceable from those that are merely hortatory. Both EU and U.S. preferential trade agreements contain a significant number of WTO+ and WTO-extra obligations, but EU agreements go much further in their WTO-extra coverage. U.S. agreements, however, typically contain more legally binding provisions, both in WTO+ and WTO-extra areas, than do EU agreements. Horn, Mavroidis, and Sapir conclude that the two have chosen markedly different strategies for including provisions in their PTAs that go beyond the WTO agreements—in particular, where legal inflation, which is almost totally absent in U.S. agreements, is concerned. The authors are not able to draw from their analysis precise conclusions about this asymmetry of behavior, but the fact that much of the legal inflation occurs in development-related provisions, which are unique to the EU agreements, suggests that the EU has a greater need than the United States to portray its PTAs as not driven solely by commercial interests. The authors speculate that this pattern may reflect a lack of consensus among EU member states about the ultimate purpose of these PTAs and that the wide variety of provisions of weak legal value may represent a compromise among various interests in the EU. Although both EU and U.S. preferential trade agreements go significantly beyond the WTO agreements, these PTAs contain only small numbers of legally enforceable WTO-extra provisions. Very few provisions can be regarded as breaking new ground: mainly, these are environment and labor standards in U.S. agreements, and competition policy in EU agreements. All such provisions can be seen as dealing with regulatory matters. Other enforceable WTO-extra provisions found in both EU and U.S. PTAs concern domains that are closely related to existing WTO disciplines. Such disciplines are dealt with in the General Agreement on Trade in Services (GATS), for investment and capital movement matters, and in the agreements on Trade-Related Investment Measures (TRIMS) and on Trade-Related Aspects of Intellectual Property Rights (TRIPS). The fact that the new, legally enforceable WTO-extra provisions all deal with regulatory issues suggests that EU and U.S. agreements can be regarded as vehicles of “regulatory projectionism,� or means for the two trade powers to export their own regulatory approaches to their PTA partners. Developing countries’ costs and benefits in assuming such regulatory obligations require closer analytical scrutiny. Source: Horn, Mavroidis, and Sapir 2010. 242 Aaditya Mattoo and Pierre Sauvé sequence in which a country chooses to harmonize (or McMillan 1993). Although in principle this is a simple progressively align) its regulations with those its trading enough criterion, its implementation is not straightfor- partners is thus a relevant consideration. One reason is ward because the focus is on trade flows at the individual that the sequence of harmonization may influence the product level. bargaining power of different country groupings in the The liberalization of services trade implies not only negotiation concerning the level at which the harmonized that measures restricting the ability of foreign suppliers standard should be set. For example, the countries in to engage in cross-border trade are reduced or eliminated Eastern Europe that acceded to the EU individually could but also that factor mobility, including, especially, the arguably have had a greater say in the EU-wide standard in establishment of a commercial presence, is allowed. specific areas if they had been original members, had nego- In determining the welfare implications for third parties tiated collectively, or both. Similarly, harmonization first of regional integration agreements covering services, conducted at the level of the Southern Cone Common account therefore needs to be taken of the impact on both Market (Mercosur, Mercado Común del Sur) and then at trade and factor flows (capital and labor). Both flows are the level of the Free Trade Area of the Americas (FTAA) endogenous and interdependent, so that simple prescrip- or the WTO, could imply different costs and produce a tions or criteria along Kemp-Wan lines are no longer different outcome from direct harmonization at the applicable (see Baldwin, ch. 3 in this volume). higher level. If trade and factor flows are substitutes, a decline in A final consideration concerning preferential regulatory trade in products need not necessarily be detrimental to convergence is the administrative burden that may be an outside country because larger factor flows substitute implied by the maintenance and administration of distinct for trade. This is the standard case in neoclassical trade regulatory requirements and procedures by members and theory, assuming constant-returns-to-scale technology. nonmembers of a PTA. Such costs may be so acute for a With increasing returns, the relationship between factor number of developing countries as to tilt negotiating movements and product-trade flows may well be comple- incentives in favor of multilateral undertakings. It may also mentary; that is, an increase in one may be associated with encourage the multilateralization of norms first brokered an increase in the other. Although the presumption is that at the regional level or incite countries to simply extend to by liberalizing both product and factor markets, the aggre- all third countries treatment similar to that afforded to gate benefits for participants will increase and that this in PTA members, bearing in mind the limits of MFN-based turn will be beneficial for the rest of the world (partly outcomes on regulatory issues. through induced growth and investment effects), straight- forward criteria with which to evaluate such integration effects ex ante do not exist. These problems are com- Third-Country Effects pounded by the difficulty of establishing clear-cut criteria PTAs between countries that are WTO members or are for product likeness in services, given the far greater accessing the WTO can be potentially harmful to nonmem- degree of product differentiation and customer tailoring ber countries because they imply preferential liberalization arising in services markets. in favor of certain member states. Such discrimination violates one of the central obligations imposed by both The Practice of Services Liberalization in PTAs the General Agreement on Tariffs and Trade (GATT) and GATS: the MFN treatment rule. GATS is similar to the In theory, the inclusion of services trade in PTAs can help GATT in permitting signatories to pursue preferential lib- achieve greater transparency by means of rules that require eralization arrangements, subject to a number of condi- mutual openness, heightened credibility of policy through tions that are intended to minimize potential adverse legally binding commitments, and more efficient protec- effects on nonmembers and on the multilateral trading tion and regulation through rules favoring the choice of system as a whole. superior policy instruments. Relatively little, however, is In the context of agreements liberalizing trade in goods, known about the actual practice of services liberalization a sufficient condition for preferential liberalization to be in PTAs. What can be learned from experience with PTAs deemed multilaterally acceptable is that it not have detri- governing services trade? Does the bilateral or regional mental impacts on third countries. That is, the volume of route to services trade and investment liberalization actu- imports by member countries from the rest of the world ally offer significant prospects for speedier or deeper liber- should not decline on a product-by-product basis after the alization and more comprehensive rule making than does a implementation of the agreement (Kemp and Wan 1976; multilateral framework? Services 243 This section examines the manner in and the degree to of a subsequent PTA would have to be granted to the which PTAs covering services have achieved the theoretical United States. objectives enumerated above. Substantive provisions and The issue of MFN treatment in services sector PTAs negotiated outcomes under GATS are compared with the has generated much policy controversy in the context of progress made under a broad sample of PTAs featuring dis- the economic partnership agreements (EPAs) that mem- ciplines on trade in services. ber states of the EU have entered into with members of the Caribbean Forum of African, Caribbean, and Pacific (ACP) States (CARIFORUM) and plan to conclude with Key Disciplines other ACP country groupings. Fears are expressed that Although PTAs covering services come in many different such a clause, which would apply only to agreements shapes and sizes, they tend to share with GATS a common set involving partners accounting for more than 1 percent of of key disciplines governing trade and investment in serv- world trade, could reduce incentives for South-South ices, although the burdens of obligation differ (table 12.1). PTAs in services if the benefits of such integration auto- The areas of greatest rule-making convergence between matically flow (for free) to EU members. (For a fuller dis- the multilateral and PTA levels relate to scope of coverage. cussion, see Messerlin 2009.) Such a debate seems to Most commonly found in both sets of agreements, and ignore an important political-economy consideration: typically drafted in an identical manner, are disciplines on third-country MFN rights allow smaller countries to ben- transparency, national treatment, MFN treatment, pay- efit from the negotiating clout of larger partners that sign ments and transfers, and monopolies and exclusive serv- agreements with common partners (see Baldwin, Evenett, ices providers, as well as general exceptions. Considerable and Low 2009). similarities also exist between the multilateral and PTA A weaker variation of the above discussion can be found levels as regards the need for sectoral specificity (i.e., indi- in the recently concluded free trade agreement (FTA) vidual sectors or horizontal issues such as labor mobility between India and the Republic of Korea, which stipulates that require special treatment in annexes). Less conver- that if any party subsequently enters into another agree- gence (and more limited PTA progress) can be observed in ment offering more favorable treatment to a nonparty, areas of rule making that have posed recurring difficulties then that party is to give consideration to a request by the in the GATS setting. These include issues such as domestic other party for the incorporation of such treatment into regulation, emergency safeguards, and subsidy disciplines the PTA. Any such incorporation should maintain the bal- for services. ance of concessions in the overall agreement. Similarly, in Most favored nation and national treatment. The prin- the FTA between the Association of Southeast Asian ciples of most favored nation and national treatment con- Nations, Australia, and New Zealand (AANZ), if any more stitute two of the basic building blocks for any agreement favorable treatment is granted in a future trade agreement on services, just as they do in the goods area. As with by one party to a nonparty to the FTA, then the other par- GATS, very few PTAs set out such principles in unquali- ties may request consultations to discuss the possibility of fied form, regardless of whether they are framed as gen- extending no less favorable treatment. In addition, the eral obligations (which is the case for MFN in virtually all requested party is to enter into consultations with the agreements and for national treatment in agreements pur- requesting party, bearing in mind the overall balance of suing a negative-list approach to liberalization) or as obli- benefits.3 (On the related issue of MFN clauses in the con- gations that apply solely in sectors where liberalization text of investment provisions, see Miroudot, ch. 14 in this commitments are positively undertaken.2 volume.) One can easily understand that within agreements Transparency. As may be expected, given the regulatory among more than two parties, MFN is required to ensure intensity of services trade, transparency disciplines are equality of preferential treatment among all signatories. common to all PTAs covering services. These typically But why should an MFN obligation be embedded in stipulate, as is the case under GATS, an obligation to pub- PTAs concluded among bilateral partners? Part of the lish relevant measures, to notify new (or changes to exist- reason lies in the principle, first addressed in NAFTA, that ing) measures affecting trade in services, and to establish members of a given PTA automatically secure any PTA+ national enquiry points to provide on request information benefits that any one party to the original PTA may be on measures affecting services trade. One innovation willing to accord to a third party in a subsequent PTA. beyond GATS is the provision in some PTAs that members For example, any NAFTA+ commitment that Canada or afford the opportunity (to the extent possible, i.e., on a Mexico might be willing to grant to the EU in the context “best-endeavors� basis) for prior comment on proposed Table 12.1. Key Disciplines in PTAs That Cover Services 244 Market access (nondiscriminatory Rule of National quantitative Domestic Emergency Subsidy Government origin (denial Agreement MFN treatment treatment restrictions) regulation safeguards disciplines procurement of benefits) GATS (1994) Yes Yes Yes Yes Future Future negotiations Future negotiations Yes a Andean Community Yes Yes Yes Yes No No No Yes ASEAN Framework Yes Yes Yes Not specified Yes No No Yes Agreement on Services (1995) ASEAN–Australia (2009) Consideration Yes Yes Yes No No No Yes on request ANZCERTA (1988) MFN for excluded Yes Yes Yes No Export subsidies No Yes sectors prohibited Other subsidies excluded Bolivia–Mexico (1994) Yes Yes Yes Yesa Future No Separate chapter Yes CAFTA–Chile (1999) Yes Yes Yes Yesa No No Separate chapter Yes CAFTA–Dominican Yes Yes Yes Yesa Future Future negotiations Separate chapter Yes Republic (1998) CAFTA–Dominican Yes Yes Yes Yes No No Separate chapter Yes Republic–U.S. (2004) Canada–Chile (1996) Yes Yes Yes Yesa No No No Yes Canada–Colombia (2008) Yes Yes Yes Yes No No Separate chapter Yes Canada–Peru (2008) Yes Yes Yes Yes No No Separate chapter Yes CARICOM (2001) Not specified Yes Not specified Yesa Yes No No Yes CARICOM–Dominican Yes Yes Yes Yesa Future No Separate chapter Yes Republic (1998) CARIFORUM–EU (2008) Yes Yes Yes Yes No No Separate chapter Yes Central American Not specified No general No No No No No Not specified Economic Integration article Chile–Colombia (2006) Yes Yes Yes Yes No No Separate chapter Yes Chile–Mexico (1998) Yes Yes Yes Yesa No No No Yes China–ASEAN (2007) No Yes Yes Yes No No No Yes China–Singapore (2008) No Yes Yes Yes No No No Yes Costa Rica–Mexico (1994) Yes Yes Yes Yesa Future No Separate chapter Yes EEA Yes Yes No Yes No Yes (covered under No Beneficiaries competition specified disciplines) through definition of “undertakings� EFTA–Colombia (2008) Yes Yes Yes Yes No No Separate chapter Yes EFTA–GCC (2009) Yes Yes Yes Yes No No Separate chapter Yes EFTA–Mexico (2000) Yes Yes Yes Yes No No No Not specified EFTA–Singapore (2002) Yes Yes Yes Yes No Requests for Separate chapter Yes consultations to be given sympathetic consideration El Salvador–Taiwan, Yes Yes Yes Yes No No No Yes China (2007) EU (1957) Yes Yes Yes Yes No Yes (covered under Yes Yes competition disciplines) EU–Mexico (1997) Yes Yes Yes No (provisions No No Separate chapter Yes on regulatory carve-out and recognition) Group of Three (1994)b Yes Yes Yes Yesa No No Separate chapter Yes Guatemala–Taiwan, Yes Yes Yes Yes No No No Yes China (2005) India–Korea, Rep. (2009) Consideration to be Yes Yes Yes No No No Yes given to request for MFN treatment India–Singapore (2005) Consideration to be Yes Yes Yes No No No Yes given to request for MFN treatment Japan–Chile (2007) Yes Yes No (only in Yes No No Separate chapter Yes relation to financial services) Japan–Indonesia (2007) Yes Yes Yes Yes No No No Yes Japan–Malaysia (2005) Yes Yes Yes Yes Future No No Yes Japan–Philippines (2006) Yes Yes Yes Yes No No Future Yes Japan–Singapore (2002) No Yes Yes Yes No No Separate chapter Yes Japan–Switzerland (2009) Yes Yes Yes Yes No No No Yes Japan–Thailand (2007) Yes Yes Yes Yes Future No No Yes Korea, Rep.–Singapore (2005) Yes Yes Yes No No Separate chapter Yes Malaysia–Pakistan 2007) Yes Yes Yes Yes Future No No Yes Mercosur (1997) Yes Yes Yes Yes No Future negotiations Future negotiations Yes Mexico–Nicaragua (1997) Yes Yes Yes Yesa No No Future negotiations Yes Mexico–Northern Yes Yes Yes Yesa Future No No Yes Triangle (2000)c (continued next page) 245 246 Table 12.1. Key Disciplines in PTAs That Cover Services (continued) Market access (nondiscriminatory Rule of National quantitative Domestic Emergency Subsidy Government origin (denial Agreement MFN treatment treatment restrictions) regulation safeguards disciplines procurement of benefits) NAFTA (1992) Yes Yes Yes Yesa No No Separate chapter Yes New Zealand–China (2008) Yes Yes Yes Yes No No No Yes Singapore–Jordan (2005) No Yes Yes Yes No No No Yes Singapore–Panama (2006) Yes Yes Yes Yes No No Separate chapter Yes Taiwan, China–Nicaragua Yes Yes Yes Yes No No No Yes (2006) U.S.–Chile (2003) Yes Yes Yes Yesa No No Separate chapter Yes U.S.–Colombia (2006) Yes Yes Yes Yes No No Separate chapter Yes U.S.–Jordan (2000) Yes Yes Yes Yes No Future negotiations Yes Yes U.S.–Oman (2006) Yes Yes Yes Yes No No Separate chapter Yes U.S.–Panama (2007) Yes Yes Yes Yes No No Separate chapter Yes U.S.–Singapore (2003) Yes Yes Yes Yesa No No Separate chapter Yes Source: Authors’ compilation. Note: ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Agreement; ASEAN, Association of Southeast Asian Nations; CAFTA, Central America Free Trade Agreement; CARICOM, Caribbean Community; CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; EEA, European Economic Area; EFTA, European Free Trade Association; EU, European Union; GATS, General Agreement on Trade in Services; GCC, Gulf Cooperation Council; Mercosur, Southern Cone Common Market (Mercado Común del Sur); MFN, most favored nation; NAFTA, North American Free Trade Agreement; PTA, preferential trade agreement. a. Rules on domestic regulation are set out more narrowly; in most cases they apply only to the licensing and certification of professional services suppliers. b. Colombia, Mexico, and República Bolivariana de Venezuela. c. The Northern Triangle countries are El Salvador, Guatemala, and Honduras. Services 247 changes to services regulations. Increasingly, such provi- burdensome restrictions on trade or investment in serv- sions are becoming legally binding. This is notably the ices, or to disguised ones. With the exception of the EU case in North-South PTAs, following a trend initiated by itself and of agreements reached between the EU and the FTAs between the United States and Chile and between countries in Central and Eastern Europe before accession the United States and Singapore. The latter development to the EU, no PTA to date has made tangible progress in offers an interesting example of what could be described delineating the elements of a necessity test aimed at ensur- as “tactical� or “demonstration-effect� regionalism, with ing broad proportionality between regulatory means and advances at the PTA level creating precedents that their objectives, as contemplated under the GATS Article VI:4 proponents hope will facilitate their subsequent replica- mandate (see Delimatsis 2008). Neither NAFTA nor the tion at the multilateral level.4 many NAFTA-type agreements reached in the Western Market access. PTAs covering services typically address Hemisphere contain in their services chapters an article on nondiscriminatory quantitative restrictions that impede domestic regulation per se. Instead, these agreements fea- access to services markets (and which are partly dealt ture more narrowly drawn disciplines relating to the licens- with under GATS Article XVI). A number of earlier PTAs, ing and certification of professionals.5 however, particularly those concluded in the Western On most matters relating to rules governing domestic Hemisphere and modeled on North American Free Trade regulation in services trade, progress has been greater at the Agreement (NAFTA), were actually weaker than GATS, multilateral than at the PTA level. Such progress has been committing parties solely to making such restrictions fully evident in the Doha Round discussions on licensing and transparent in annexes listing nonconforming measures qualification requirements and procedures, technical stan- and to a best-endeavors approach to their progressive dis- dards, transparency, and special and differential treatment. mantling in the future. (In contrast, under GATS, WTO The disciplines on licensing procedures found in a number members undertake policy bindings in sectors, subsectors, of recent PTAs entered into by members of the Friends of and modes of supply against which market access commit- Domestic Regulation (the main demandeurs for domestic ments are scheduled.) The WTO-minus provisions of earlier regulation disciplines at the multilateral level) reflect the agreements are no longer found in the newest-generation progress made in the Doha Round negotiations. For PTAs entered into by the United States, Canada, and other instance, Article 65:3(a) of the China–Singapore FTA countries that had initially agreed to the lower standard of makes specific provision allowing applicants to remedy treatment. The newer language is similar to that found deficiencies in their applications. In addition, in cases early on in Mercosur and in the various PTAs to which EU where an application was denied or terminated, the appli- members are party; it prohibits the introduction of new cant is afforded the possibility of resubmitting a new nondiscriminatory quantitative restrictions on any sched- application at its own discretion. Similar provisions are uled commitments and sectors, mirroring the similar found in Article 10:5 of the AANZ FTA and Article 111:3 requirement in GATS Article XVI. of the New Zealand–China FTA. Although the provisions Domestic regulation. The assumption behind the argu- on licensing procedures in these PTAs do not go as far as ment, discussed above, that PTAs in the services field pro- the draft Doha Round proposals, they do offer an inter- vide scope for creating optimal regulatory areas, is that esting illustration of the iterative relationship between the aggregate adjustment costs of regulatory convergence PTAs and the WTO and of the way in which PTAs can and policy harmonization are likely to be smaller when reverse-engineer developments originating at the multilat- foreign regulatory preferences are similar and regulatory eral level. The area of domestic regulation is one in which institutions are broadly compatible. Both sets of condi- increasing numbers of PTAs embed existing GATS provi- tions are likelier, on balance, to obtain among countries sions and signal members’ desire to incorporate by refer- that are “closer� geographically, linguistically, culturally, or ence the ultimate outcome of the ongoing GATS Article VI:4 historically (Mattoo and Fink 2002). In practice, the broad negotiating mandate. In so doing, PTAs cease to assume the intersect between domestic regulation and services trade role of rule-making laboratories. This trend is most com- has tended to prove intractable (just as it has under mon with regard to the “unfinished agenda� of GATS GATS), even among the smaller subset of countries engag- (apart from government procurement, where significant ing in PTAs. PTA-induced movement has occurred in the services area). In many instances, PTAs address domestic regulation in Harmonization, mutual recognition, and regulatory coop- a manner analogous to that found in GATS Article VI— eration. With a few notable exceptions—notably, the EU that is, with a focus on procedural transparency and on and Australia–New Zealand Closer Economic Relations ensuring that regulatory activity does not lead to unduly Trade Agreement (ANZCERTA), both of which involve a 248 Aaditya Mattoo and Pierre Sauvé level of integration that extends to a common labor in which authority is delegated to licensing bodies at the market—tangible progress on matters of regulatory har- subnational level. monization or mutual recognition within PTAs has gen- Still, in comparison with the progress registered under erally proved more arduous than might be expected in GATS Article VII (recognition), reliance on the subsidiary theory (see box 12.2). approaches afforded by PTAs has led to some tangible Although a number of PTAs call on members to progress. Such advances appear more pronounced within recognize—at times, on the basis of explicit timetables— South-South PTAs such as Mercosur and the Association of foreign educational credentials and professional qualifica- Southeast Asian Nations (ASEAN) Framework Agreement tions in selected professions, progress in concluding mutual on Services (AFAS), both of which have seen the conclu- recognition agreements has often proved slow, difficult, sion of mutual recognition agreements in several regulated and partial. This is particularly noteworthy in agreements professions—for example, for nurses, engineers, account- between countries with federal political regimes and systems ants, architects, and lawyers. Box 12.2. Harmonization and Mutual Recognition in Services: Promise and Pitfalls The pessimism that often greets calls for regulatory harmonization is based on the absence of widely accepted international standards in services. Where such standards do exist, as in financial services and maritime transport, meeting them tends to be seen as a first step toward acceptability, rather than as a sufficient condition for market access. The General Agreement on Trade in Services (GATS), like the General Agreement on Tariffs and Trade (GATT), does not specifically require the use of international standards. It generally provides weaker incentives for the use of such standards than do the Sanitary and Phytosanitary (SPS) and Technical Barriers to Trade (TBT) Agreements, and it does not include a presumption of compliance; those agreements do. It is unlikely that meaningful international standards for most services will be developed soon. Still, in those areas in which global standards do exist, the likelihood of disguised or needlessly restrictive impediments to trade and investment may be significantly lessened. The presumption must also be that the existence of such standards may significantly facilitate trade and investment by promoting cross-border trade and by helping to overcome the various information asymmetries that hold back such trade and its commensurate liberalization under GATS. Accordingly, efforts should be directed toward ensuring that trade agreements create a stronger presumption in favor of genuinely international standards in services trade. As with recognition agreements, efforts to develop international standards for services trade will likely require greater doses of technical assistance and capacity building. This may be usefully done at the national and regional levels, particularly because proximity, whether geographic, historical, or cultural, may be expected to facilitate regulatory convergence. Efforts to promote the adoption of international standards will invariably be carried out outside a trade policy framework. Contrary to much popular belief, trade agreements are not in the business of making regulatory standards. Rather, they specify how such standards are to be implemented, should they affect trade. The relevant institutions for promoting international standards for services are to be found in the various specialized regulatory institutions, such as the Bank for International Settlements for banking standards, the International Telecommunications Union for telecommunications, and the International Standardization Organization (ISO) for various categories of services, including the means of producing and supplying them. A multilateral agreement such as GATS cannot mandate countries to conclude MRAs. As in the case of regional agreements, multilateral disciplines can be more or less permissive with regard to mutual recognition. The practice of MRAs suggests that their scope is often quite limited; they are invariably concluded between very similar countries. Even in a region with as strong an integrationist dynamic as Europe, and despite a significant level of prior or complementary (minimal) regulatory harmonization, the effect of MRAs has been limited by the unwillingness of many host-country regulators to cede full control. It should come as no surprise that MRAs have yet to exert significant effects on services trade. Such an outcome in turn raises the question of the benefits and costs of MRAs. Here, the analogy with regional integration agreements is again useful, as MRAs can be likened to sector-specific preferential arrangements. Where regulatory barriers are prohibitively high—one can imagine autarky as the ultimate example—recognition can only be trade creating. If they are not, selective recognition can have discriminatory effects and can lead to trade diversion. The result may well be to create trade according to a pattern of mutual trust rather than on the basis of the forces of comparative advantage. GATS Article VII (recognition) strikes a delicate balance by allowing such agreements, provided that third countries have the opportunity to accede or to demonstrate equivalence. Article VII thus has a desirable open-ended aspect that Article V (dealing with integration agreements) does not. This makes it particularly worrisome that many MRAs have been notified by WTO members under Article V rather than Article VII. In any case, the key concern for any multilateral agreement should be not how those who enjoy preferential access are treated, but how those who do not enjoy such access are treated. Ironically, the only line of defense of the rights of third countries could well come from a necessity test aimed at ensuring that such countries are not subjected to unnecessarily burdensome regulation even if they are not parties to an MRA. Because of the potential of MRAs to create trade and investment distortions, bilateral or plurilateral recognition agreements should respect the nondiscrimination principle, as mandated by Article VII of GATS. Such agreements should not, as a rule, be notified under GATS Article V but should, rather, be open to all eligible participants under the terms of Article VII. Source: Mattoo and Sauvé 2003. Services 249 Most PTAs feature provisions calling for greater insti- services example on a sectoral basis for areas in which tutional cooperation between the parties’ domestic regula- market opening may be prone to unanticipated disloca- tors in implementing agreements, typically setting up tions, with injurious consequences for smaller domestic joint regulatory commissions and periodic meetings at the firms—in distribution, for instance. The quest for a senior or ministerial level. Such cooperation, even though it generic emergency safeguard measure applicable to all proceeds from soft-law undertakings, can still yield impor- sectors and predicated on the GATT model has led to a tant trade- and investment-facilitating benefits and can negotiating stalemate at both the PTA and WTO levels. help build trust, enhance enforcement capacities, and iden- (For a fuller discussion, see Sauvé 2002; Marconini 2005; tify postnegotiation implementation bottlenecks. All of Pierola 2008.) these may be key ingredients—indeed, preconditions—for On the issue of disciplines for service-related subsidies, regulatory harmonization and effective mutual recognition with the exceptions of the EU (including its preaccession initiatives. agreements with countries in Central and Eastern Europe) Emergency safeguard mechanisms, subsidy disciplines, and of ANZCERTA, the quest for rule-making advances and government procurement. With few exceptions, PTAs has proved just as elusive at the PTA as at the WTO level. have made little headway in tackling the key “unfinished� This is particularly so within countries with federal systems rule-making items on the GATS agenda. This is most of governance. notably the case for disciplines on an emergency safeguard Whereas a number of PTAs (e.g., Mercosur) replicate mechanism (ESM) and subsidy disciplines for services the exhortation in GATS to develop future disciplines on trade, where governments confront the same conceptual subsidies in services trade, most PTAs that cover services challenges, data limitations, feasibility challenges, and specifically exclude subsidy practices. Paralleling provi- political sensitivities at the regional level as they do on the sions found in GATS, the FTA between the European Free multilateral front. For example, the countries of Southeast Trade Association (EFTA) and Singapore requires that Asia, which until recently were among the most vocal sympathetic consideration be given to requests by a party proponents of an ESM in GATS, have yet to adopt such a for consultations in instances where subsidy practices provision within AFAS. To date, in the Western Hemi- affecting trade in services may be deemed to have injurious sphere, only members of the Caribbean Community effects. The area of subsidy disciplines, like that of domestic (CARICOM) have adopted (but have not yet used) such regulation, is one in which many PTAs signal a desire to an instrument, and questions remain as to the necessity incorporate by reference the outcome of any agreed out- and the operational feasibility of an ESM in services come from ongoing (but largely stagnant) discussions at trade, given the flexibilities embedded in the very conduct the WTO level. (For a fuller discussion, see Sauvé 2002; of market opening under most agreements. Adlung 2007; Poretti 2008.) NAFTA provides one example of sector-specific More progress has been made at the PTA level in open- experimentation with safeguard-type measures. Under ing up government procurement markets in services. This the terms of the agreement’s chapter on financial serv- has typically been achieved through negotiations in the ices, Mexico was allowed to impose market-share caps if area of government procurement per se—as with the the specific foreign ownership thresholds agreed to WTO’s Government Procurement Agreement (GPA)— (25 percent for banks and 30 percent for securities firms) rather than through services negotiations.6 were reached before 2004. Mexico could have recourse to The approach taken in PTAs is for the most part very such market-share limitations only once during the similar to that adopted in the WTO; that is, nondiscrimina- transition period and could only impose them for a tion among members within the scope of scheduled three-year period. Under no circumstances could such commitments, and procedures to enhance transparency measures be maintained beyond the end of the transition and due process. PTAs whose members are all parties to the period foreseen for market opening under NAFTA (e.g., GPA, such as EFTA and the Singapore–Japan FTA, specifi- 2001). Mexico, in fact, never made use of such provisions cally mention that the relevant GPA articles apply, and even as the aggregate share of foreign participation in its most agreements concluded in the Western Hemisphere financial system became significantly higher than the basically replicate GPA disciplines at the regional level. In specified thresholds (Sauvé and Gonzalez-Hermosillo contrast to the GPA, which applies in principle to pur- 1993; Sauvé 2002). It is interesting, if somewhat surpris- chases by both central and subnational governments, most ing, that no further attempt has been made, either in PTAs provide for binding government procurement disci- PTAs or at the WTO level, to consider the scope for, and plines at the national level only (OECD 2002c; see also the practical means of, replicating the Mexican financial Dawar and Evenett, ch. 17 in this volume). 250 Aaditya Mattoo and Pierre Sauvé Investment. A policy domain in which PTAs have in other countries, with a view to extending or denying the achieved considerable progress, while forward move- benefits foreseen under such agreements. ment has stalled in the WTO, is that of investment rule Experience shows that rules of origin for services and making. Most PTAs feature comprehensive disciplines on investment can play a significant role in determining the the protection and liberalization of cross-border invest- degree to which regional trading arrangements discrimi- ment activity. There is scope in this area for investor-state nate against nonmember countries, and hence the extent of arbitration alongside WTO-like state-to-state dispute set- potentially costly trade and investment diversion. When tlement, together with extensive liberalization commit- levels of protection differ among participating countries, ments, most often brokered on a negative-list basis. Given the effective preference granted to a trading partner will the central role assumed by investment as the most impor- depend on the restrictiveness of the applied rule of origin. tant mode of service supply, such developments are of In the extreme, if one participant has a fully liberalized some significance for the operation of services markets market, the adoption of a liberal rule of origin by the other and for the promotion of more contestable entry condi- participants can be likened to MFN liberalization because tions in them. (For a fuller discussion of the evolution of services and services suppliers can enter or establish international rules on investment, see Beviglia Zampetti themselves in the liberal jurisdiction and from there move and Sauvé 2007.) The extent to which PTAs featuring to, or service, the partner countries. comprehensive investment norms might influence the From an efficiency perspective, origin rules for services WTO’s evolving architecture of rules will very much should allow for third-country services suppliers, particu- depend on prospects for crafting a multilateral regime larly those operating through a commercial presence (via for investment. Any such agreement at the WTO level mode 3 entry) to take advantage of, and contribute to, the would likely raise a number of intractable questions benefits of an integrating area. Under a liberal rule of ori- regarding the scope of GATS—notably, its coverage of gin for services and investment aimed at ensuring that commercial presence as a mode of supplying services. established foreign operators are not mere shell companies Starting with NAFTA in 1994, a large and growing num- but conduct substantial business operations in the host- ber of PTAs has shown how the treatment of investment country market, third-country investors and service in services need not be distinguished from that in other providers can take full advantage of the expanded market sectors subject to trade disciplines. The issue of preferen- opportunities afforded by the creation of a PTA by estab- tial advances in investment rule making is taken up by lishing a commercial presence within the integration area.7 Miroudot, ch. 14 in this volume. Not surprisingly, participants that seek to benefit from Rules of origin and denial of benefits. A final area of rule preferential access to a protected market and deny benefits making concerns rules of origin, which determine who to third-country competitors are likely to argue for the ultimately qualifies for preferential treatment under PTAs. adoption of restrictive rules of origin. This may be the atti- In services agreements, this matter is generally addressed tude, in particular, of regionally dominant but nonglobally under provisions dealing with denial of benefits (see competitive service providers toward third-country com- Beviglia Zampetti and Sauvé 2007). petition within a regionally integrating area. The literature on rules of origin has focused almost The adoption of restrictive rules of origin is permissible exclusively on merchandise trade flows and hence on poli- under GATS Article V:3, which allows PTAs concluded cies for determining the origin or nationality of tangible between developing countries (South-South PTAs) to products. Much less attention has been paid to the increas- restrict the benefits of integration to service suppliers that ingly important issue of how to determine the origin of are owned and controlled by citizens of the integrating producers, which is the primary concern of the study of area. It is not clear whether such flexibility serves the rules of origin in services trade and investment. The ability development interests of those that use the rules. Several to contest services markets often requires the physical pres- South-South PTAs, notably ASEAN, Mercosur, the Andean ence of suppliers in the territory of consumers, either in the Pact, and the FTAs between China and Hong Kong SAR, form of individual services providers performing cross- China, and between China and Macao SAR, China, have border transactions on a temporary (contract) basis, or as opted for a restrictive policy stance in this area. entities servicing a foreign market on the basis of a com- The policy stance on rules of origin for services and mercial presence in that market. Governments that are investment in a PTA can play an important role in pro- signatories of trade and investment agreements may, moting or inhibiting access to the most efficient suppliers accordingly, need to ascertain whether suppliers originate of services. In many services sectors, the most efficient (or Services 251 most globally competitive) suppliers tend to be either are contemplating the scope that may exist in the current developed-country firms or firms originating outside an negotiations for making possible improvements to the integrating area. Accordingly, the adoption of rules of ori- GATS architecture. gin that restrict benefits to nationals of member states can Under a GATS-like positive or hybrid approach to have detrimental effects by potentially locking integrating scheduling liberalization commitments, countries agree partners into suboptimal patterns of production and con- to undertake national treatment and market access com- sumption. This problem may be compounded, and may mitments specifying (through reservations in scheduled generate longer-term deadweight losses, because many areas) the nature of treatment or access offered to foreign services, particularly network-based services, involve sig- services or foreign service suppliers.9 Countries retain nificant location-specific sunk costs, such that first movers the full right to undertake no commitments. In such (even if relatively inefficient) can exert long-term domi- instances, they are under no legal obligation to supply nance and extract monopolistic rents. Then, a country information to their trading partners on the nature of dis- may be stuck with inferior suppliers for a long time even if criminatory or access-impeding regulations maintained at it subsequently liberalizes on an MFN basis. Indeed, the domestic level. because of the importance of sunk costs in many services A related feature of GATS that tends to be replicated in industries, sequential entry (which preferential liberaliza- PTAs that espouse a bottom-up or hybrid approach to lib- tion with restrictive rules of origin can easily promote) eralization is that countries have the possibility of making can produce very different results from simultaneous commitments that do not reflect (i.e., are made below) the entry. If entry is costly, an incumbent may indeed be able regulatory status quo—a long-standing practice in tariff to deter entry, leading to greater market concentration negotiations that was replicated in a GATS setting. and a reduction in consumer welfare. (For a fuller discus- The alternative, top-down approach to liberalization of sion, see Mattoo and Fink 2002.) Some measure of solace services trade and investment is based on the concept of can be taken from the observation, confirmed in this negative listing, whereby all sectors and nonconforming chapter’s sample of reviewed PTAs, that to date, most pref- measures are to be liberalized unless otherwise specified in erential agreements have adopted the most liberal rule of a transparent manner in reservation lists appended to an origin for mode 3 suppliers, whereby any juridical person agreement. Nonconforming measures contained in reser- incorporated in any party to an integration agreement and vation lists are then usually liberalized through consulta- conducting substantial business operations there receives tions or, as in GATS, periodic negotiations. full treaty benefits. It is interesting to note that despite the strong opposi- tion that such an approach generated when first mooted by a few GATT contracting parties during the Uruguay Modalities of Liberalization: Negative-List versus Round, the negative-list approach to services liberalization Positive-List Approaches has recently been adopted by a majority of PTAs covering Two major approaches toward the liberalization of trade services that have been notified to the WTO. In the sample and investment in services have been manifest in PTAs and of 55 PTAs under review in this chapter, 33 (60 percent of in the WTO: (a) the positive-list, or “bottom-up,� approach the total) follow a negative-list approach. (typically, a hybrid approach featuring a voluntary, positive First used (for trade in services only, in the absence of choice of sectors, subsectors, and modes of supply in which an investment chapter) by Australia and New Zealand in governments are willing to make binding commitments, ANZCERTA, the approach was further developed by together with a negative list of nonconforming measures to Canada, Mexico, and the United States in NAFTA in be retained in scheduled areas), and (b) the negative-list, or 1994. Mexico has played a pivotal role in extending this “top-down/list it or lose it,� approach. Although both liberalization approach and similar types of services disci- negotiating modalities can be made to produce broadly plines (i.e., right of nonestablishment) to other PTAs it equivalent liberalization outcomes, the two approaches has signed with countries in South and Central America.10 may generate qualitative differences that may be significant The pattern has been replicated in PTAs signed between from both a domestic and an international governance Central and South American countries, on the one hand, point of view.8 Even though the debate over these compet- and Asian economies, on the other hand, particularly in ing approaches appears settled at the level of GATS, it is PTAs involving the region’s most advanced partners— useful to recall these differences because the issue is still among them, Australia, Japan, Korea, New Zealand, very much alive in the PTA context, and WTO members Singapore, and Taiwan, China. 252 Aaditya Mattoo and Pierre Sauvé A number of distinguishing features of negative listing A further liberalizing feature found in a number of can be identified. First, such an approach enshrines and PTAs using a negative-list approach to liberalization con- affirms the up-front commitment of signatories to an sists of a ratchet mechanism (table 12.2), whereby any overarching set of general obligations (subject to reserva- autonomous liberalization measure undertaken by a PTA tions). This is currently the case under GATS, primarily member between periodic negotiating rounds is automati- with respect to the agreement’s provisions on MFN treat- cally reflected in that member’s schedule of commitments ment (Article II, with scope for one-time exceptions) or lists of reservations. A provision of this kind typically and transparency (Article III). Most other disciplines aims at preventing countries from backsliding with respect apply in an à la carte manner to sectors and modes of to autonomously decreed policy changes. It may also pro- supply on the terms inscribed in members’ schedules of vide negotiating credit for autonomous liberalization by commitments.11 developing country members, an issue currently under dis- A second, and perhaps more immediately operational, cussion in the GATS context. defining characteristic of negative listing lies in its ability to Such ratchet provisions are found in many South-South generate a standstill—that is, to establish a stronger floor of PTAs covering services in the Western Hemisphere. For liberalization by locking in the statutory or regulatory sta- instance, Article 10 of the Andean Community’s Decision tus quo. Such an approach avoids the GATS pitfall of allow- 439, on services, applies to all new measures adopted by ing a wedge to arise between applied and bound regulatory member countries that affect trade in services and does not or statutory practices.12 An important caveat concerns the allow for the establishment of new measures that would propensity of negative-list agreements to allow parties to increase the degree of nonconformity or fail to comply lodge reservations that preserve future regulatory freedom with the commitments contained in Article 6 (market in a manner analogous to unbound or nonscheduled com- access) and Article 8 (national treatment) of the decision. mitments under GATS. Here again, and unlike GATS, Article 36 of the CARICOM Protocol, another status quo which yields no information on the nature of noncon- or standstill provision, prohibits members from introduc- forming measures retained in what are typically sensitive ing any new restrictions on the provision of services in the sectors, negative-list agreements oblige signatories to reveal community by CARICOM nationals. the nature of existing nonconforming measures in such A provision of this type can exert positive effects on the reserved sectors. investment climate of host countries by signaling to foreign The main governance-enhancing feature arising from suppliers the host countries’ commitment not to reverse the adoption of a negative-list approach is thus the greater liberalization (see Hoekman and Sauvé 1994; Stephenson level of transparency it can generate if adhered to prop- 2001c). Such credibility-enhancing provisions may be erly.13 The information contained in reservation lists will especially important for smaller countries that often find it be important to prospective traders and investors, who difficult to attract larger doses of foreign direct investment. value the one-stop shopping attributes of a comprehensive A recent comprehensive review of East Asian PTA com- inventory of potential restrictions in foreign markets. mitments in services suggests that some qualification is Such an inventory is also likely to benefit home-country required to the common belief that negative listing inher- negotiators, assisting them in establishing a hierarchy of ently yields greater transparency (see Fink and Molinuevo impediments to tackle in future negotiations. This infor- 2007). Some agreements that use negative listing provide a mation can in turn lend itself more easily to formula-based clearer road map of existing regulatory impediments, but liberalization—for instance, by encouraging members others fall short of the expected transparency because, as to agree to reduce or progressively phase out “revealed� noted above, they use sweeping sectoral or mode-specific nonconforming measures that may be similar across coun- carve-outs or exclude entire categories of measures, such as tries, such as quantitative limitations on foreign ownership subnational measures.15 in selected sectors (see Sauvé 1996). Evidence on the impact of negative listing on induced The production of a negative list may also help generate levels of liberalization is also mixed. Some parties have a useful domestic policy dialogue between the trade- concluded hybrid-list agreements that achieve greater lib- negotiating and regulatory communities, thereby encour- eralization than their negative-list agreements with other aging countries to perform a comprehensive audit of existing partners. For example, Singapore’s positively listed com- trade- and investment-restrictive measures, benchmark mitments in its PTA with Japan provide significantly domestic regulatory regimes against best international greater coverage than its negatively listed commitments in practices, and revisit the rationale for, and the most effi- its PTA with Australia. There is little doubt, however, that cient means of, satisfying domestic policy objectives.14 when done properly, negative-list agreements may yield Services 253 Table 12.2. Negotiating Approaches in Services Trade Main features GATS-like (hybrid approach) Negative-list approach General description Schedule of commitments positively lists sectors, Free trade and investment in services are subsectors, and modes of supply in which assumed unless specific existing measures are commitments on market access, national inscribed in reservation lists indicating the treatment, and any additional commitments sector, subsector, industrial classification, are undertaken and negatively lists any nature of the treaty provision that is violated, nonconforming treatment or measures description of the measure in question, and retained therein. nature of the measure’s nonconformity with regard to specific treaty provisions. A la carte liberalization: members retain the right List or lose: all nonconforming measures not to choose sectors, subsectors, and modes of notified at the moment of a PTA’s entry into supply in which they are prepared to undertake force are automatically bound as “free� legally binding market access, national (signatories lose the right to invoke treatment, and any other additional nonconforming measures that are not commitments. inscribed in their lists of reservations on an agreement’s entry into force). Locking in of regulatory Not guaranteed unless otherwise specified. Generally guaranteed. status quo Members are typically allowed to schedule commitments below the regulatory status quo, regardless of the level of market openness flowing from existing domestic regulatory measures. Transparency Generally more limited, as signatories retain the Generally greater, given focus on reserving flexibility not to schedule commitments or to existing nonconforming measures, but some schedule commitments below the regulatory agreements allow signatories to lodge status quo. sweeping reservations (e.g., all nonconforming measures that exist at the subnational level). Scope for introducing future Can be secured either by omitting a sector, Specific annexes allow signatories to a negative nonconforming subsector, or mode of supply from a list PTA to retain future policy flexibility in measures member’s schedule of commitments or by sectors, subsectors, and modes of supply. scheduling an unbound commitment. These become the GATS equivalent of No information on the nature of unbound measures. Parties are normally nonconforming measures is generated in required to describe the current level of nonscheduled or unbound sectors, nonconformity prevailing in such reserved subsectors, or modes of supply. areas. Ratchet mechanism None. Many negative-list PTAs feature a ratchet provision aimed at ensuring that any autonomous measure of a liberalizing nature enacted after a PTA’s entry into force or, where envisaged, between periodic negotiating rounds, becomes the liberalizing party’s commitment under the PTA, with market-opening benefits automatically extended to PTA partners on an MFN basis in the case of plurilateral PTAs. Source: Authors’ compilation. Note: GATS, General Agreement on Trade in Services; MFN, most favored nation; PTA, preferential trade agreement. important benefits in regulatory transparency and locking- must also be weighed against some of the benefits in in of the regulatory status quo. governance and best regulatory practices described Two potential pitfalls arising from the use of negative above. A second concern is that the adoption of a nega- listing can be identified. First, such an approach may be tive list implies that governments ultimately forgo the administratively burdensome, particularly for developing right to introduce discriminatory or access-impairing countries. The burden may be mitigated by allowing for measures in the future—even in sectors that do not yet progressivity in the completion of members’ negative lists exist or are not regulated at the time of an agreement’s of nonconforming measures.16 The costs of compliance entry into force. 254 Aaditya Mattoo and Pierre Sauvé To assuage the latter concerns while promoting the opening advances achieved in services-related PTAs across transparency-enhancing properties associated with the use the sample of agreements under review, but it does draw on of negative listing, a suggestion has been made that coun- a number of important recent contributions to the litera- tries be encouraged (possibly in the WTO context) to ture to offer stylized facts about the WTO+ nature of PTAs exchange comprehensive and nonbinding lists of noncon- in services trade. forming measures—something that has been done in the Andean Community and is being considered for Mercosur. Sequencing (See Sauvé and Wilkie 2000 for a fuller description of such a proposal.) Preferential market opening in services rarely, if ever, pre- In an important new development along the above dates preferential talks in goods trade. Countries that lines, exemplified in the most recent Japanese PTAs, nego- engage in services-related PTAs either conduct such nego- tiators have sought to combine the best properties of nega- tiations alongside merchandise trade negotiations in a tive and hybrid listing. Recent Japanese PTAs maintain a manner analogous to the WTO’s single undertaking GATS-like hybrid approach to scheduling, preserving the approach or pursue services talks sequentially once a PTA right of countries to pick and choose those sectors, subsec- in goods trade has been agreed.17 The latter approach is tors, and modes of supply in which they desire to make more common among South-South PTAs, whereas agree- commitments. The country’s PTAs balance this flexibility ments involving members of the Organisation for Eco- with the twin obligations of scheduling the regulatory sta- nomic Co-operation and Development (OECD) typically tus quo (to prevent members from offering less access than espouse the single undertaking route. currently exists) and exchanging nonbinding lists of non- Countries preferring sequential liberalization may wish conforming measures, to promote greater regulatory trans- to test the waters in goods trade first, raising comfort levels parency. The economic partnership agreement between the with new trading partners. They may also wish to limit the EU and CARIFORUM takes a similar approach by allowing scope for bargaining across several policy areas that takes parties to schedule status quo commitments on a GATS place in WTO-type negotiations, although that choice may basis. (For a fuller discussion, see Sauvé 2009; Sauvé and well constrain negotiations, given the narrowness of the Ward 2009b; Marconini and Sauvé 2010.) resulting negotiating remit. Sequential liberalization may PTAs have become more flexible, and important varia- also allow partners to identify key services inputs and tions are being introduced, depending on the negotiating address potential services sector bottlenecks holding back partners. For instance, Japan’s PTAs with Indonesia, trade in manufactured products or primary commodities. Malaysia, the Philippines, and Thailand have been con- The greater degree of liberalization achieved to date under ducted along positive-list lines, while those with Chile and North-South PTAs, particularly those based on a negative- Switzerland take a negative-list approach. PTAs increas- list approach, helps explain why PTAs predicated on a sin- ingly mix positive- and negative-list approaches within the gle undertaking approach have tended to attain greater same agreement. Recourse to negative listing is particu- levels of market opening. larly pronounced in the investment area. Some agree- The fact that PTAs have achieved significant progress in ments use both approaches, depending on the sector or market-opening terms when compared with GATS should mode of supply (e.g., positive listing for cross-border come as no surprise when one considers that the WTO trade and negative listing for commercial presence, or neg- commitments under GATS were brokered in the early ative listing for banking services and positive listing for 1990s—in the case of telecommunications and financial insurance services). services, where the most progress was made, in 1997. By far the greatest number of services-related PTAs postdates the entry into force of GATS. Such agreements have taken Assessing the Depth of Preferential advantage of the rising comfort levels afforded by the ped- Liberalization in Services Trade agogical journey undertaken during the Uruguay Round, The depth of services liberalization varies considerably the increased services-specific trade-related technical assis- across PTAs, with notable differences across sectors, modes tance dispensed at both the multilateral and PTA levels of supply, and approaches to scheduling commitments in the post–Uruguay Round period, and the far-reaching (i.e., hybrid versus negative-list approaches), as well degree of unilateral liberalization in services markets that as across country groups (North-North, North-South, and characterized the period and which the WTO-GATS has South-South) and partner pairings. This chapter does not yet been able to catch up with and consolidate. A more not attempt a comprehensive assessment of the market- analytically meaningful comparison would thus be one Table 12.3. Key Features of PTAs That Cover Services Right of nonestab- Ratchet Agreement Scope or coveragea Negotiating approach Treatment of investment in services lishment mechanism GATS Universal Positive list Covered as “commercial presence� (mode 3) No No Andean Community Universal Negative list Covered as “commercial presence� No No ANZCERTA Universal Negative list Covered as “commercial presence� but no common Yes No disciplines on investment ASEAN–Australia–New Zealand Universal Positive list Covered as “commercial presence� and under investment No No ASEAN Framework Agreement Universal Positive list Covered as “commercial presence� and under a separate No No on Services investment chapter Bolivia–Mexico Universal Negative list Separate chapter Yes Yes CACM Construction services Positive list Not specified No No CAFTA–Chile Universal Negative list Separate chapter Yes Yes CAFTA–Dominican Republic Universal Negative list Separate chapter Yes Yes CAFTA–Dominican Republic–U.S. Universal Negative list Separate chapter Yes Yes Canada–Chile Universal Negative list Separate chapter Yes Yes Canada–Colombia Universal Negative list Separate chapter Yes Yes Canada–Peru Universal Negative list Separate chapter Yes Yes CARICOM Universal Negative list Covered as “commercial presence� and in separate chapters No No (on right of establishment and movement of capital) CARICOM–Dominican Republic Universal Negative list Separate chapter Yes No CARIFORUM–EU Universal (audiovisual Positive list Covered as “commercial presence� No No services explicitly excluded)a Chile–Colombia Universal Negative list Separate chapter Yes Yes Chile–Mexico Universal Negative list Separate chapter Yes Yes China–ASEAN Universal Positive list Covered as “commercial presence� No No China–Singapore Universal Positive list Covered as “commercial presence� No No Costa Rica–Mexico Universal Negative list Separate chapter Yes Yes EEA Universal Negative list Separate chapter Yes No EFTA–Colombia Universal Positive list Covered as “commercial presence� No No EFTA–GCC Universal Positive list Covered as “commercial presence� No No EFTA–Mexico Universal Positive list Covered as “commercial presence� and under a separate No No investment chapter EFTA–Singapore Universal Positive list Covered as “commercial presence� and under a separate No No investment chapter El Salvador–Taiwan, China Universal Negative list Separate chapter Yes Yes EU Universal Negative list Treated as freedom to establish Yes No EU–Mexico Universal (audiovisual Standstill (+ future Covered as “commercial presence� and under a separate No No services explicitly negotiation of investment chapter excluded) commitments, as in GATS) (continued next page) 255 Table 12.3. Key Features of PTAs That Cover Services (continued) 256 Right of nonestab- Ratchet Agreement Scope or coveragea Negotiating approach Treatment of investment in services lishment mechanism Group of Threeb Universal Negative list Separate chapter Yes Yes Guatemala–Taiwan, China Universal Negative list Separate chapter Yes Yes India–Korea, Rep. Universal Positive list Covered as “commercial presence� and under a separate No No investment chapter India–Singapore Universal Positive list Covered as “commercial presence� No No Japan–Chile Universal Negative list Separate chapter Yes Yes Japan–Indonesia Universal Positive list Covered as “commercial presence� and under a separate No No investment chapter Japan–Malaysia Universal Positive list Covered as “commercial presence� and under a separate No No investment chapter Japan–Philippines Universal Positive list Covered as “commercial presence� and under a separate No Yes investment chapter Japan–Singapore Universal Positive list Covered as “commercial presence� and under a separate No No investment chapter Japan–Switzerland Universal Negative list Covered as “commercial presence� and under a separate No Yes investment chapter Japan–Thailand Universal Positive list Covered as “commercial presence� No No Korea, Rep.–Singapore Universal Negative list Separate chapter Yes Yes Malaysia–Pakistan Universal Positive list Covered as “commercial presence� and under a separate No No investment chapter Mercosur Universal Positive list Separate protocols No No Mexico–Nicaragua Universal Negative list Separate chapter Yes Yes Mexico–Northern Trianglec Universal Negative list Separate chapter Yes Yes NAFTA Universal Negative list Separate chapter Yes Yes New Zealand–China Universal Positive list Covered as “commercial presence� No No Singapore–Jordan Universal Positive list Covered as “commercial presence� No No Singapore–Panama Universal Negative list Separate chapter Yes Yes Taiwan, China–Nicaragua Universal Negative list Separate chapter Yes Yes U.S.–Chile Universal Negative list Separate chapter Yes Yes U.S.–Colombia Universal Negative list Separate chapter Yes Yes U.S.–Jordan Universal Positive list Covered as “commercial presence� No No U.S.–Oman Universal Negative list Separate chapter Yes Yes U.S.–Panama Universal Negative list Separate chapter Yes Yes U.S.–Singapore Universal Negative list Separate chapter Yes Yes Source: Authors’ compilation. Note: ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Agreement; ASEAN, Association of Southeast Asian Nations; CACM, Central American Common Market; CAFTA, Central America Free Trade Agreement; CARICOM, Caribbean Community; CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; EFTA, European Free Trade Association; EU, European Union; GATS, General Agreement on Trade in Services; GCC, Gulf Cooperation Council; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement. a. Except for CACM (construction services), excludes air transport, and in certain cases, maritime cabotage (transport of goods or passengers between two points in the same country by a vessel registered in another country). b. Colombia, Mexico, and República Bolivariana de Venezuela. c. The Northern Triangle countries are El Salvador, Guatemala, and Honduras. Services 257 between the level and nature of PTA commitments and of that achieved by the latest DDA offers. For mode 3, it negotiating offers made by WTO members in the ongoing reaches 85 percent, almost double the average proportion Doha Development Agenda (DDA). of sectors covered by existing GATS commitments and significantly higher than what has been offered in the DDA to date. Comparison with the WTO The study by Roy, Marchetti, and Lim (2008) of the Today’s PTAs may have taken services liberalization fur- WTO+ nature of PTA commitments regarding mode 3 ther than the situation prevailing at the end of the last (commercial presence)—by far the most important round of multilateral negotiations, but there are marked means of accessing services markets—reveals significant differences in outcomes that bear noting, Using a sample variance across country groupings (table 12.4). Despite of 28 PTAs concluded among 29 WTO members, Roy, the considerable diversity in additional sectoral coverage Marchetti, and Lim (2008) identify differences in the for individual countries, PTAs are found to go beyond level of commitments scheduled under GATS for modes 1 existing GATS commitments and DDA offers across all (cross-border supply) and 3 (commercial presence) and country groupings. The PTA-induced jump in sectoral those flowing from GATS offers made, as of January 2008, coverage for developing countries is much larger than for under the Doha Development Agenda or included in PTAs. developed countries, which had higher sectoral coverage In computing sector coverage, the authors look at the best levels to start with under their GATS schedules. commitments undertaken by each country in any of its Table 12.4 further highlights sizable differences in agreements. The results are summarized in figure 12.3. mode 3 liberalization patterns between agreements pursuing Figure 12.3 shows that the average level of subsectors hybrid and negative-list approaches to liberalization, with bound in the prevailing GATS schedules is rather low far greater commitments scheduled under the former. It (31 percent for mode 1 and 44 percent for mode 3), reflect- also shows that PTAs conducted along North-South lines ing the caution that characterized the first-ever multilateral achieve deeper liberalization than those involving South- negotiation in the services field. The figure further reveals South partnerships. Such a result is broadly commensurate that DDA offers have not modified this landscape in a with the continued dominance of OECD countries in dramatic fashion; mode 1 increases 7 percentage points, to world services trade and investment, even as a growing 38 percent, and mode 3 gains 9 percentage points, to number of developing countries are fast acquiring signifi- 53 percent. The value added of PTAs is significant. For cant levels of comparative advantage across a wide range of mode 1, coverage is 73 percent, on average—almost double sectors. The findings may also confirm the superior negoti- ating leverage that large countries have in preferential con- fines relative to what is possible at the WTO level. Figure 12.3. Sectoral Coverage of PTAs and of Several PTAs, particularly (but not only) those negoti- GATS Offers and Schedules, Selected Countries ated along South-South lines, show a tendency to deepen 100 liberalization commitments only marginally beyond the GATS. This raises serious questions about the agreements’ 90 very rationale, all the more so when signatories resort to committed by countries reviewed average percentage of subsectors 80 the GATS framework for rule-making purposes without 70 attempting to craft new or PTA-specific rules to govern 60 services trade and investment among themselves.18 50 A factor that clearly influences the level of commit- ments undertaken in services-related PTAs is the economic 40 importance of the trading partners involved. Marchetti and 30 Roy (2008) and Roy, Marchetti, and Lim (2008) show how 20 the United States invariably secures from its trading part- 10 ners greater commitments than those the same countries 0 are willing to undertake in PTAs with other trading part- mode 1 mode 3 ners, including OECD countries. Marchetti and Roy with PTA with GATS offer GATS (2008) argue that such a finding can be traced to a mix of political influence and foreign policy factors and to the Source: Roy, Marchetti, and Lim 2008. Note: GATS, General Agreement on Trade in Services; PTA, preferential relative importance of the U.S. market for its trading part- trade agreement. ners’ key goods exports, as in the cases of the Central 258 Aaditya Mattoo and Pierre Sauvé Table 12.4. Average Percentage of Subsectors Subject to Market Access Commitments on Mode 3, Selected Country Groupings (percent) Difference from best Item GATS With GATS DDA offer With PTA WTO treatment All 44 53 85 32 Developing countries 36 46 86 40 Developed countries 67 74 82 8 Hybrid listing 57 66 69 3 Negative listing 37 47 93 46 U.S. PTA partner 30 39 93 54 Non-U.S. PTA partner 56 66 76 10 Source: Adapted from Roy, Marchetti, and Lim 2008. Note: DDA, Doha Development Agenda; GATS, General Agreement on Trade in Services; PTA, preferential trade agreement; WTO, World Trade Organization. American countries, the Dominican Republic, and the Table 12.5. GATS Commitments, GATS DDA Offers, and Andean countries. “Best� PTA Commitments for All Members Reviewed, Selected Sectors Sectoral and modal patterns. Turning to sectoral patterns (scale: 1–100) of liberalization, the available empirical evidence attests to Sector GATS GATS DDA offers PTAs significant WTO+ advances across the full range of traded services. Using an index that ranks scheduled commit- Professional 30 39 67 ments on a scale from 1 to 100, Marchetti and Roy (2008) Computer 55 74 93 Postal and courier 14 20 53 show how PTAs have registered far-reaching advances in Telecommunications 51 58 80 comparison with GATS commitments and DDA offers Audiovisual 17 20 50 across all sectors (see table 12.5). These include sectors that Construction 40 46 75 have attracted fewer commitments and DDA offers under Distribution 32 41 76 GATS, such as audiovisual services, as well as those that Education 18 25 57 have generally proved more attractive in a multilateral set- Environment 20 30 62 ting, including computer, tourism, and telecommunica- Financial 36 40 53 tions services. Health 8 11 34 Tourism 51 61 83 With the notable exception of land transport issues, Maritime 12 23 57 where physical proximity stands out as a determinative Rail 14 20 52 trade-facilitating feature driving cross-border liberalization, Road 16 18 56 PTAs continue to encounter resistance in opening up those Auxiliary transport 21 24 58 services sectors that have proved difficult to address at the Source: Marchetti and Roy 2008. multilateral level. Thus, most PTAs tend to exclude from Note: DDA, Doha Development Agenda; GATS, General Agreement on their coverage the bulk of air transport services (with the Trade in Services; PTA, preferential trade agreement. Scores for modes 1 and 3 are combined. PTA data are based on the best (most liberalizing) notable exception of the EU for intra-EU traffic), as well as a PTA commitments. broad swath of public services. Relatively limited progress—although still WTO+ in most areas—has similarly been achieved in PTAs in sec- characterized by rapid technological and commercial tors in which particular policy sensitivities arise, such as change. A prominent example is e-commerce (digital maritime transport, audiovisual services, energy, and, to trade), which encompasses a broad range of business and some extent (though more so for some countries than information technology (IT)–related services and which others), the movement of services suppliers. Other sectors was not yet a commercial reality at the time of the Uruguay that generally fit this overall pattern include postal and Round (see box 12.3). courier services (but not express delivery) and health and Market-opening advances are also notable in sectors in education services. which new post–Uruguay Round proliberalization con- By contrast, PTAs have proved useful settings in which stituencies have emerged that seek to use trade agreements to advance liberalization prospects in market segments to secure expanded opportunities in world markets. This Services 259 Box 12.3. PTAs and Digital Trade Increasingly, preferential trade agreements (PTAs) acknowledge the growing cross-border electronic delivery of services and digital products (e.g., software) by incorporating trade rules for e-commerce. A nonbinding e-commerce chapter in the U.S.–Jordan PTA of 2000 was followed by the first legally binding U.S. e-commerce chapter in bilateral trade agreements—in the 2003 free trade agreement (FTA) between the United States and Singapore—and by a further flurry of U.S.-led bilateral PTAs incorporating e-commerce chapters subject to dispute settlement provisions. The trend has spread, and PTAs between Singapore and Australia, Thailand and Australia, India and Singapore, and other parties contain digital trade rules. Other PTAs—the Maghreb Arab Union and agreements between, among others, India and Thailand, Japan and Mexico, Japan and the Association of Southeast Asian Nations (ASEAN), India and ASEAN, and China and ASEAN—contain binding and nonbinding pledges related to information and communication technology (ICT) and e-commerce, as do trade-related statements by Asia-Pacific Economic Cooperation (APEC) and other cooperation agreements. PTAs thus function as a laboratory for digital trade rules, with a demonstration effect that is potentially useful for future multilateral or other preferential trade negotiations. Digital products. E-commerce chapters of PTAs that follow the U.S. model formalize a definition of digital content products; confirm the applicability of World Trade Organization (WTO) rules to e-commerce and the applicability of provisions on cross-border trade in services to electronically supplied services; ensure a zero-duty rate on the content of digital trade; and provide for nondiscriminatory and most favored nation (MFN) treatment for digital products such as music, films, and software (see table 12.1). Interestingly, from the point of view of rules of origin, in this template digital products, to benefit from nondiscrimination or MFN, must not be fully created in and exported by a contracting party of the PTA. Electronic trade in chapters on cross-border trade in services. The cross-border trade in services chapters of newly agreed U.S.-led PTAs also innovate to the benefit of the digital delivery of services. These PTAs use a negative-list approach to schedule services trade commitments. Assuming that no reservations are made, this top-down approach guarantees that narrow or outdated classification schemes and uncertainties relating to the mode of delivery do not limit the applicability of commitments to digitally delivered services. Importantly, the PTAs specify that “neither Party may require a service supplier of the other Party to establish or maintain a representative office or any form of enterprise, or to be resident, in its territory as a condition for the cross-border supply of a service.� MFN exemptions are ruled out. Again the devil is in the details, in the form of specified nonconforming measures. (For example, in the case of U.S.-led PTAs, U.S. state-level regulations are sometimes listed as exceptions.) On the side of services rules, the PTAs introduce strengthened transparency requirements, sector-specific mutual recognition annexes (e.g., for professional services), and binding rules on domestic regulation useful for digital trade. “Deep� digital trade rules. Two other developments in PTAs foster digital trade rules: • Nonbinding joint understandings on e-commerce calling for liberal digital trade principles and rules that foster the diffusion of ICT and e-commerce. A number of PTAs spell out a cooperation agenda on various aspects of the information society—in particular, in areas such as telecommunications policy, ICT standards and conformity assessments, interconnection and interoperability issues, cybersecurity, electronic signatures, the balance between privacy protection and the free cross-border flow of information, intellectual property rights, and consumer confidence in e-commerce. • Incorporation of “deep integration� digital trade principles as an integral component of bilateral trade agreements, with digital trade sometimes subject to dispute settlement provisions. Such deep digital trade provisions may relate to, for example, domestic regulation, transparency, consumer protection, data protection, authentication and digital signatures, and paperless trading. Source: Wunsch-Vincent 2008. is notably the case for express delivery services, which fea- opening and novel (procompetitive) rule making can be ture prominently in a number of recent PTAs. Similarly, found in the tourism sector, where the EPA between many recent PTAs, including South-South agreements, the EU and CARIFORUM recently blazed a new trail (see are giving greater attention to new areas of financial box 12.4). It is possible that many such advances could be services such as asset management and financial services replicated at the WTO level in the Doha Round or beyond, delivered through electronic means; to trade in some all the more so as most of them are being actively discussed segments of higher education and related services (e.g., in ongoing negotiations under the GATS and have been the vocational training and educational testing); and to the object of collective requests advanced by various coalitions wellness industry situated at the interface of tourism and of like-minded WTO members. health services. The sectoral patterns of PTA-induced market opening In some cases, market-opening advances rest on a com- in services trade described above appear to hold at the plementary set of new disciplines. Such a trend is most regional level. This is evidenced, for instance, in the work visible in the field of investment, where PTAs have of Fink and Molinuevo (2007), which offers an aggregated achieved significant forward movement over the WTO. It measure of the GATS+ nature of market-opening commit- is also notable in the area of procurement liberalization. ments in a sample of service-related PTAs concluded A further example of the close nexus between market among Asian countries (see figure 12.4). Such findings 260 Aaditya Mattoo and Pierre Sauvé Box 12.4. Tourism Liberalization in the EU–CARIFORUM EPA The tourism sector stands out as one in which developing countries possess clear comparative advantages in services trade. Accordingly, several developing-country governments have for some time been clamoring for provisions in trade agreements dedicated to the sector and its specificities. Such calls led, in the course of the World Trade Organization (WTO) Doha Round of trade negotiations, to the formulation of a draft annex to the General Agreement on Trade in Services (GATS) covering tourism services and to a collective request sponsored by a majority of developing-country members of the WTO, several of them in the Caribbean region. The collective inability of WTO members to complete the Doha Round has so far stymied progress in this area. Not surprisingly, proponents of tourism trade liberalization have taken their case to subsidiary settings. The precedent set in the EPA between the European Union (EU) and the Caribbean Forum of African, Caribbean, and Pacific (ACP) States (CARIFORUM) is likely to influence the treatment of tourism in other PTAs involving developing countries, as well as at the WTO level. Initially, CARIFORUM members had proposed the inclusion of a distinct tourism annex in the EPA. The origin of this initiative seems to have been the WTO Doha Round proposal submitted in 2001 by several Latin American and Caribbean countries. That proposal was the inspiration for the draft text on tourism formulated by the Caribbean Hotel and Tourism Association and adopted by the Caribbean Regional Negotiation Machinery in the EPA context. Mutual recognition. On the question of the negotiation of a mutual recognition agreement (MRA) for tourism service providers, the EPA states, in Article 114, that “the Parties shall co-operate towards the mutual recognition of requirements, qualifications, licenses or other regulations in accordance with Article 85 . . .� Article 85, which deals with mutual recognition in general, reaffirms the parties’ right to require that natural persons possess the necessary qualifications or professional experience (or both) to supply covered services. It also commits the parties to encourage the relevant professional bodies in their respective territories to jointly develop and propose recommendations on mutual recognition to the CARIFORUM–EU Trade and Development Committee to determine the criteria to be applied by the parties for the authorization, licensing, operation, and certification of investors and services suppliers. Tourism is identified, in Article 85(3), as one of the priority sectors for the development of mutual recognition arrangements under the agreement. The EPA specifically mandates the EU and CARIFORUM to encourage the relevant professional bodies in their territories to start negotiations three years after the EPA’s entry into force in order to jointly develop and propose recommendations on mutual recognition; this represents a hortatory, or “best-endeavors,� commitment. Competition policy disciplines. An important element of the Doha Development Agenda proposal which the Caribbean Hotel and Tourism Association retained in its EPA draft was the creation of a competitive safeguard for tourism. The inclusion of disciplines on anticompetitive practices was of key importance to CARIFORUM states because the global tourism industry is characterized by vertically integrated market structures and consolidated distribution channels controlled by a limited number of large international players, many of them based in the EU. Specifically, in accordance with the chapter dealing with competition policy, Article 111 of the EPA commits the parties to maintain or introduce measures designed to prevent suppliers from materially affecting “the terms of participation in the relevant market for tourism services by engaging in or continuing anti-competitive practices, including, inter alia, abuse of dominant position through imposition of unfair prices, exclusivity clauses, refusal to deal, tied sales, quantity restrictions or vertical integration.� The EPA provision on the prevention of anticompetitive practices, unlike other provisions on the tourism sector, is legally binding. Trade-related capacity building. The EPA’s treatment of tourism services features distinct development cooperation provisions, in contrast to other sectors, where such issues are addressed generically. The EPA contains an explicit commitment by the EU to help advance the tourism sector in the CARIFORUM states, as well as a nonexhaustive list of specific areas in which the parties agree to cooperate. Among these areas are capacity building for environmental management and the development of Internet-based marketing strategies for small and medium-size tourism enterprises. Also included is the upgrading of national accounts systems with a view toward facilitating the introduction of tourism satellite accounts (statistical instruments for analyzing the economic importance of tourism) at the regional and local levels. Source: Sauvé and Ward 2009b, 2009c; World Trade Organization, “Communication by Bolivia, Dominican Republic, Ecuador, El Salvador, Honduras, Nicaragua, Panama, Peru, and Venezuela: Draft Annex on Tourism,� S/CSS/W/107, September 26, 2001. reveal that although GATS+ advances are significant across negotiations have to date been most successful. Perhaps the all sectors, they are particularly noticeable in the areas of GATS advances lessened the scope or the perceived need for business services (reflecting the emergence of digital trade, significant new advances in PTAs. e-commerce, and the outsourcing revolution in services); Using the Fink-Molinuevo methodology and applying distribution; and education, health, and transport services— it to the four largest members of CARIFORUM in the con- all areas that proved difficult in the WTO context during text of the EPA with the EU, Sauvé and Ward (2009b) reveal the Uruguay Round and in the more recent context of the a broad pattern of WTO+ or WTO-extra advances arising at Doha Development Agenda. the preferential level (see annex figures 12A.1–12A.4 ). Less progress, relatively, has been seen in Southeast Asian The East Asian PTA experience described in Fink and PTAs in areas such as telecommunications and financial Molinuevo (2007) confirms the partial, incremental services. These are precisely the sectors in which GATS nature of market opening in services trade. This feature Services 261 Figure 12.4. GATS+ Advances in East Asian PTAs with Services Provisions, by Sector business services communication services construction and related engineering services distribution services educational services environmental services financial services health and related social services recreational, cultural, and sporting services tourism and travel-related services transport services other services not included elsewhere 0 20 40 60 80 100 share of total (percent) partial full partial full partial full unbound GATS provisions PTAs with GATS+ PTAs with GATS-extra only provisions provisions Source: Fink and Molinuevo 2007. Note: GATS, General Agreement on Trade in Services; PTA, preferential trade agreement. “Partial� and “full� refer to the degree of market opening. is particularly apparent when commitments are analyzed As in other preferential settings and at the WTO level, less on a modal basis (see figure 12.5). On the positive side, progress has been made in East Asian PTAs on mode 1 (cross- the most significant GATS+ advances in East Asian PTAs border supply) than mode 3 (commercial presence). This relate to the two modes of supply likely to generate the difference reflects the generally greater caution shown strongest developmental returns. These are mode 4 toward transactions that many host countries feel they can- (movement of natural persons), the least committed of not regulate, or at least, not easily. all modes under GATS, and mode 3 (commercial pres- Evidence of iterative learning by doing. The relationship ence), the most committed of the modes subject to GATS between PTAs and the WTO is not unidirectional in char- bindings, and the principal means through which serv- acter but involves iterative, two-way interaction between ices are traded internationally. This outcome is not the two layers of trade governance in ways that can inform entirely surprising, given the reluctance of countries to subsequent patterns of rule making and market opening contemplate mode 4 liberalization on an MFN basis in at both levels. Examples of such interaction are notably the WTO.19 found in areas in which WTO jurisprudence has clarified 262 Aaditya Mattoo and Pierre Sauvé Figure 12.5. GATS+ Advances in East Asian PTAs with Services Provisions, by Mode mode 1 mode 2 mode 3 mode 4 total 0 20 40 60 80 100 share of total (percent) partial full partial full partial full unbound GATS provisions PTAs with GATS+ PTAs with GATS-extra only provisions provisions Source: Fink and Molinuevo 2007. Note: GATS, General Agreement on Trade in Services; PTA, preferential trade agreement. “Partial� and “full� refer to the degree of market opening. Modes are defined as follows: mode 1, cross-border supply of services; mode 2, consumption abroad; mode 3, commercial presence; mode 4, movement of natural persons. or interpreted the scope of key provisions governing serv- dimensions is the increasing reliance, particularly in agree- ices trade that are typically found both in the WTO-GATS ments brokered by the EU, on nonbinding provisions and in the services and investment chapters of PTAs. Bap- embedded in PTAs alongside treaty provisions that are tista Neto (2009), for instance, has found evidence of legally binding and enforceable. Advances of this type NAFTA-minus treatment of recreational services in U.S. reflect the ever-broadening remit of trade rule making and reservation lists following the decision by the WTO’s Dis- the commensurate desire of parties to assign to regulatory pute Settlement Body and Appellate Body (DSB and AB) cooperation a number of trade- and investment-facilitating concerning online gambling (United States–Gambling roles. For various reasons, such PTA advances may well be Services). Similarly, the most recent DSB and AB decisions limited to preferential settings and may encounter difficul- in the China–Publications and Audiovisual Products dis- ties in being agreed in the WTO. This may notably be the pute have prompted some observers to note the need for case when parties harbor concerns over MFN-induced China to adjust its future PTA commitments with a view free riding, when particular policy sensitivities arise at the to ensuring the preservation of adequate policy space WTO level that can be contained or addressed in a PTA set- within which to pursue cultural policy objectives (Chen ting, or when parties may simply not desire binding and and Shi 2011). enforceable obligations. Cultural cooperation appears to Coexistence of hard- and soft-law provisions. A final, correspond to the former category of policy domains, salient, trend that emerges from the most recent generation while aid for trade and its design in the services field likely of PTAs and that has both rule-making and market-opening falls more within the latter category (see box 12.5). Services 263 Box 12.5. Cultural Cooperation and Aid for Trade in the EU–CARIFORUM EPA Cultural cooperation A novel feature of the economic partnership agreement (EPA) between the European Union (EU) and the Caribbean Forum of African, Caribbean, and Pacific (ACP) States (CARIFORUM) is its inclusion of a protocol on cultural cooperation between the parties. The protocol establishes a clear precedent for addressing matters relating to cultural industries within preferential trade agreements (PTAs), laying the basis for the incorporation of similar provisions in other EPAs. The inclusion of language on cultural cooperation matters marks a significant evolution in EU attitudes toward the subject in a trade policy context. Earlier, the EU had tried to preserve maximum policy autonomy toward cultural industries by eschewing any commitments in trade agreements and, in the case of the Doha Development Agenda (DDA), by refusing to direct negotiating requests to its trading partners or to entertain offers in response to requests by trading partners. The advances made in the protocol respond to CARIFORUM members’ strong offensive interests in this area, notably the music industry. The EPA protocol establishes a framework within which the parties can work together to facilitate exchanges of cultural activities, goods, and services, as well as the movement of artists and other cultural professionals, and to improve cinematographic cooperation. The protocol can be viewed as the first concrete response to Article 16, on preferential treatment, of the United Nations Educational, Scientific, and Cultural Organization (UNESCO) Convention on the Protection and Promotion of the Diversity of Cultural Expressions. According to CARIFORUM officials, the conclusion of the coproduction agreements called for under the protocol will enable Caribbean audiovisual producers to access new sources of funding for creative projects. Given the EU’s long-standing sensitivities in the audiovisual sector, the protocol likely comes as close to opening new market access opportunities as the EU’s EPA partners could have hoped, without actually resulting in new liberalization commitments on national treatment or market access. Aid for trade The cooperation elements of the EU–CARIFORUM EPA reflect the desire of EU members to infuse the agreement with a concrete development dimension. In so doing, the EPA charts useful new territory at a time when the multilateral community is struggling to give operational meaning to the concept of aid for trade. The EPA text does not contain explicit language on the level of development financing to be made available overall, nor does it spell out the specific issues and sectors subject to coverage by the agreement. These omissions have sparked much criticism throughout the CARIFORUM region over the allegedly unbalanced nature of the agreement, in that its development provisions remain somewhat abstract and not legally enforceable, while its liberalization commitments are up front, legally binding, and enforceable. Responding to such critiques, the Caribbean Regional Negotiating Machinery (CRNM), which led the negotiations on the CARIFORUM side, cautioned that “any perceptions about the EPA’s practical deficiencies with respect to the treatment of development and development cooperation and assistance should first be tempered by the recognition that as a trade agreement, the EPA should not be perceived to be the primary vehicle through which development may be achieved.� Rather, it should be considered “one strategic instrument in a range of economic development strategies.� According to the Joint Declaration on Development Co-operation, which is annexed to the EPA and constitutes an integral part of it, a package of 165 million euros has been set aside for the six years following the agreement’s entry into force to fund activities that are identified and rank-ordered in the Caribbean regional indicative plan. In addition to funding for that plan, each CARIFORUM state will receive funds for its national indicative plans (NIP); two priority projects must be identified for such additional funding. The Dominican Republic and Jamaica announced that they would be using some of the financing under their respective NIPs for purposes of EPA implementation. The minimum cost of implementing the EPA’s provisions on investment, trade in services, and e-commerce and of addressing capacity constraints at the national and regional levels has been estimated at 15.6 million euros. The constraints identified include insufficient numbers of specialists and experts; limited human resources within both public and private sectors; the absence of an organized services sector body through which stakeholders can be mobilized; general absence of infrastructure; and inadequacy of financial resources. Key tasks include building regulatory capacity, overcoming information asymmetries to assist CARIFORUM firms and entities in identifying business opportunities in the European market, and developing productive capacity in goods and cultural services. Specific technical assistance efforts are to be directed at the following objectives: (a) improving the ability of CARIFORUM service suppliers to gather information on and to meet EU members’ regulations and standards; (b) increasing the export capacity of local services suppliers; (c) facilitating interaction and dialogue between services suppliers in both parties; (d) addressing quality and standards in those areas in which the CARIFORUM states have undertaken commitments; (e) developing and implementing regulatory regimes for specific services at the CARIFORUM level and in the signatory CARIFORUM states; (f) establishing mechanisms for promoting investment and joint ventures between service suppliers of the parties; and (g) enhancing the capacities of investment promotion agencies in CARIFORUM states. Source: Sauvé and Ward 2009b. The quotation from the CRNM is from RNM Update 0802, http://www.crnm.org/documents/updates_2008/ rnmupdate0802.htm (accessed April 19, 2008). The information on costs is from CARICOM Secretariat 2008, 10, and the listing of objectives is from CARICOM Secretariat 2008, Article 121 (2). 264 Aaditya Mattoo and Pierre Sauvé Annex Annex Table 12A.1 (continued) Agreement Typea Annex Table 12A.1. Preferential Trade Agreements (PTAs) Mexico–Nicaragua FTA, EIA That Include Provisions on Trade in Services NAFTA FTA, EIA New Zealand–Singapore FTA, EIA Agreement Typea Nicaragua–Taiwan, China FTA, EIA ANZCERTA (S) EIA Pakistan–Malaysia FTA, EIA ASEAN–China (S) EIA Panama–Chile FTA, EIA CARICOM (S) EIA Panama–Costa Rica (Panama–Central America) FTA, EIA EFTA (S) EIA Panama–El Salvador (Panama–Central America) FTA, EIA Mercosur (S) EIA Panama–Honduras (Panama–Central America ) FTA, EIA Australia–Chile FTA, EIA Panama–Singapore FTA, EIA Brunei Darussalam–Japan FTA, EIA Panama and Taiwan, China FTA, EIA CAFTA–DR FTA, EIA Peru–China FTA, EIA Canada–Chile FTA, EIA Peru–Singapore FTA, EIA Canada–Peru FTA, EIA Singapore–Australia FTA, EIA Chile–Colombia FTA, EIA Thailand–Australia FTA, EIA Chile–Costa Rica (Chile–Central America) FTA, EIA Thailand–New Zealand FTA, EIA Chile–El Salvador (Chile–Central America) FTA, EIA Trans-Pacific Strategic Economic Partnership FTA, EIA Chile–Japan FTA, EIA U.S.–Australia FTA, EIA Chile–Mexico FTA, EIA U.S.–Bahrain FTA, EIA China–Hong Kong SAR, China FTA, EIA U.S.–Chile FTA, EIA China–Macao SAR, China FTA, EIA U.S.–Jordan FTA, EIA China–New Zealand FTA, EIA U.S.–Morocco FTA, EIA China–Singapore FTA, EIA U.S.–Oman FTA, EIA Costa Rica–Mexico FTA, EIA U.S.–Peru FTA, EIA EEA (S) EIA U.S.–Singapore FTA, EIA EFTA–Chile FTA, EIA EU-15 enlargement CU, EIA EFTA–Korea, Rep. FTA, EIA EU-25 enlargement CU, EIA EFTA–Mexico FTA, EIA EU-27 enlargement CU, EIA EFTA–Singapore FTA, EIA EU Treaty CU, EIA EU–Albania FTA, EIA Total number of services PTAs 76 EU–CARIFORUM EPA FTA, EIA Total number of services-only PTAs 6 EU–Chile FTA, EIA Total number of goods PTAs 196 EU–Croatia FTA, EIA Total number of PTAs in force 272 EU–Macedonia, FYR FTA, EIA Total number of services PTAs as a percentage EU–Mexico FTA, EIA of total PTAs in force 27.94 Iceland–Faroe Islands FTA, EIA India–Singapore FTA, EIA Source: World Trade Organization, Regional Trade Agreement (RTA) database, http://rtais.wto.org/UI/PublicMaintainRTAHome .aspx. Japan–Indonesia FTA, EIA Note: The notation (S) after an agreement denotes services only. WTO Japan–Malaysia FTA, EIA statistics on PTAs are based on notification requirements rather than on Japan–Mexico FTA, EIA physical numbers of PTAs. Thus, for a PTA that includes both goods and services, two notifications are counted (one for goods and the other Japan–Philippines FTA, EIA services), even though there is actually only one PTA. ANZCERTA, Japan–Singapore FTA, EIA Australia–New Zealand Closer Economic Relations Trade Agreement; Japan–Switzerland FTA, EIA ASEAN, Association of Southeast Asian Nations; CAFTA–DR, Dominican Japan–Thailand FTA, EIA Republic–Central America Free Trade Agreement; CARICOM, Caribbean Community; CARIFORUM, Caribbean Forum of African, Caribbean, and Japan–Vietnam FTA, EIA Pacific (ACP) States; EEA, European Economic Area; EFTA, European Free Jordan–Singapore FTA, EIA Trade Association; EU, European Union; Mercosur, Southern Cone Korea, Rep.–Chile FTA, EIA Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement. EU-15 refers to the EU after the 1995 enlargement; the Korea, Rep.–Singapore FTA, EIA members were Austria, Belgium, Denmark, Finland, France, Germany, Mexico–El Salvador (Mexico–Northern Triangle) FTA, EIA Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, Mexico–Guatemala (Mexico–Northern Triangle) FTA, EIA and the United Kingdom. EU-25 and EU-27 refer to successive enlarge- Mexico–Honduras (Mexico–Northern Triangle) FTA, EIA ments of the EU, in 2004 and 2007. a. CU, customs union; FTA, free trade agreement; EIA, economic (continued) integration agreement. Services 265 Annex Table 12A.2. Classification of Preferential Trade Agreements (PTAs) Featuring Services Provisions by Country Group North-North North-South South-South ANZCERTA Australia–Chile ASEAN–China EEA Brunei Darussalam–Japan CARICOM EFTA CAFTA–DR Chile–Colombia EU-15 enlargement Canada–Chile Chile–Costa Rica (Chile–Central America) EU-25 enlargement Canada–Peru Chile–El Salvador (Chile–Central America) EU-27 enlargement Chile–Japan Chile–Mexico EU Treaty China–New Zealand China–Hong Kong SAR, China Iceland–Faroe Islands EFTA–Chile China–Macao SAR, China Japan–Switzerland EFTA–Korea, Rep. China–Singapore U.S.–Australia EFTA–Mexico Costa Rica–Mexico EFTA–Singapore India–Singapore EU–Albania Jordan–Singapore EU–CARIFORUM EPA Korea, Rep.–Chile EU–Chile Korea, Rep.–Singapore EU–Croatia Mercosur EU–Macedonia, FYR Mexico–El Salvador (Mexico–Northern Triangle) EU–Mexico Mexico–Guatemala (Mexico–Northern Triangle) Japan–Indonesia Mexico–Honduras (Mexico–Northern Triangle) Japan–Malaysia Mexico–Nicaragua Japan–Mexico Nicaragua–Taiwan, China Japan–Philippines Pakistan–Malaysia Japan–Singapore Panama–Chile Japan–Thailand Panama–Costa Rica (Panama–Central America) Japan–Vietnam Panama–El Salvador (Panama–Central America) NAFTA Panama–Honduras (Panama–Central America) New Zealand–Singapore Panama–Singapore Singapore–Australia Panama–Taiwan, China Thailand–Australia Peru–China Thailand–New Zealand Peru–Singapore Trans-Pacific Strategic Economic Partnership U.S.–Bahrain U.S.–Chile U.S.–Jordan U.S.–Morocco U.S.–Oman U.S.–Peru U.S.–Singapore Source: World Trade Organization, regional trade agreement (RTA) database, http://rtais.wto.org/UI/PublicMaintainRTAHome.aspx. Note: PTAs include all categories of preferential agreements. ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Agreement; ASEAN, Association of Southeast Asian Nations; CAFTA–DR, Dominican Republic–Central America Free Trade Agreement; CARICOM, Caribbean Community; CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; EEA, European Economic Area; EFTA, European Free Trade Association; EU, European Union; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement. EU-15 refers to the EU after the 1995 enlargement; the members were Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Netherlands, Portugal, Spain, Sweden, and the United Kingdom. EU-25 and EU-27 refer to successive enlargements of the EU, in 2004 and 2007. 266 Aaditya Mattoo and Pierre Sauvé Annex Figure 12A.1. GATS and EU–CARIFORUM Commitments Compared: Barbados business services communication services construction and related engineering services distribution services educational services environmental services financial services health-related and social services tourism and travel-related services recreational, cultural, and sporting services transport services other services not included elsewhere 0 20 40 60 80 100 share of total (percent) partial full partial full partial full unbound GATS only EPA improvements EPA new subsectors Source: Sauvé and Ward 2009b. Services 267 Annex Figure 12A.2. GATS and EU–CARIFORUM Commitments Compared: Dominican Republic business services communication services construction and related engineering services distribution services educational services environmental services financial services health-related and social services tourism and travel-related services recreational, cultural, and sporting services transport services other services not included elsewhere 0 20 40 60 80 100 share of total (percent) partial full partial full partial full unbound GATS only EPA improvements EPA new subsectors Source: Sauvé and Ward 2009b. 268 Aaditya Mattoo and Pierre Sauvé Annex Figure 12A.3. GATS and EU–CARIFORUM Commitments Compared: Jamaica business services communication services construction and related engineering services distribution services educational services environmental services financial services health-related and social services tourism and travel-related services recreational, cultural, and sporting services transport services other services not included elsewhere 0 20 40 60 80 100 share of total (percent) partial full partial full partial full unbound GATS only EPA improvements EPA new subsectors Source: Sauvé and Ward 2009b. Services 269 Annex Figure 12A.4. GATS and EU–CARIFORUM Commitments Compared: Trinidad and Tobago business services communication services construction and related engineering services distribution services educational services environmental services financial services health-related and social services tourism and travel-related services recreational, cultural, and sporting services transport services other services not included elsewhere 0 20 40 60 80 100 share of total (percent) partial full partial full partial full unbound GATS only EPA improvements EPA new subsectors Source: Sauvé and Ward 2009b. 270 Aaditya Mattoo and Pierre Sauvé Notes series of negotiations, ultimately resulting in a common market free of barriers to services trade by a target date (set initially at 2005 but not yet The authors are grateful to Natasha Ward and Sacha Wunsch-Vincent for achieved). their valuable comments on and written contributions to an earlier draft 11. Most PTAs that employ a negative-list approach to liberaliza- of this chapter. tion feature unbound reservations listing sectors in which members 1. Evidence of such causality can be found in recent PTAs involving wish to preserve the right to introduce new nonconforming measures the United States, notably in reservations made with regard to online in the future. In many PTAs, particularly those modeled on NAFTA, gambling services (Baptista Neto 2009). It may further be expected in such reservations nonetheless oblige member countries to list existing the aftermath of the most recent WTO ruling in the dispute brought by discriminatory or access-impairing measures whose effects on foreign the United States against China in the area of publications and audiovisual services or service suppliers might in the future become more burden- products (Chen and Shi 2011). some. 2. Only the Mercosur Protocol and Decision 439 of the Andean Com- 12. The suggestion has been made that WTO members could address munity provide that no deviation from MFN and national treatment is to this issue in GATS without revisiting the agreement’s negotiating modal- be allowed among members. ity by agreeing to a new framework provision designed to encourage gov- 3. There is a measure of asymmetry in this provision, as the obligation ernments to reflect the statutory or regulatory status quo in their sched- does not apply to bilateral or plurilateral agreements between an Associa- uled commitments, while retaining the voluntary nature of such tion of Southeast Asian Nations (ASEAN) state or the ASEAN states, on the commitments. See Sauvé and Wilkie (2000). one hand, and a nonparty or Australia or New Zealand, on the other. 13. This caveat is important because a number of PTAs, particularly 4. Other examples of demonstration-effect or precedent-setting those conducted along North-South lines, have seen powerful partners regionalism are the provisions on the linkages between trade and labor reserve all measures of a subnational nature through one sweeping reser- standards inserted in the waning days of the Clinton administration into vation that yields no information on the nature and sectoral incidence of the U.S.–Jordan FTA and the recurring tendency of the European Com- nonconforming measures maintained by subnational governments. Such mission to insert disciplines on trade and competition into the EU’s PTAs reservations also greatly reduce the potential scope of the PTA in ques- with developing countries. tion, to the extent that in many federal countries, important pockets of 5. Whereas similar GATS language states that the measures in ques- services regulation apply at the subnational level. The insurance sector in tion should not restrict the supply of a service under any of the four GATS the United States and many energy-related services in Canada are cases in modes, the NAFTA-type agreements narrow this requirement to the point. cross-border supply of a service. No comparable provision can be found 14. For a fuller discussion of the modalities and uses to which a trade- in these agreements’ investment chapters. related regulatory audit may be put, see Marconini and Sauvé (2010); 6. Despite notable progress in PTAs, government procurement prac- Sauvé (2010). tices continue in most instances to be the province of discriminatory 15. One troubling example is recent PTA practice by the United practices. In the case of NAFTA, for instance, even though the scope of States, which increasingly uses sweeping negative-list reservations that covered purchases was quadrupled from that in the 1987 Canada–U.S. exclude all measures affecting services at the subnational level. Recent U.S. FTA, covered entities represented only a tenth of North America’s civilian PTAs are also notable for excluding mode 4 (movement of natural per- procurement market at the time of the agreement’s entry into force. See sons) commitments. Hart and Sauvé (1997). 16. In NAFTA, for instance, subnational governments were initially 7. More restrictive rules of origin conditioning the receipt of prefer- given an extra two years to complete their lists of nonconforming meas- ences may relate to factors such as local incorporation (if it denies benefits ures pertaining to services and investment. The NAFTA parties subse- to branches of third-country invested entities), place of incorporation or quently decided not to complete the lists at the subnational level, opting location of headquarters or ownership, and control tests aimed at limiting instead for a standstill on existing nonconforming measures. Compliance PTA benefits to local juridical persons. Examples of the latter can be with the production of negative lists has been similarly problematic else- found under Mercosur and the Andean Pact. where in the Western Hemisphere, as a number of agreements were con- 8. The purpose of the discussion here is to note such differences with- cluded without such lists being finalized and without firm deadlines for out advocating any implicit hierarchy of policy desirability. Both doing so. The inability of users to access the information contained in the approaches have strengths and weaknesses. The governance-enhancing negative lists to such agreements deprives the latter of an important good- aspects of negative listing have, however, been noted by several observers. governance-promoting feature. See, in particular, Sauvé (1996); Snape and Bosworth (1996); WTO (2001); 17. In trade negotiations, the term “single undertaking� means that Stephenson (2002). For a fuller discussion of the good-governance- participants accept or reject the outcome of multiple negotiations in a sin- promoting aspects of PTAs, see Baldwin, Evenett, and Low (2009), as well gle package, rather than select among them. as Chauffour and Maur, ch. 1 in this volume. 18. Countries reluctant to go beyond GATS include, for instance, 9. Members of Mercosur adopted a slightly different version of the EFTA countries; India; the EU prior to the EPA with CARIFORUM and, positive-list approach, with a view toward liberalizing services trade to some extent, the FTA with Chile; and the ASEAN countries other than within the region. According to Mercosur’s Protocol of Montevideo on Singapore. In some particular cases, the reason for these more limited Trade in Services, annual rounds of negotiations based on the scheduling advances may be that PTA negotiations took place before the last DDA of increasing numbers of commitments in all sectors (with no exclusions) offer, and what was conceded in the PTA may later find its way into a are to result in the elimination of all restrictions on services trade among revised GATS offer. members within 10 years of the entry into force of the protocol. The latter 19. An example drawn from the Asian experience relates to the labor has yet to enter into force. See Peña (2000); Stephenson (2001b). mobility provisions found in recent Japanese PTAs. These include novel 10. The Andean Community has adopted a somewhat different ver- features aimed at assisting partner countries with training in the home sion of the negative-list approach. Decision 439, on trade in services, spec- country prior to their admission as professionals in the Japanese labor ifies that the process of liberalization is to begin when comprehensive market, with a view to meeting Japanese licensing requirements in nurs- (nonbinding) national inventories of measures affecting trade in services ing and other health-related occupations. Although the numerical quotas for all members of the Andean Community are finalized. Discriminatory agreed by Japan in these areas remain low relative to the supply capacity restrictions listed in these inventories were to be lifted gradually through a (and negotiating interests) of sending countries, such provisions Services 271 nonetheless represent a step forward in the treatment of mode 4 issues in Curzon, Gerard. 1965. Multilateral Commercial Diplomacy: The General a context of population aging and labor market shortages in OECD Agreement on Tariffs and Trade and Its Impact on National Commercial countries. Policies and Techniques. London: Michael Joseph. Dam, Kenneth W. 1970. The GATT: Law and International Economic Organization. Chicago, IL: University of Chicago Press. Bibliography Delimatsis, Panagiotis. 2008. “Towards a Horizontal Necessity Test for Services: Completing the GATS Article VI:4 Mandate.� In GATS and Abugattas Majluf, Luis, and Sherry M. Stephenson. 2003. “Liberalization the Regulation of International Trade in Services, ed. Marion Panizzon, of Trade in Services: Options and Implications for Latin American Nicole Pohl, and Pierre Sauvé, 370–96. Cambridge, U.K.: Cambridge and Caribbean Countries.� In Trade Negotiations in Latin America: University Press. Problems and Prospects, ed. Diana Tussie. Houndsmills, Basingstoke, Doove, Samantha, Owen Gaabitas, Duc Nguyen-Hong, and Joe Owen. Hampshire, U.K.: Palgrave Macmillan. 2001. “Price Effects of Regulation: Telecommunication, Air Transport Adlung, Rudolf. 2007. “Negotiations on Safeguards and Subsidies in Serv- and Electricity Supply.� Productivity Commission Working Paper ices: A Never-Ending Story?� Journal of International Economic Law 1682, AusInfo, Canberra. 10 (2, June): 235–65. Ethier, Wilfred J. 1982. “National and International Returns to Scale in the Baier, Scott L., and Jeffrey H. Bergstrand. 2001. “International Trade in Modern Theory of International Trade.� American Economic Review Services, Free Trade Agreements, and the WTO.� In Services in the 72 (3, June): 389–405. International Economy, ed. Robert M. Stern, ch. 7. Ann Arbor, MI: Evenett, Simon J. 2005. “‘Competitive Liberalization’: A Tournament University of Michigan Press. Theory-Based Interpretation.� University of St. Gallen, St. Gallen, Baldwin, Richard E. 2000. “Regulatory Protectionism, Developing Nations Switzerland. http://www.alexandria.unisg.ch/publications/22172. and a Two-Tier World Trade System.� CEPR Discussion Paper 2574, Evenett, Simon J., and Michael Meier. 2007. “An Interim Assessment of the Centre for Economic Policy Research, London. U.S. Trade Policy of ‘Competitive Liberalization.’� University of St. ———. 2006. “Multilateralising Regionalism: Spaghetti Bowls as Build- Gallen Discussion Paper 2007-18, St. Gallen, Switzerland. http://ssrn ing Blocs on the Path to Global Free Trade.� 2006 World Economy .com/abstract=985521. Annual Lecture, Nottingham, England, June 22. Feketekuty, Geza. 1998. “Setting the Agenda for Services 2000: The Next Baldwin, Richard, Simon Evenett, and Patrick Low. 2009. “Beyond Tariffs: Round of Negotiations on Trade in Services.� Presented at conference on Multilateralizing Non-Tariff RTA Commitments.� In Multilateralizing “The Future of Services Trade Liberalization: Services 2000,� sponsored Regionalism: Challenges for the Global Trading System, ed. Richard E. Bald- by the Coalitions of Service Industries, Ditchley Park, England, April. win and Patrick Low. Cambridge, U.K.: Cambridge University Press. ———. 2000a. “Assessing Improving the Architecture of GATS.� In GATS Baptista Neto, J. A. 2009. “United States Preferential Trade Agreements in 2000: New Directions in Services Trade Liberalization, ed. Pierre Sauvé Services: How Far Beyond the NAFTA?� Master in international law and Robert M. Stern, 85–111. Washington, DC: Brookings Institution and economics thesis, World Trade Institute, University of Bern, Bern. Press. Beviglia Zampetti, Americo. 2000. “Market Access through Mutual Recogni- ———. 2000b. “Regulatory Reform and Trade Liberalization in Services.� tion: The Promise and Limits of GATS Article VII.� In GATS 2000: New In GATS 2000: New Directions in Services Trade Liberalization, ed. Directions in Services Trade Liberalization, ed. Pierre Sauvé and Robert Pierre Sauvé and Robert M. Stern, 225–40. Washington, DC: Brook- M. Stern, 283–306. Washington, DC: Brookings Institution Press. ings Institution Press Beviglia Zampetti, Americo, and Pierre Sauvé. 2006. “Rules of Origin for Findlay, Christopher, and Tony Warren. 2000. Impediments to Trade in Services: Economic and Legal Considerations.� In The Origin of Services: Measurement and Policy Implications. London and New York: Goods: Rules of Origin in Regional Trade Agreements, ed. Olivier Cadot, Routledge. Antoni Estevadeordal, Akiko Suwa Eisenmann, and Thierry Verdier, Finger, J. Michael. 1992. “GATT’s Influence on PTAs.� Presented at a 114–45. Oxford, U.K.: Oxford University Press. World Bank and Centre for Economic Policy Research (CEPR) confer- ———. 2007. “International Investment.� In Research Handbook in Inter- ence on “New Dimensions in Regional Integration,� Washington, DC, national Economic Law, ed. Alan O. Sykes and Andrew Guzman, April 2–3. 211–70. London: Edgar Elgar. Fink, Carsten. 2007. “Liberalizing Services Trade in the ASEAN Region.� Bhagwati, Jagdish. 1990. “Multilateralism at Risk: The GATT Is Dead, In A Handbook of International Trade in Services, ed. Aaditya Mattoo, Long Live the GATT.� World Economy, 13 (2): 149–69. Robert M. Stern, and Gianni Zanini. Oxford, U.K.: Oxford University CARICOM (Caribbean Community) Secretariat. 2008. “Implementation Press. of the CARIFORUM-EC Economic Partnership Agreement.� ———. 2008. “PTAs in Services: Friends or Foes of the Multilateral Trad- CARICOM Secretariat, Georgetown. ing System?� In Opening Markets for Trade in Services: Countries and CEC (Commission of the European Communities). 2002. “The State of Sectors in Bilateral and WTO Negotiations, ed. Juan A. Marchetti and the Internal Market for Services: Report from the Commission to the Martin Roy, 113–47. Cambridge, U.K.: Cambridge University Press. Council and the European Parliament, Presented under the First Stage Fink, Carsten, and Marion Jansen. 2008. “Services Provisions in Regional of the Internal Market Strategy for Services.� COM(2002)41 Final, Trade Agreements: Stumbling Blocks or Building Blocks for Multilat- July 30, Commission of the European Communities, Brussels. eral Liberalization?� Multilateralizing Regionalism: Challenges for the Chen, Weidong, and Jingxia Shi. 2011. “The ‘Specificity’ of Cultural Prod- Global Trade System, ed. Richard Baldwin and Patrick Low, 221–61. ucts versus the ‘Generality’ of Trade Obligations: Reflecting on Cambridge U.K.: Cambridge University Press. ‘China–Publications and Audiovisual Products.’� Journal of World Fink, Carsten, and Martin Molinuevo. 2007. “East Asian FTAs in Services: Trade 45 (1): 159–86. Roaring Tigers or Timid Pandas?� East Asia and Pacific Region, Report Coe, David T., Elkanah Helpman, and Alexander W. Hoffmaister. 1997. 40175, World Bank, Washington, DC. http://siteresources.worldbank “North-South R&D Spillovers.� Economic Journal 107 (440): 134–49. .org/INTEAPSUMEASPR/Resources/2576847-1163691185244/East_ Copeland, Brian R. 2001. “Benefits and Costs of Trade and Investment Asian_FTAs_in_Services.pdf. Liberalization in Services: Implications from Trade Theory.� Paper Frankel, Jeffrey A., Ernesto Stein, and Shang-Jin Wei. 1995. “Trading Blocs prepared for the Department of Foreign Affairs and International and the Americas: The Natural, the Unnatural, and the Super-Natural.� Trade, Government of Canada, Ottawa. Journal of Development Economics 47: 61–95. 272 Aaditya Mattoo and Pierre Sauvé Glassman, Cynthia A. 2000. “Customer Benefits from Current Informa- Leebron, David. 2003. “Regulatory Discrimination in Domestic United tion Sharing by Financial Services Companies.� Study for the Finan- States Law: A Model for the GATS?� In Domestic Regulation and Serv- cial Services Roundtable, Washington, DC. ice Trade Liberalization, ed. Aaditya Mattoo and Pierre Sauvé, 43–56. Greenaway, David. 1993. “Study on Regional Integration: Third Country Washington, DC: World Bank and Oxford University Press. Responses.� Background paper for the OECD Trade Directorate project Leroux, Eric. 1995. Le libre-échange nord-américain et les services finan- on “Regional Integration,� Organisation for Economic Co-operation ciers. Collection Minerve. Montréal: Editions Yvon Blais. and Development, Paris. Low, Patrick, and Aaditya Mattoo. “Is There a Better Way? Alternative Hamilton, Bob, and John Whalley. 1984. “Efficiency and Distributional Approaches to Liberalization under GATS.� In GATS 2000: New Direc- Implications of Global Restrictions on Labour Mobility: Calculations and tions in Services Liberalization, ed. Pierre Sauvé and Robert M. Stern. Policy Implications.� Journal of Development Economics 14 (1): 61–75. Washington, DC: Brookings Institution Press. Hardin, Alexis, and Leanne Holmes. 1997. “Service Trade and Foreign Lumengo-Neso, Olivier, Marcelo Olarreaga, and Maurice Schiff. 2001. Direct Investment.� Industry Commission Staff Research Paper, Pro- “On ‘Indirect’ Trade-Related Research and Development Spillovers.� ductivity Commission, Government Publishing Services, Canberra. Policy Research Working Paper 2580, World Bank, Washington, DC. Hart, Michael, and Pierre Sauvé. 1997. “Does Size Matter? Canadian Perspec- Marchetti, Juan A. 2004. “Developing Countries in the WTO Services tives on the Development of Government Procurement Disciplines in Negotiations.� ERSD Staff Working Paper ERSD-2004-06, Economic North America.� In Law and Policy in Public Purchasing: The WTO Agree- Research and Statistics Division, World Trade Organization, Geneva. ment on Government Procurement, ed. Bernard M. Hoekman and Petros Marchetti, Juan A., and Martin Roy. 2008. “Services Liberalization in the C. Mavroidis, 203–21. Ann Arbor, MI: University of Michigan Press. WTO and in PTAs.� Opening Markets for Trade in Services: Countries Hoekman, Bernard. 1996. “Assessing the General Agreement on Trade in and Sectors in Bilateral and WTO Negotiations, ed. Juan A. Marchetti Services.� In The Uruguay Round and the Developing Countries, and Martin Roy, 61–112. Cambridge, U.K.: Cambridge University ed. Will Martin and L. Alan Winters. Cambridge, U.K.: Cambridge Press. University Press. Marconini, Mario. 2005. “Emergency Safeguard Measures in the GATS: Hoekman, Bernard, and Carlos A. Primo Braga. 1997. “Protection and Trade Beyond Feasible and Desirable.� UNCTAD/DITC/TNCD/2005/4, United in Services: A Survey.� Open Economies Review 8 (3, July): 285–308. Nations Conference on Trade and Development (UNCTAD), Geneva. Hoekman, Bernard M., and Pierre Sauvé. 1994. Liberalizing Trade in Serv- Marconini, Mario, and Pierre Sauvé. 2010. “Negotiating Trade in Services: ices. World Bank Discussion Paper 243. Washington, DC: World Bank. A Practical Guide for Developing Countries.� In Trade in Services Holmes, Leanne, and Alexis Hardin. 2000. “Assessing Barriers to Services Negotiations: A Guide for Developing Countries, ed. Sebastián Sáez, Sector Investment.� In Impediments to Trade in Services: Measurement 19–86. Washington, DC: World Bank. and Policy Implications, ed. Christopher Findlay and Tony Warren. Markusen, James. 1983. “Factor Movements and Commodity Trade as London and New York: Routledge. Complements.� Journal of International Economics 14: 341–56. Horn, Henrik, Petros C. Mavroidis, and André Sapir. 2010. “Beyond the Mattoo, Aaditya. 1998. “Financial Services and the WTO: Liberalization in WTO? An Anatomy of EU and US Preferential Trade Agreements.� the Developing and Transitional Economies.� Presented at workshop World Economy 33 (11): 1565–88. on “Measuring Impediments to Trade in Services,� Productivity Com- IADB (Inter-American Development Bank). 2002. Beyond Borders: The mission, Canberra, April 30–May 1. New Regionalism in Latin America. Economic and Social Progress in Mattoo, Aaditya, and Carsten Fink. 2002. “Regional Agreements and Latin America, 2002 Report. Washington, DC: IADB. Trade in Services: Policy Issues.� Policy Research Working Paper 2852, Iida, Keyia, and Julia Nielson. 2003. “Strengthening Regulatory Trans- World Bank, Washington, DC. parency.� In Domestic Regulation and Service Trade Liberalization, Mattoo, Aaditya, and Pierre Sauvé. 2002. “Regionalism and Trade in Serv- ed. Aaditya Mattoo and Pierre Sauvé, 7–26. Washington, DC: Oxford ices in the Western Hemisphere: A Policy Agenda.� Prepared for forum University Press for the World Bank. on “FTAA and Beyond: Prospects for Integration in the Americas,� IMF (International Monetary Fund). 2009. Balance of Payments Manual. organized by the Inter-American Development Bank and Harvard 6th ed. Washington, DC: IMF. http://www.imf.org/external/pubs/ft/ University, Punta del Este, December 15–16. bop/2007/bopman6.htm. ———. 2003. “Domestic Regulation and the GATS: Looking Ahead.� In Kalirajan, Kaliappa. 2000. “Restrictions on Trade and Distribution Services.� Domestic Regulation and Service Trade Liberalization, ed. Aaditya Productivity Commission Staff Research Paper, AusInfo, Canberra. Mattoo and Pierre Sauvé. Washington, DC: Oxford University Press Kemp, Murray C., and Henry Y. Wan, Jr. 1976. “An Elementary Proposi- for the World Bank. tion Concerning the Formation of Customs Unions.� Journal of Inter- ———. 2008. “Regionalism in Services Trade.� In A Handbook of Interna- national Economics 6 (1): 95–97. tional Trade in Services, ed. Aaditya Mattoo, Robert M. Stern, and Kemp, Steven. 2000. “Trade in Education Services and the Impacts of Bar- Gianni Zanini. Oxford, U.K.: Oxford University Press. riers on Trade.� In Impediments to Trade in Services: Measurement and Mattoo, Aaditya, Randeep Rathindran, and Arvind Subramanian. 2001. Policy Implications, ed. Christopher Findlay and Tony Warren. London “Measuring Services Trade Liberalization and its Impact on Economic and New York: Routledge. Growth: An Illustration.� World Bank Policy Research Working Paper Kessides, Ioannis N., Roger G. Noll, and Nancy C. Benjamin. 2009. 2380, World Bank, Washington, DC. “Regionalizing Telecommunications Reform in West Africa.� Policy McGuire, Greg, and Michael Schuele. 2000. “Restrictiveness of Interna- Research Working Paper 5126, World Bank, Washington, DC. tional Trade in Banking Services.� In Impediments to Trade in Services: http://go.worldbank.org/31T8TCK3G0. Measurement and Policy Implications, ed. Christopher Findlay and Koncz, Jennifer, and Anne Flatness. 2008. “U.S. International Services: Tony Warren. London and New York: Routledge. Cross Border Trade in 2007 and Services Supplied through Affiliates in McGuire, Greg, Michael Schuele, and Tina Smith. 2000. “Restrictiveness 2006.� In Survey of Current Business (October), U.S. Department of of International Trade in Maritime Services.� In Impediments to Trade Commerce, Bureau of Economic Analysis, Washington, DC. in Services: Measurement and Policy Implications, ed. Christopher http://www.bea.gov/scb/pdf/2008/10%20October/services_text.pdf. Findlay and Tony Warren. London and New York: Routledge. Kowalczyk, Carsten. 1990. “Welfare and Customs Unions.� NBER Working McLaren, John. 1997. “Size, Sunk Costs, and Judge Bowker’s Objection to Paper 3476, National Bureau of Economic Research, Cambridge, MA. Free Trade.� American Economic Review 87 (3, June): 400–20. Lawrence, Robert Z. 1991. “Emerging Regional Arrangement: Building ———. 1999. “A Theory of Insidious Regionalism.� Department of Eco- Blocks or Stumbling Blocks?� In Finance and the International Econ- nomics, Columbia University, New York. omy 5: The AMEX Bank Review Prize Essays, ed. Richard O’Brien, McMillan, John. 1993. “Does Regional Integration Foster Open Trade? 23–35. New York: Oxford University Press. Economic Theory and GATT’s Article XXIV.� In Regional Integration Services 273 and the Global Trading System, ed. Kym Anderson and Richard Black- Nicole Pohl, and Pierre Sauvé, 434–65. Cambridge, U.K.: Cambridge hurst. London: Harvester Wheatsheaf. University Press. Mengozzi, Paolo. 1993. “Trade in Services and Commercial Policy.� In The Poretti, Pietro. 2008. “Waiting for Godot? Subsidy Disciplines in Serv- European Community’s Commercial Policy after 1992: The Legal ices Trade.� In GATS and the Regulation of International Trade in Dimension, ed. Marc Maresceau. Dordrecht: Martinus Nijhoff. Services, Regulation and Development, ed. Marion Panizzon, Nicole Messerlin, Patrick. 2009. “Economic Partnership Agreements: How to Pohl, and Pierre Sauvé, 466–88. Cambridge, U.K.: Cambridge Rebound?� In Updating Economic Partnership Agreements to Today’s University Press. Global Challenges, ed. Emily Jones and Marti F. Darlan, 22–27. Eco- Prieto, Francisco Javier. 2000. “The GATS, Subregional Agreements, and nomic Policy Paper Series 09. Washington, DC, and Brussels: the FTAA: How Much Is Left to be Done?� In Services Trade in the German Marshall Fund of the United States. http://www.gmfus.org/ Western Hemisphere: Liberalization, Integration, and Reform, ed. publications/index.cfm. Sherry M. Stephenson, 217–41. Washington, DC: Brookings Institu- Monge-Naranjo, Alexander. 2003. “Lessons from the NAFTA for Latin tion Press. America and the Caribbean.� World Bank, Washington, DC. Primo-Braga, Carlos Alberto. 1992. “NAFTA and the Rest of the World.� Nguyen-Hong, Duc. 2000. “Restrictions on Trade in Professional Services.� In North American Free Trade: Assessing the Impact, ed. Nora Lustig, Productivity Commission Staff Research Paper, AusInfo, Canberra. Barry Bosworth, and Robert Z. Lawrence, 210–49. Washington, DC: Nicolaïdis, Kalypso, and Joel P. Trachtman. 2000. “From Policed National Brookings Institution. Treatment to Managed Recognition: Mapping the Boundary in Qian, Ying. 2000. “Analysis of the Commitments under the GATS and the GATS.� In GATS 2000: New Directions in Services Trade Liberalization, World Trade Organization.� World Bank, Washington, DC. ed. Pierre Sauvé and Robert M. Stern, 241–82. Washington, DC: Robert, Maryse, Pierre Sauvé, and Karsten Steinfatt. 2001. “Negotiating Brookings Institution Press. Investment Rules: Possible Scenarios for an EU-MERCOSUL Agree- Nielson, Julia, and Daria Taglioni. 2004. “Services Trade Liberalisation: ment.� Institut d’Etudes Politiques, Paris. Identifying Opportunities and Gains.� OECD Trade Policy Working Robinson, Sherman, Zhi Wang, and Will Martin. 1999. “Capturing the Paper 1, Organisation for Economic Co-operation and Development, Implications of Services Trade Liberalization.� Presented at the Second Paris. http://www.oecd.org/officialdocuments/displaydocumentpdf/? Annual Conference on Global Economic Analysis, Ebberup, Denmark, cote=td/tc/wp%282003%2923/final&doclanguage=en. June 20–22. OAS (Organization of American States). 1998. Sectoral Agreements on Roy, Martin, Juan Marchetti, and Hor Lim. 2006. “Services Liberalization Services in the Western Hemisphere. Washington, DC: OAS. in the New Generation of Preferential Trade Agreements: How Much ———. 1999. “Provisions on Trade in Services in the Trade and Inte- Further than the GATS?� WTO Staff Working Paper ERSD-2006-07, gration Agreements of the Western Hemisphere.� OAS, Washington, World Trade Organization, Geneva. DC. ———. 2008. “The Race towards Preferential Trade Agreements in Serv- OECD (Organisation for Economic Co-operation and Development). ices: How Much Further Than the GATS?� In GATS and the Regulation 2002a. GATS: The Case for Open Services Markets. Paris: OECD. of International Trade in Services, ed. Marion Panizzon, Nicole Pohl, ———. 2002b. “Labor Mobility in PTAs.� TD/TC/WP(2002)16, OECD, and Pierre Sauvé, 77–110. Cambridge, U.K.: Cambridge University Paris. Press. ———. 2002c. “The Relationship between the Multilateral Trade System Salazar-Xirinachs, José M., and Maryse Robert. 2001. Toward Free Trade in and Regional Trade Agreements: Government Procurement.� TD/TC/ the Americas. Washington, DC: Brookings Institution Press. WP(2002)24, OECD, Paris. Sapir, Andre. 1998. “GATS 1994–2000.� Journal of World Trade 33: 51–66. ———. 2002d. “The Relationship between Regional Trade Agreements Sauvé, Pierre. 1996. “Services and the International Contestability of and the Multilateral Trading System: Investment.� TD/TC/WP(2002) Markets.� Transnational Corporations 5 (1, April): 37–56. 18, OECD, Paris. ———.1997. Preparing for Services 2000. CSI Occasional Paper 4. ———. 2002e. “The Relationship between Regional Trade Agreements Washington, DC: Coalition of Services Industries. and the Multilateral Trading System: Services.� TD/TC/WP(2002)21, ———. 2000. “Making Progress on Trade and Investment: Multilateral vs. OECD, Paris. Regional Perspectives.� In Services Trade in the Western Hemisphere: ———. 2002f. “Service Providers on the Move: Taking a Closer Look at Liberalization, Integration, and Reform, ed. Sherry M. Stephenson, Labor Mobility and the GATS.� OECD Trade Directorate, Paris. 72–85. Washington, DC: Brookings Institution Press. ———. 2003. “Quantifying the Benefits of Liberalising Trade in Services.� ———. 2002. “Completing the GATS Framework: Safeguards, Subsidies OECD, Paris. and Government Procurement.� In Development, Trade and the WTO: ———. 2007. “The Services Trade Restrictiveness Index: Issues and Chal- A Handbook, ed. Bernard Hoekman, Aaditya Mattoo, and Philip lenges in Its Design.� Meeting of the Experts on the STRI, English, 326–35. Washington, DC: World Bank. TAD/TC/SXM(2007)2, OECD, Paris. ———. 2003. “Assessing the General Agreement on Trade in Services: ———. 2009. “Assessing Barriers to Trade in Services in the MENA Half-Full or Half-Empty?� In Trade Rules behind Borders: Essays Region.� OECD Trade Policy Working Paper 84, TAD/TC/WP(2008) on Services, Investment and the New Trade Agenda, ed. Pierre Sauvé. 25/FINAL, OECD, Paris. London: Cameron May. Panagariya, Arvind. 2000. “Preferential Trade Liberalization: The Tradi- ———. 2009. “Trade and Investment in Services: An ADB-ITD Train- tional Theory and New Developments.� Journal of Economic Literature ing Module for the Greater Mekong Sub-Region.� Asian Develop- 38 (2): 287–331. ment Bank, Manila, and International Institute for Trade and Patterson, Gardner. 1966. Discrimination in International Trade: The Policy Development, Bangkok. http://www.adb.org/Documents/Guidelines/ Issues, 1945–1965. Princeton, NJ: Princeton University Press. Trade-Investment-services/default.asp. PECC (Pacific Economic Cooperation Council). 1995. “Survey of the ———. 2010. “Conducting a Trade-Related Regulatory Audit in Financial Impediments to Trade and Investment in the APEC Region.� PECC, Services.� In Financial Services and Preferential Trading Arrangements: Singapore. Lessons from Latin America, ed. Mona Haddad and Constantinos Peña, María-Angélica. 2000. “Services in MERCOSUR: The Protocol of Stephanou. Directions in Development series. Washington, DC: Montevideo.� In Services Trade in the Western Hemisphere: Liberalisa- World Bank. tion, Integration, and Reform, ed. Sherry M. Stephenson, 154–68. Sauvé, Pierre, and Denis Audet. 2009. “The Treatment of Trade in Services Washington, DC: Brookings Institution Press. in Economic Partnership Agreements: Implications for Prospective Pierola, Fernando. 2008. “A Safeguards Regime for Services.� In GATS and EPA Partners.� In Updating Economic Partnership Agreements to the Regulation of International Trade in Services, ed. Marion Panizzon, Today’s Global Economic Challenges, ed. Emily Jones and Darlan 274 Aaditya Mattoo and Pierre Sauvé F. Marti, 85–91. Economic Policy Paper Series 09, German Marshall ———. 2002. “Regional versus Multilateral Liberalization of Services.� Fund of the United States, Washington, DC and Brussels. http:// World Trade Review 1 (2): 187–209. www.scribd.com/doc/38477671/Updating-Economic-Partnership- Stephenson, Sherry M., and S. Yi. 2002. “Transparency in Regulation of Agreements-to-Today-s-Global-Challenges-Essays-on-the-Future- Services.� Presented to the meeting of the Asia-Pacific Economic of-Economic-Partnership-Agreements. Cooperation (APEC) Group on Services, Mexico City, February. Sauvé, Pierre, and Brenda Gonzalez-Hermosillo. 1993. “Implications of Thompson, Rachel. 2000. “Formula Approaches to Improving GATS the NAFTA for Canadian Financial Institutions.� In C. D. Howe Com- Commitments.� In GATS 2000: New Directions in Services Trade Liber- mentary 44 (April). Toronto, Canada: C. D. Howe Institute. alization, ed. Pierre Sauvé and Robert M. Stern, 473–86. Washington, Sauvé, Pierre, and Karsten Steinfatt. 2001. “Financial Services and the DC: Brookings Institution Press. WTO: What Next?� In Open Doors: Foreign Participation in Financial Tirole, Jean. 1998. The Theory of Industrial Organization. Cambridge, MA: Systems in Developing Countries, ed. Robert E. Litan, Paul Masson, and MIT Press. Michael Pomerleano, 351–86. Washington, DC: World Bank Group, Trachtman, Joel P. 2003. “Lessons for the GATS from Existing WTO Rules International Monetary Fund, and Brookings Institution Press. on Domestic Regulation.� In Domestic Regulation and Service Trade Sauvé, Pierre, and Natasha Ward. 2009a. “The EC–CARIFORUM Partner- Liberalization, ed. Aaditya Mattoo and Pierre Sauvé. Washington, DC: ship Agreement: Assessing the Outcome on Services and Investment.� Oxford University Press for the World Bank. ECIPE Discussion Paper. European Centre for International Political United Nations. 2002. Manual on Statistics of International Trade in Serv- Economy, Brussels. http://www.ecipe.org/publications/ecipe-working- ices. Geneva: United Nations. http://unstats.un.org/unsd/tradeserv/ papers/the-ec-cariform-economic-partnership-agreement-assessing- TFSITS/manual.htm. the-outcome-on-services-and-investment. United Nations and World Bank. 1993. Liberalizing International Transac- ———. 2009b. “Services and Investment in the EC–CARIFORUM Eco- tions in Services: A Handbook. Geneva: United Nations Conference on nomic Partnership Agreement: Innovation in Rule-Design and Impli- Trade and Development. cations for Africa.� In Beyond Market Access for Economic Develop- U.S. Coalition of Service Industries. 2003. “Reports of ISAC 13 on the ment: EU–Africa Relations in Transition, ed. G. Faber and J. Orbie, US–Chile and US–Singapore Free Trade Agreements.� http://www 137–72. London: Routledge. .uscsi.org. ———. 2009c. “Tourism Services Liberalization: The Role of Business Viner, Jacob. 1950. The Customs Union Issue. New York: Carnegie Endow- Advocacy in the Caribbean–European Union Economic Partnership ment for International Peace. Agreement.� Business and Trade Policy Case Study, International Trade Walmsely, T., and L. Alan Winters. 2002. “Relaxing the Restrictions on the Centre, Geneva. http://www.intracen.org/btp/business_advocacy/ Temporary Movement of Natural Persons: A Simulation Analysis.� publications/tourism_services_liberalization..pdf. University of Sussex, Brighton, U.K. Sauvé, Pierre, and Christopher Wilkie. 2000. “Investment Liberalization in Warren, Tony. 2000. “The Identification of Impediments to Trade and Invest- GATS.� In GATS 2000: New Directions in Services Trade Liberalization, ment in Telecommunications and Financial Services.� Impediments to ed. Pierre Sauvé and Robert M. Stern, 331–63. Washington, DC: Cen- Trade in Services: Measurement and Policy Implications, ed. Christopher tre for Business and Government, Harvard University, and Brookings Findlay and Tony Warren. London and New York: Routledge. Institution Press. Winters, L. Alan. 1996. “Regionalism versus Multilateralism.� Policy Schiff, Maurice, and W. Chang. 2003. “Market Presence, Contestability, Research Working Paper 1687, World Bank, Washington, DC. and the Terms-of-Trade Effects of Regional Integration.� Journal of ———. 2001. “Harnessing Trade for Development.� University of Sussex, International Economics 60 (1, May): 161–75. Brighton, U.K. Schott, Jeffrey, ed. 1989. Free Trade Areas and U.S. Trade Policy. Washing- Wonnacott, Paul, and Ronald J. Wonnacott. 1981. “Is Unilateral Tariff ton, DC: Institute for International Economics. Reduction Preferable to a Customs Union? The Curious Case of the Silver, Sara. 2002. “US under Pressure over Mexican Trucking.� Financial Missing Foreign Tariffs.� American Economic Review 71 (4, September): Times (November 4). 704–14. Snape, Richard. 1993. “History and Economics of GATT’s Article XXIV.� World Bank. 2000. Trade Blocs. New York: Oxford University Press. In Regional Integration and the Global Trading System, ed. Kym Ander- ———. 2002. Global Economic Prospects and the Developing Countries son and Richard Blackhurst. London: Harvester Wheatsheaf. 2002: Making Trade Work for the Poor. Washington, DC: World Bank. Snape, Richard, Jan Adams, and David Morgan. 1993. Regional Trading ———. 2003. Global Economic Prospects and the Developing Countries Arrangements: Implications and Options for Australia. Canberra: 2003: Investing to Unlock Global Opportunities. Washington, DC: Australian Government Publishing Service. World Bank. Snape, Richard, and Malcolm Bosworth. 1996. “Advancing Services Negotia- ———. 2004. Global Economic Prospects and the Developing Countries 2004: tions.� In The World Trading System: Challenges Ahead, ed. Jeffrey Schott, Realizing the Promise of the Doha Round. Washington, DC: World Bank. 185–203. Washington, DC: Institute for International Economics. WTO (World Trade Organization). 1995. Regionalism and the World Trad- Stephenson, Sherry M. 1999. “PTAs on Services within APEC and Disci- ing System. Geneva: WTO. plines of the Multilateral Trading System.� Presented to the meeting of ———. 1999. “The Fundamental WTO Principles of National Treatment, the Asia-Pacific Economic Cooperation (APEC) Group on Services, Most-Favored-Nation Treatment and Transparency.� WT/GTCP/ Wellington, February. W/114, WTO, Geneva. ———. 2000. “PTAs on Services and Multilateral Disciplines: Interpret- ———. 2001. Market Access: Unfinished Business—Post–Uruguay Round ing and Applying GATS Article V.� In GATS 2000: New Directions in Issues. Special Study 6. Geneva: WTO. Services Trade Liberalization, ed. Pierre Sauvé and Robert M. Stern. ———. 2003. Regional Trade Agreements, Geneva: WTO. Washington, DC: Brookings Institution Press. Wunsch-Vincent, Sacha, 2003. “The Digital Trade Agenda of the U.S.: Par- ———. 2001a. “Deepening Disciplines for Trade in Services.� OAS Trade allel Tracks of Bilateral, Regional and Multilateral Liberalization.� Unit Studies: Analyses on Trade and Integration in the Americas Aussenwirtschaft 58: 7–46. (March). Washington, DC: Organization of American States. ———. 2006. The WTO, the Internet and Digital Products: EC and US ———. 2001b. “Multilateral and Regional Services Liberalization by Latin Perspectives. Oxford, U.K.: Hart Publishing. America and the Caribbean.� OAS Trade Unit Studies, OEA/Ser.D/XXII, ———. 2008. “Trade Rules for the Digital Age.� In GATS and the Regulation SG/TU/TUS-9, Organization of American States, Washington, DC. of International Trade in Services, ed. Marion Panizzon, Nicole Pohl, and ———. 2001c. “Services.� In Toward Free Trade in the Americas, ed. José Pierre Sauvé, 497–529. Cambridge, U.K.: Cambridge University Press. M. Salazar-Xirinachs and Maryse Robert, 163–85. Washington, DC: Zoellick, Robert B. 2001. “Free Trade and the Hemispheric Hope.� Brookings Institution Press. Remarks to Council of the Americas, Washington, DC, May 7. 13 LABOR MOBILITY Sherry Stephenson and Gary Hufbauer Of all the liberalization dimensions of international institution has been designated as a repository of agree- trade, the opening of labor markets is probably the most ments and promoter of labor mobility. And whereas most sensitive. As a result, agreements to open (generally, only governments sponsor investment promotion agencies to partially) labor markets are not as plentiful as other liber- encourage inward flows of capital and foreign direct alization agreements, and they are typically more restric- investment, there is no similar institution for workers. In tive. They are also less well surveyed. This chapter presents nearly all countries, the agency that deals with the influx of an overview of provisions for opening labor markets that foreign labor is the immigration authority, whose concern are found in preferential trade agreements (PTAs). is to regulate and restrict, not to promote. To complicate Movement of labor is one of the four fundamental eco- matters further, immigration authorities are primarily nomic freedoms, along with free movement of goods, focused on setting rules for permanent rather than tempo- services, and capital. Of the four, it has met with the least rary migration. Temporary migration, however, is the receptivity on the part of countries in the international object of international trade policy and is the focus of this economy, whether developed or developing. Even the discussion. most spirited free traders—Chile; Singapore; the United The problems with the present situation are twofold. Kingdom; and Hong Kong SAR, China, for example— The first has to do with its unbalanced nature. Developed have been reticent about opening their borders more economies have a comparative advantage in the export of to admit labor from abroad. These economies and many capital and thus benefit tremendously from the openness others shy away from significant opening for natural per- of capital markets and the welcoming character of most sons from other countries, even in the face of labor short- investment regimes. That kind of receptiveness does not ages at home. exist for labor movements. Developing countries have a The contrast between the desire to promote capital comparative advantage in the export of their people, but mobility and investment flows and the reluctance to envis- they are constrained from realizing the gains from trade age corresponding labor mobility is stark. More than 2,800 that they might otherwise enjoy. bilateral investment treaties have been signed to date, but The second problem is that the entire world suffers nothing equivalent exists in the area of labor (Vis-Dunbar from a loss of potential income that could be realized and Nikiema 2009). The number of trade agreements cov- through greater mobility. Depending on what assump- ering services is growing rapidly, yet willingness to incor- tions are made by researchers, the potential gains could be porate meaningful provisions on labor mobility into the quite substantial and could easily surpass the combined services package is limited, and most agreements contain gains anticipated from freer trade in agriculture and man- very modest market access opportunities for foreign ufactured goods—as currently proposed in the Doha workers. Several recent free trade agreements (FTAs) con- Development Round sponsored by the World Trade tain no provisions at all in this area. Leaving aside formal Organization (WTO). Nevertheless, proposals for greater bilateral and regional trade agreements, it is extremely dif- market access for foreign workers are limited, and this has ficult to determine the number of bilateral agreements been a central obstacle to progress in the services compo- worldwide that incorporate arrangements for temporary nent of Doha Round negotiations. worker programs. There appears, however, to be not It has been argued by many, including, prominently, more than a handful. Information is scarce, and no single Lant Pritchett and L. Alan Winters, that greater mobility of 275 276 Sherry Stephenson and Gary Hufbauer labor would be the first-best development promotion strat- The Concept of Labor Mobility egy (Pritchett 2006; Winters 2008). Pritchett writes that it is In international services trade, labor mobility is conceptu- hard to imagine a policy more directly at odds with poverty alized as the temporary movement of natural persons and reduction or pro-poor growth objectives than one limiting is categorized as mode 4. Article I.2 (d) of the WTO Gen- the demand for lower-skilled labor. This limitation can be eral Agreement on Trade in Services (GATS) defines mode viewed as the principal way that rich countries are cur- 4 as the supply of a service “by a service supplier of one rently inhibiting the development possibilities of poorer Member, through presence of natural persons of a Member countries—much more than through restrictive agricul- in the territory of any other Member.� (See box 13.1 for tural policies or nontariff barriers. estimates of the size of this mode.) A natural person of In such a challenging and often hostile environment another member is defined as for labor mobility, what options might developing coun- tries have for increasing the scope for the movement of a natural person who resides in the territory of that other their workers? Given the impasse in the Doha Round and Member or any other Member, and who under the law of the lack of any progress on services in multilateral nego- that other Member: tiations for the past several years, preferential trade (i) is a national of that other Member; or agreements might offer a more promising channel for (ii) has the right of permanent residence in that other greater labor mobility, even if among a limited number Member . . .� (Article XVIII[k]) of partners. Other options that have been relatively unexplored to date may also be available, such as the Temporary versus Permanent Workers promotion of circular migration through temporary worker agreements—time-bound instruments that allow For services trade, and for our purposes here, labor mobil- greater flexibility for both labor-sending and labor- ity is understood as the movement of workers to carry out receiving countries. employment in another country for a time-limited period. In this chapter, we do not delve into the Doha quagmire Although the term “temporary� is not defined under but, rather, explore these other options. We first discuss the GATS, the notion of moving in order to work for a limited concept and magnitude of labor mobility and the potential period, as opposed to moving with the intention of emi- benefits from greater liberalization. We then review the grating permanently, is what distinguishes mode 4. This is various ways in which members of PTAs have treated the affirmed in the GATS “Annex on Movement of Natural issue of labor mobility to assess whether these preferential Persons Supplying Services under the Agreement,� which agreements have effectively promoted temporary entry. specifies that GATS shall not apply “to measures affecting Only PTAs between developed and developing countries natural persons seeking access to the employment market are examined. of a Member, nor shall it apply to measures regarding citi- The questions we attempt to answer are the following: zenship, residence or employment on a permanent basis.� Which developed countries are more amenable to greater All subsequent trade agreements, following the WTO openness for natural persons, and what are the possible approach, consider only the temporary movement of reasons? Can recent PTAs be emulated? What have tem- workers, but governments have been unwilling to define in porary worker programs in bilateral and plurilateral precise terms what period of time is meant by “temporary.� agreements with developing-country partners achieved, The official statistical definition of temporary migration is and could such agreements usefully supplement the a stay of less than one year, but for trade policy purposes, a PTA approach? After exploring these issues, we look at temporary stay can vary anywhere from a few weeks to a policy suggestions that might be implemented by the few years, depending on the commitments governments World Bank to promote labor mobility for its developing- are prepared to undertake. This lack of precision has been country members. both a strength and a weakness in defining mode 4 treat- We recognize that developed countries are currently ment within trade agreements. experiencing very high levels of unemployment and that The great political sensitivity surrounding interna- these conditions will probably continue through 2011 tional labor mobility is not helped by the very frequent and possibly into 2012. Accordingly, political resistance confusion, in both statistical analysis and political to all forms of labor mobility is extremely high. This debate, between temporary and permanent migration. chapter, however, is written with a view to the longer Immigration authorities deal with both simultaneously, term, when normal economic conditions will have been and at times the character of “temporary� labor movement restored. is disregarded and all immigrants are treated as though Labor Mobility 277 Box 13.1. Labor Mobility in Statistical Terms Historically, one of the main ways that temporary labor migration has been captured in the data is through recorded “transfers and payments� in balance of payments statistics; this category is what we and others term “remittances.� According to the International Monetary Fund (IMF) Balance of Payments Manual, “remittances� mainly comprise “compensation of employees� and “personal transfers� (IMF 2010). Transactions are recorded in the balance of payments when money is paid by residents to nonresidents, or vice versa. Determining the magnitude of mode 4 (temporary movement of services providers) by examining transactions is problematic, since remittances are made by both temporary and permanent migrants and by workers in the manufacturing and agricultural segments of the economy, as well as in services sectors. In addition, payments to undocumented foreign workers are not captured statistically when they are spent in the country in which the person works (see also Carzaniga 2008). Looking at volume rather than value is hardly more satisfactory. The Organisation for Economic Co-operation and Development (OECD) collects information on the number of temporary foreign workers in surveys of firms, visas, border crossings, and so on. These data cover intracorporate transferees and other temporary workers. They are, however, far from a perfect match for mode 4. Furthermore, the scope may differ from one country to another, and in any event, figures are only available for a subgroup of OECD members. Karsenty (2000) calculated services trade through mode 4 to be no more than 1 to 2 percent of total two-way trade in services. Applying this range to the most recent trade data available (2008) would put the value of services trade through mode 4 somewhere between US$70 billion and US$150 billion annually. Those figures may underestimate the actual value of this form of services trade, since remittances alone in 2007 amounted to more than US$200 billion. Remittances, however, likely overestimate mode 4 services trade. Source: Karsenty 2000; OECD data. they were seeking permanent status. Moreover, inside the ual in a host country. In most trade agreements, these host country, the line between permanently and temporar- have been limited to professional workers, but commit- ily resident migrants often becomes blurred. ments can also be extended to lower-skilled categories of workers. • Contractual services suppliers (CSSs). Employees of a Categories of Labor Included in Trade Agreements foreign services company with no local presence or Although the GATS text does not define specific categories commercial presence in the host country who are of labor, WTO members have accepted four widely used engaged under contract to provide a service to a firm in categories for the purpose of inscribing commitments the destination country.1 under mode 4. These categories are not comprehensive, as they cover only skilled professionals. In a few recent trade Developing countries’ interest in promoting greater agreements, as we will see in the next section, countries labor mobility most concerns the independent profes- have begun to move beyond this limited range of cate- sional and contractual services supplier categories, gories to broaden their consideration of labor categories rather than employees of multinational corporations for market access. The four traditional mode 4 categories (MNCs). This is because most developing countries, are the following: with notable exceptions such as Brazil, China, and India, have not yet become home bases for MNCs. Greater flex- • Business visitors and salespersons (BVs). Foreign nation- ibility in the independent professional and contractual als who travel abroad to negotiate the sale of a service services supplier categories would allow most developing or to explore the possibility of making a foreign direct countries to send a larger number of professionals investment (of establishing a commercial presence, abroad for temporary employment. The business visitor in GATS terminology) for their company in the desti- and intracorporate transferee categories are of interest nation country. Their main purpose is to facilitate to successful emerging countries such as Brazil, China, future transactions rather than actually to carry out and India. transactions. • Intracorporate transferees (ICTs). Employees of a foreign Potential Economic Gains from Greater services company that has set up a commercial presence Labor Mobility abroad and that transfers these employees to its foreign location. The fact that we cannot accurately know the real statisti- • Independent professionals (IPs). Self-employed persons cal importance of temporary workers in the world econ- who are supplying a service to a company or an individ- omy is secondary to the fact that impediments to labor 278 Sherry Stephenson and Gary Hufbauer mobility suppress trade to the disadvantage of everyone, for native workers is shown by area ACDE. The gain for but particularly developing countries. Goods move freely capitalists is shown by area EABD, with most of this gain in response to price differentials, and capital flows effort- coming from the loss for native workers. Since the gain for lessly around the globe in response to profit and interest capitalists is larger than the loss for native workers, the lib- rate differentials, but workers are not allowed to move eralization of mode 4 leads to an overall gain, shown by readily in response to wage differentials. Consequently, area ABC. very large wage differentials exist in the world today, as Effect on developing countries. The effect of the liberal- shown in figure 13.1. The benefit to be derived from the ization of mode 4 on developing countries is the exact exploitation of comparative advantage is directly propor- opposite to that for developed countries. With restrictions tional to the size of wage, price, or profit differences prior on mode 4, the equilibrium in the labor market is at point to trade or investment liberalization; thus, considerable B in figure 13.3. After liberalization, the equilibrium point gains could be realized if workers were permitted to moves to point A, reflecting an increase in the wage per exploit wage differentials among countries. (See box 13.2 hour and a decrease in the number of hours worked. for estimates of these gains.) As will be apparent later, the gains for migrants in developed countries are much larger than the loss that their departure inflicts on developing countries. Nonmi- Theoretical Model of the Distributional Effects of grant workers also experience gains, shown by area ACDE Mode 4 Liberalization in figure 13.3, since the wage rate has increased in devel- Like trade in goods, labor mobility can create losers as oping countries. But nonmigrant capitalists experience a well as winners. In the overall balance, gains usually very large loss, shown by area ABDE (most of the loss cor- exceed losses by a wide margin, but political sensitivities responds to the wage gain for nonmigrant workers). focus on those who lose. In simple theoretical terms, Because the loss for nonmigrant capitalists is larger than migration can be modeled as an increase of supply in the the gain for nonmigrant workers, the group of nonmi- labor markets of developed countries and a decrease of grants as a whole experiences an overall loss of income, supply in developing countries. Here, we use that frame- shown by area ABC. In other words, the effect on total work to examine the effects of those supply changes on welfare of liberalizing mode 4 is negative for nonmigrants the incomes of capitalists and workers, in both the send- in developing countries. Income per capita, however, is ing and the host countries, and on the incomes of the likely (although not guaranteed) to rise as marginal pro- migrants themselves. ductivity increases. Effect in developed countries. Given the restrictions on Overall outcome. Migrants lose their erstwhile wages in labor mobility, the equilibrium in the labor market is at developing countries but enjoy larger wages in developed point A in figure 13.2. After liberalization, the equilibrium countries. They therefore experience a gain, measured by moves to point B, reflecting an increase in the number of the wage difference between the destination and source hours worked and a decrease in the wage per hour. The loss countries. Figure 13.1. Theoretical Gains from Liberalization of Mode 4 restrictions on labor mobility productivity/ productivity/ wages in wages in developed developing countries countries wage potential gain differential number of workers in number of workers in developed countries number of developing countries migrants Labor Mobility 279 Box 13.2. Quantitative Estimates of Overall Gains from Greater Labor Mobility Complete liberalization of mode 4 would result in very large gains. Hamilton and Whalley (1984) use a partial equilibrium (PE) model and 1977 data to estimate the benefits from the complete elimination of all immigration restrictions, for skilled and unskilled labor alike. The potential gains are enormous, ranging from 60 to almost 205 percent of world gross domestic product (GDP). Millions of workers would move from low-productivity to high- productivity jobs in countries with high salaries, until wages in labor-sending and labor-receiving countries equalized. Iregui (1999) revisits the question using a computable general equilibrium (CGE) model and more precise measures of elasticities and population characteristics. Here again, the gains are large, ranging from 15 to 67 percent of world GDP. Moses and Letnes (2004), using more precise values for productivities, confirm large gains, ranging from 4.3 to about 112 percent of world GDP in 1977. According to these authors, the ‘’most reasonable’’ gain would be 7.5 percent of world GDP. The large differences between these estimates, both within and between studies, can be explained by the differences in modeling frameworks (partial versus general equilibrium) and assumed parameters. Some estimates assume that migrants can achieve the average productivity of workers in the destination country; others assume that additional education and training will be needed. Gains from less than complete liberalization of mode 4 are still large. Because full liberalization is politically unacceptable, some economists have estimated the potential outcome of more modest liberalization of mode 4. Moses and Letnes (2004) estimate the gains from eliminating 10 percent of the wage inequality between countries and find that potential gains would still be large, corresponding to around 2.2 percent of world GDP. Walmsley and Winters (2002) estimate the potential gain from a 3 percent increase in the workforce in developed countries, a movement of 14.2 million workers, and a 50 percent increase in the current number of immigrants in developed countries at US$156 billion in 2002, representing 0.6 percent of world GDP. World Bank (2006) reaches a very similar result. Most of the gains come from the movement of unskilled labor. According to Iregui (1999), the potential gains from the migration of skilled labor only are much smaller: 3 to 11 percent of world GDP, in comparison with 13 to 59 percent for all skills. Walmsley and Winters (2002) show that the potential gain from the movement of unskilled workers would account for US$110 billion, or 70 percent of the total. This reflects the fact that inequality in wages worldwide is larger for unskilled than for skilled workers. Source: Annex table 13A.1. Figure 13.2. Theoretical Effect on Developed Figure 13.3. Theoretical Effect on Developing Countries of Liberalization of Mode 4 Countries of Liberalization of Mode 4 wages wages supply supply (with restrictions) (liberalization) supply supply (liberalization) (with restrictions) A A E E loss for native workers overall gain for nonmigrant overall gain for natives workers loss for nonmigrants D D C B C B demand demand number of hours worked number of hours worked According to the theoretical model, the liberalization of • In developing countries, most of the losses to capitalists mode 4 has the following distributional consequences: are mirrored by gains to nonmigrant workers. • In developed countries, the gains for capitalists are • In developed countries, most of the gains for capitalists larger than the losses for native workers. Therefore, total are balanced by losses to native workers. income in developed countries rises. 280 Sherry Stephenson and Gary Hufbauer • In developing countries, the losses for capitalists are between 1990 and 2008, rising from US$69 billion to larger than the gains for nonmigrant workers. There- US$397 billion (adjusted for inflation). In 2007, migrant fore, total income in developing countries falls. compensation and remittances accounted for around 0.7 percent of world GDP, but for developing countries, the relative importance of remittances in GDP in 2007 was Distributional Effects of Mode 4 Liberalization much higher. Remittances were 2.1 percent of the GDP of The theoretical and empirical prediction of large gains developing countries as a whole, but 1.9 percent of the from full or partial liberalization of mode 4 outlined in GDP of middle-income countries and 5.8 percent of the box 13.2 do not hide the fact that labor mobility will have GDP of the least-developed countries (a UN category). distributional consequences. Migrants are the main win- An increasing share of remittances goes to developing ners; the results for natives in both the sending and the host countries, which accounted for 46 percent of this flow in countries are mixed. 1990 but for 76 percent by 2007. It is estimated that remit- Gains for migrants. Walmsley and Winters (2002) calcu- tances touch 1 in 10 people worldwide. Dependence on late that benefits to migrants (US$171 billion) actually remittances is especially high in certain countries. The account for more than the total gain from increased labor main receiving countries in absolute terms are India mobility (US$156 billion). Total gains are smaller than the (US$27 billion), China (US$26 billion), Mexico (US$25 gains to migrants because of the losses to the sending billion), and the Philippines (US$17 billion). For many countries, discussed below. smaller countries, remittances represent a very large frac- Losses for developing countries, before remittances. The tion of GDP, accounting for more than 36 percent of the departure of migrants reduces the number of workers in GDPs of Moldova and Tajikistan and about 25 percent of the sending countries, which increases hourly wages of the GDPs of Guyana, Honduras, and Lesotho. nonmigrant workers but diminishes total output. Walms- Mixed picture in developed countries. Outcomes of ley and Winters (2002) calculate that Brazil would see its migration for the developed countries are mixed, although welfare reduced by US$7 billion if the workforce going to slightly positive. Workers, especially unskilled ones, face developed countries increased by 3 percent, and China increased competition from migrants and see their wages would experience a decline of US$2 billion, notwith- decline. For example, Hatton and Williamson (1998) esti- standing the compensation received from remittances. mate that in 1910, American wages would have been 11 to The authors’ calculations suggest that unskilled workers 14 percent higher in the absence of the immigration wave in India would see a wage increase of 0.7 percent and that that set in after 1870. Borjas (1999) calculates that immi- skilled workers in Mexico would enjoy an increase of gration to the United States between 1980 and 1998 4.5 percent. Returns to capital would, however, decrease resulted in a decrease in native wages amounting to by, for example, 0.4 percent in Mexico. Exploring a more 1.9 percent of GDP and that the losses were concentrated extreme scenario, Moses and Letnes (2004) arrive at simi- among low-skilled U.S. workers, whereas skilled workers lar results. In their calculations, a 10 percent elimination actually benefited from immigration. Immigration of wage inequality leads to an 11.4 percent increase in the reduced the wages of native high-school dropouts in the wages of nonmigrant workers in the poorest countries in United States by 8.9 percent between 1980 and 2000 but 1998, while the return to capital in those countries falls increased the return to capital by 2 percent of GDP. The like a stone, by 21 percent. net gain from the 1980–98 migration wave for all U.S. The importance of remittances for developing countries. natives is the difference between the decrease in wages If the gains to migrants themselves are included in the and the increase in returns to capital, or 0.1 percent of overall balance sheet for developing countries, the pic- U.S. GDP per year over the period. This net gain repre- ture changes completely. (Pritchett 2006 makes this sents about US$10 billion a year, accounting for about point.) When the gains to migrants are combined with 5 percent of U.S. economic growth over a 20-year period. the national income losses to the sending countries, the Moses and Letnes (2004) find the same pattern in the developing countries experience a significant gain in case of a 10 percent elimination of wage inequality. They plausible scenarios—the equivalent of 1.8 percent of their calculate that liberalization of this magnitude would gross domestic product (GDP), according to the World reduce wages in developed countries by 3.1 percent, Bank’s Global Economic Prospects 2006, which explores while increasing the return to capital by 7.2 percent. the “3 percent scenario.� Walmsley and Winters (2002) reach similar results in the World Bank estimates of global remittances show that case of a 3 percent increase in the workforce of devel- globally, compensation and remittances increased sixfold oped countries: that scenario leads to a 0.8 percent Labor Mobility 281 decrease in U.S. and European wages and a 0.8 percent (CARIFORUM) do permit limited access. Japanese PTAs increase in return to capital in the United States. The World allow the usual professional categories and contain innova- Bank’s Global Economic Prospects 2006 study shows that in tive provisions for semiskilled workers. Australia and New the 3 percent scenario, the incomes of all natives combined Zealand have negotiated highly innovative agreements in developed countries would rise by 0.4 percent (World with China and Chile. By contrast, the Trans-Pacific Strate- Bank 2006). gic Economic Partnership Agreement between Brunei Darussalam, Chile, New Zealand, and Singapore is quite restrictive. This feature may ease future accession by Aus- Labor Mobility in Preferential Trade tralia, Japan, the United States, and Vietnam, but it does Agreements nothing for labor mobility among the Pacific members. A stalemate on services liberalization at the multilateral level has clouded the Doha Round for the past several PTAs Negotiated by the United States and Canada years, with no moves to put new services offers on the table or to improve existing ones made since the end of NAFTA was the pioneer agreement and template for 2005.2 In contrast, an increasing amount of activity has many subsequent PTAs. It contains a chapter entitled taken place at the regional level, with the negotiation of “Temporary Movement of Business Persons,� designed to numerous PTAs, a large number of which have incorpo- facilitate temporary entry to member countries for rated mode 4 as part of the package.3 business people involved in goods or services trade or The members of some recent PTAs have accepted in investment activities. The categories defined under greater labor mobility at the regional level. Although NAFTA are traders and investors, business visitors, intra- progress is still relatively modest, interesting initiatives corporate transferees, and professionals. There is no limit have been taken. Developing countries in the Americas on the number of visas for business visitors, and a work and in Asia have entered into several free trade agree- permit is not required. According to Martin and Lowell ments that contain provisions to facilitate procedures (2008), the novel migration component of NAFTA is the for temporary labor movement and open up market Trade NAFTA, or TN, visa (see Stephenson 2007 for a access opportunities. Some agreements include guaran- summary). This visa was uncapped in 1994 for Canadians teed numerical quotas for certain categories of skilled and has been uncapped for Mexicans since 2004. When labor. In this section, we compare approaches to labor proof of a job offer is demonstrated, the TN visa permits mobility in PTAs. Only PTAs between developed and employment for one year, with unlimited renewal. developing economies, and only those negotiated since the In addition to the chapter on temporary entry, NAFTA, entry into force in January 1994 of the North American like subsequent agreements with a similar structure, Free Trade Agreement (NAFTA), which signaled an era of contains an annex on professionals that is specifically deeper and more comprehensive preferential trade agree- targeted at professional services suppliers. These annexes ments, are considered. are intended to promote the development of mutually Our discussion is divided along geographic lines, distin- acceptable standards and criteria for licensing and certi- guishing between PTAs negotiated by the United States, fication of professional services suppliers, on the basis of Canada, the European Union (EU), Japan, and Australia factors such as educational background, qualifying and New Zealand. Annex tables 13A.2–13A.6 summarize examinations, and experience. In addition, the NAFTA the salient provisions of the PTAs that are examined. annex encourages members to submit recommendations Before delving into the details, a broad overview may for furthering the process of mutual recognition. A qual- be useful. NAFTA and other first-generation U.S. PTAs ifying list of 62 professions is set out in an appendix to allowed limited mobility for professional workers, but sec- the agreement; applicants must fulfill the necessary qual- ond-generation U.S. PTAs are quite restrictive, as a result ification requirements. The United States originally of congressional opposition. Canadian PTAs are much placed a quota of 5,500 per year on the number of pro- more liberal for skilled workers, as well as professionals, fessionals who could be admitted from Mexico, but that and contain some innovative provisions. Early EU agree- quota has been eliminated. ments with developing countries did not have provisions Besides NAFTA, the United States has negotiated several allowing labor mobility because the subject was reserved other bilateral free trade agreements with developing coun- to the competence of member states, but more recent EU tries. The agreements selected for examination are those agreements with Chile and the countries of the Caribbean with Chile, the five-country Dominican Republic–Central Forum of African, Caribbean, and Pacific (ACP) States America Free Trade Agreement (CAFTA–DR), Morocco, 282 Sherry Stephenson and Gary Hufbauer Peru, and Singapore (annex table 13A.2). Bilateral agree- negotiate greater labor mobility in trade agreements with ments with Colombia, the Republic of Korea, and Panama the United States. have been finalized but are awaiting ratification by the U.S. In Canada, the situation has evolved in the opposite Congress. direction (see annex table 13A.3). Interestingly, and per- Under the agreements with Chile and Singapore, both haps because of pressures from the private sector and of which were concluded in 2002 and entered into force apparent labor shortages in the Canadian market prior in 2004, labor mobility was expanded slightly for profes- to the current economic crisis, the government has negoti- sional workers, and a path to a special visa for profession- ated recent FTAs that go quite far toward providing als (the H-1B1 visa) was created. The visa provided for an increased access not only for professionals but also for semi- initial stay of 18 months, but with unlimited extensions. skilled foreign workers. Although the FTA that Canada In addition, an annual quota of 1,400 visas for profes- negotiated with Chile in 1997 looks very much like NAFTA, sionals from Chile was granted, as was an annual quota of with the only categories of workers covered being investors, 5,400 visas for professionals from Singapore, on top of traders, business visitors, intracorporate transferees, and the fixed total of H-1B visas from all countries. The new professionals, it is notable in that no numerical limits were visa category created under these agreements is meant for placed on 72 of these categories of professional labor. temporary migrants, for stays of up to 18 months ini- Strikingly, the two very recent FTAs negotiated by tially, but with the possibility of unlimited extensions. Canada with Colombia (2008) and Peru (2009) go much In brief, the current provisions governing labor move- farther. They cover all professional categories, with no ment to the United States under NAFTA and the agree- numerical limits and no specified length of stay, meaning ments with Chile and Singapore are as follows: that visas could in theory be renewed indefinitely. For the first time, they also expand coverage of worker categories • NAFTA: TN visa; uncapped for both Canadians and beyond highly trained professionals to include “techni- Mexicans cians.� In both the Colombia and Peru FTAs, Canada has • Chile FTA: H-1B1 visa; capped at 1,400 professionals listed 50 categories of technicians to be admitted into the • Singapore FTA: H-1B1 visa; capped at 5,400 profes- Canadian market with no specified length of stay. These sionals. workers must have a high school degree, with two years of technical training. Technician categories include, As mentioned, these visa numbers are additional among others, mechanics, construction inspectors, food to whatever entries occur under other visa categories, and beverage supervisors, chefs, plumbers, and oil and most importantly, the H-1B visa for skilled workers and gas well drillers. This recent development constitutes a professionals. major step forward for the expansion of temporary entry Unfortunately, the opposition of the U.S. Congress to in trade agreements. these arrangements, and in particular to the agreements with Chile and Singapore, was loud and clear. Key mem- PTAs Negotiated by the European Union bers of Congress objected that the trade agreements had encroached on the realm of immigration matters. As a In this section, we examine PTAs between the EU and third consequence of this outcry, no free trade agreement nego- countries (see annex table 13A.4) Total labor mobility is tiated by the United States since 2002 has contained guaranteed within the EU itself, although only after 10 a chapter to facilitate the temporary movement of skilled years for some of the newest members. workers.4 Thus, the FTAs with Morocco, with the The form of PTAs negotiated by the EU differs from CAFTA–DR members, and with Peru, like those negoti- that pioneered by the United States. Provisions for services ated with Colombia, Korea, and Panama, contain no and investment liberalization are set out in a title (or sec- chapter on temporary entry. Each does contain an annex tion) of the EU agreements entitled “Trade in Services and on professionals, with objectives similar to those set out Establishment.� The European Commission does not as yet in NAFTA, but these annexes explicitly state that “no pro- have negotiating authority from the EU member states in vision shall impose any obligation on a party regarding all service areas.5 The European Commission consequently its immigration measures,� and they contain no market always follows a positive-list approach in its trade agree- access commitments. Thus, public and official attitudes ments, with lists of commitments attached to the main text in the United States with respect to labor mobility have of the agreements. Thus, in terms of market access, mode 4 regressed since 2002. Until political opinion changes, it is brought within the scope of EU PTAs in a way similar to will be close to impossible for developing countries to that followed under GATS. Categories of workers included Labor Mobility 283 in mode 4 commitments by the EU include the four that In the annex on professionals attached to the CARIFO- are traditional for PTAs: traders and investors, business vis- RUM agreement, the European Union committed to accept itors, intracorporate transferees, and independent profes- 29 categories of professional services providers without sionals. The EU has negotiated relatively few PTAs with numerical limit, provided that they have a university developing countries that cover services. Although it has degree and three years’ experience. The CARIFORUM numerous association agreements in place with neighbor- members did not commit reciprocally to accept any EU ing Mediterranean countries (the Arab Republic of Egypt, professionals. Jordan, Morocco, the Syrian Arab Republic, Tunisia, Turkey, and others), these agreements focus on goods and PTAs Negotiated by Japan have not yet incorporated services provisions. The EU has negotiated association agreements with Japan has negotiated four PTAs that are of interest for the Mexico and Chile and has more recently finalized an eco- question of labor mobility. These are summarized in nomic partnership agreement (EPA) with CARIFORUM. annex table 13A.5. The PTAs with Mexico and Chile are There are no in-depth services provisions in the EU very similar in form and content to the NAFTA-type agreement with Mexico, which was concluded in March approach and agreements, with a negative list of noncon- 2000 when the GATS negotiations were just beginning, forming measures and with mode 4 treated in a chapter but the PTA with Chile is substantial. In addition to the on temporary movement of business persons. That chap- coverage of mode 4 in the text of the agreement, there is ter defines movement for the same four categories usually a specific article entitled “Movement of Natural Persons� seen in trade agreements: traders and investors, business in the EU–Chile association agreement, as well as an visitors, intracorporate transferees, and independent pro- annex on professionals.6 In the annex, the EU specifies fessionals. Japan has set a time limit of three years for 33 categories of professional services providers that it three of these categories (all except business visitors), will accept from Chile without numerical limit, for a which is a fairly generous interpretation of length of stay. time period of three months, subject to the “necessary The two more recent PTAs negotiated by Japan with academic qualification and experience.� Interestingly, countries in Southeast Asia—those with Indonesia and Chile did not commit reciprocally to accepting any pro- the Philippines—are notable for their innovations in cov- fessionals from the EU. ering, for the first time, specific categories of nurses and The more recent EU–CARIFORUM economic partner- health care workers. These PTAs have also expanded the ship agreement follows a similar structure, but, in addi- categories of workers in the chapter on mode 4 to include tion to the usual categories of workers defined under “professionals with personal contracts� (essentially, the mode 4, the EU has expanded coverage of workers to three same as independent professionals). All these categories additional categories important for CARIFORUM mem- (except for business visitors) are allowed a stay of up to bers: contractual services suppliers, independent profes- three years. Japan has also increased the number of pro- sionals, and graduate trainees. The following applies to fessional categories covered in the annex on professionals these three categories:7 in these two PTAs, to 14 in the case of Indonesia and 10 for the Philippines. No numerical limits are placed • Contractual services suppliers. This category applies to a on these professional categories, except for nurses and specific list of activities and permits temporary entry for health care workers, for whom an annual quota is in a cumulative period of six months. A contractual serv- effect. For those professions, specific educational and ices supplier must fulfill certain requirements; the terms training requirements are included in the agreements: a and conditions are set out in EU schedules for its mem- health degree plus two years of prior work experience and ber states. six months of language training in Japanese. The specifi- • Independent professionals. The provisions for the CSS cation of particular categories of work with annual quo- category also apply to independent professionals, again, tas and training requirements is an innovative approach subject to EU schedules. that has not yet been seen in other PTAs. • Graduate trainees. Graduate trainees, a new category, are workers from CARIFORUM states who have a uni- PTAs Negotiated by Australia and New Zealand versity degree and are temporarily transferred to the parent company or to a commercial establishment for Four PTAs negotiated by Australia and New Zealand are career development or to obtain training in business relevant for this study (annex table 13A.6). One PTA was methods. They may enter for a period of up to one year. negotiated jointly by the two countries with the 10-member 284 Sherry Stephenson and Gary Hufbauer Association of Southeast Asian Nations (ASEAN). Three includes the four usual categories of labor plus the addi- of the four are very recent, having been signed or having tional category of “installers� added by New Zealand. The entered into force since the end of August 2008. The oldest length of stay offered by the partners to the PTA is vari- of the four, and the one with the least ambitious provi- able, with Australia and New Zealand allowing the longest sions for labor mobility, is the Trans-Pacific Strategic stay, of three or four years, respectively, for intracorporate Economic Partnership Agreement between Brunei transferees and one year for independent professionals Darussalam, Chile, New Zealand, and Singapore. This and contractual services suppliers. It is notable that the agreement, which entered into force in 2006, follows a ASEAN members committed to much less generous dura- NAFTA-type structure, but with lighter content. The tions of stay for all labor categories than did their devel- only category of workers specified in the temporary oped partners. In another innovative decision, Australia entry (or labor mobility) chapter is that of professionals, also included “spouses� in its categories of temporary and no length of stay is specified. The annex on profes- labor permitted entry. sional services primarily sets out a best-endeavors clause for the development of “mutually acceptable standards Comparison and Assessment of PTAs and criteria for licensing and certification of professional service providers.� No professional categories of services An overall comparison of the PTAs negotiated with devel- providers are listed, and so the annex has no market oping countries by the United States, Canada, the EU, access component. Japan, and Australia and New Zealand is shown in the bar In the New Zealand–China PTA that entered into force graphs set out in annex figures 13A.1–13A.6. The bars indi- in October 2008, the chapter on labor mobility specifies cate the relative magnitude of the developed economies’ five categories of labor: business visitors, investors, intra- commitments on labor mobility. The higher the bar for a corporate transferees, contractual services supplies, and a particular agreement, the more access it provides for work- new category of “installers.� The CSS category includes ers from its developing-country partner. The number of artisans with Chinese cultural expertise such as theater categories of workers encompassed within the chapter on artists, Mandarin language teachers, and Chinese medical temporary entry or movement of natural persons is indi- specialists. China made no commitment with regard to cated, as well as the number of professionals allowed, professional service providers; New Zealand allows entry under specified quotas or without numerical limitation, of designated professionals from China for up to three through the annex on professionals. years. Intracorporate transferees from China are also per- In general, it can be said that trade agreements con- mitted a three-year stay. The new category of installers is cluded by developed countries with developing countries allowed a three-month stay. focus almost exclusively on professional services providers. The Australia–Chile PTA that entered into force in Many, however, go well beyond GATS in providing access March 2009 follows a NAFTA-type structure, and the chap- for a greater number of categories of professional services, ter on temporary entry specifies the four usual categories through expanded numbers of covered categories or of labor. An annex on professional services does not through the provision of unlimited access. They also often include a market access component, and no numbers are offer the possibility of long-term visa renewals once profes- attached to any category of worker. Australia allows intra- sionals are settled in the country. Thus, distinct progress corporate transferees a stay of up to four years and con- has been made with respect to professional services. tractual services suppliers, a stay of one year, with the pos- A few developed countries have been willing very sibility of renewal. This recent PTA is quite original in its recently to go beyond the expansion of access for profes- treatment of spouses and accompanying family members; sional services providers. These include, notably, Canada, they are granted the right to join the worker after he or she in two recent FTAs negotiated with Colombia and Peru has been in Australia for more than one year. Dependents that extend access to the Canadian market to 50 categories and spouses of corporate executives, intracorporate trans- of technicians. The innovative group also includes the ferees, and contractual services suppliers from the other EU’s EPA with CARIFORUM, which extends market party to the agreement are allowed to enter and reside in access to contractual services suppliers and independent Australia or Chile. Moreover, the spouse is given the right professionals (for stays of six months) and to graduate to enter, stay, and work, for a period of time equal to that trainees (for stays of one year). Japan has moved to liberalize of the national. access to its labor market for nurses and health care workers The ASEAN–Australia–New Zealand PTA, signed in in its recent EPAs with Indonesia and the Philippines. August 2008, contains a chapter on temporary entry, which Finally, both Australia and New Zealand have expanded Labor Mobility 285 the categories of labor in their PTAs to include contractual developed-country trading partners should find opportu- service suppliers and “installers� (for New Zealand) in nities that did not exist in the past.9 their recent agreement with ASEAN members and in the New Zealand PTA with China. The latter also contains PTAs Negotiated among Developing Countries novel provisions for artisans who are proficient in Chinese cultural occupations, such as theater, language, and medi- Labor mobility is not confined to PTAs involving devel- cine. Australia’s PTA with Chile covers the spouses and oped countries; it is also an important feature of several dependents of intracorporate transferees and contractual South-South PTAs. This section presents a brief overview services suppliers residing in the country longer than one of labor mobility provisions in several agreements. year. Thus trade agreements have moved over the past two The Caribbean Community (CARICOM) has had one years beyond the purely professional categories of labor to of the most successful experiences with liberalizing the include within their scope contractual services suppliers, movement of service providers at the regional level, dat- semiprofessionals and technicians, nurses and health care ing from the signing in 1998 of Protocol II, on “Establish- workers, and even spouses and dependents. ment, Services and Capital.� The objective of the protocol The trading partners that have been the most willing to is to bring the CARICOM Single Market Economy into open their markets wider for foreign workers from devel- effect. (For further discussion, see Stephenson 2007.) oping PTA partners have been countries that face consid- CARICOM provisions rest on two pillars: (a) facilitation erable labor shortages. Canada has shown itself the most of travel, common travel documents, and national treat- generous in this respect, with Japan being selective and ment at the port of entry (Article 46 of the CARICOM sector-specific in responding to its labor market needs. Single Market Economy treaty) and, since 2005, a com- Australia has been willing to consider family dependents mon passport; and (b) the free movement of skilled per- as part of the labor categories defined under its most sons within the community (Articles 32, 34d, 35d, 36, recent PTA. The United States and the EU have faced 36a, and 37 of the treaty). Five categories of skilled work- heavy inward migration flows, both documented and ers were initially identified: undocumented, from Latin America (in the U.S. case) and from North Africa and Eastern Europe (for the EU), • Graduates of the Universities of the West Indies, and they are less willing to contractually bind greater Guyana, and Suriname market openness for foreign workers in their PTAs.8 • Graduates of approved universities outside the region Nonetheless, the EU did expand its coverage of labor • Media workers categories in the recent EPA with CARIFORUM mem- • Musicians bers. In the United States, official and public attitudes • Artists have turned sour, and no agreements have been negoti- • Sports persons ated containing mode 4 coverage since 2002. • Workers in the tourism and entertainment industries The story of labor mobility within trade agreements is • Any other skilled person eligible under Articles 35d and still being written; the situation continues to evolve. Cur- 36a of Protocol II. rently, several PTAs between developed and developing economies are under negotiation. The EU is negotiating Effective January 1, 2010, domestic helpers were added with ASEAN, Colombia, Ecuador, India, Korea, Peru, five to the list.10 Since 2007, discussions have been under way countries in Central America, and the four members of the on adding teachers and nurses. Southern Cone Common Market (Mercosur, Mercado A certificate of recognition must be obtained from the Común del Sur). Canada is negotiating with the Caribbean respective national labor ministries by those wishing to Community (CARICOM), four countries in Central move abroad. A six-month temporary residency permit is America, the Dominican Republic, Jordan, Korea, Panama, issued while the certificate is reviewed by the receiving and Singapore. Japan is negotiating with India and Peru, country, after which an indefinite work and residence and Australia is negotiating with China, the Gulf Coopera- permit is granted. CARICOM has recognized the impor- tion Council (GCC), Korea, and Malaysia. Only the United tance of transferability of social security benefits, but States is currently abstaining from further involvement in progress on this matter has been slower than expected. regional trade negotiations. Thus, the sample for evaluat- Achieving the free movement of workers is also stated ing the treatment of labor mobility in PTAs will continue as a goal of the East African Economic Community in its to expand in the coming years. Developing countries that protocol on establishing a common market, expected to are able to proactively define and push their interests with enter into force on July 1, 2010. Many hurdles remain, 286 Sherry Stephenson and Gary Hufbauer however, on the path to full free movement. Work per- of residence.14 It further provides for nondiscrimination mits are not harmonized across countries; they are rela- with respect to the right to seek and engage in employ- tively restrictive and remain difficult to obtain; they are ment. In 2005 the conference of WAEMU/UEMOA heads subject to delays and administrative requirements; and of state and government approved a progressive approach rejections are numerous. Portability of social benefits is toward the implementation of freedom of movement for very limited. Progress has been made on harmonization persons, the right of residence, the provision of services, of standards and mutual recognition for graduates, but and the right of establishment. This suggests the adoption much less so for technical and vocational training. of regional codes of freedoms and rights of movement, as The treaty of the Economic Community of West well as harmonization measures. The codes concern four African States (ECOWAS) requires the community to areas (OECD 2008): ensure the removal of obstacles to the free movement of persons, goods, services, and capital and to guarantee the • Right of establishment for the freedom to carry out self- right of residence and establishment. To date, ECOWAS employed professions has signed three supplementary protocols on this subject. • Under equal conditions, access to higher-education The first provides for the free entry of community citizens establishments for a period of 90 days without a visa, provided that they • Establishment of a community visa for nationals of possess travel documents, and also grants them the rights countries outside the WAEMU/UEMOA or ECOWAS of entry, residence, and establishment.11 The second pro- zones tocol allows community residents to reside, seek employ- • Building of control posts juxtaposed on both sides of ment, and engage in income-earning employment in any the border of member countries. member state.12 The protocol specifically refers to migrant workers, defined as nationals of community member In 2006 regulations were adopted on free movement and states who seek or propose to hold employment, are right of establishment for workers in specific professions already holding employment, or have in the past held (for example, accountants and pharmacists). Today, the employment in a member country. Special provisions are commission is working on a draft common policy in the made for four protocol categories: migrant workers, itin- areas of movement and stay by third-country nationals. erant workers, seasonal workers, and border workers In Latin America, as in Southeast Asia, progress in lib- (Mattoo and Sauvé 2010). eralizing labor mobility has been slow within regional Despite early provisions on the free movement of arrangements. Mercosur members have included free- persons, implementation within ECOWAS has been dom of movement among their integration goals. In the- slow—hampered, in particular, by the efforts of young ory, Mercosur nationals may currently move among member states to affirm their sovereignty. At times, slow member states, although the right to work is regulated by progress in other areas of economic integration and host governments. Progress in liberalizing labor mobility adverse reactions to the influx of foreign labor in periods has been sluggish. A Mercosur social security agreement of recession have hindered implementation (OECD was signed in 1997, but many of the steps aimed at facil- 2008). In recent years, several measures have been under- itating migration within the community are taking far taken—in particular, since 2000, in the harmonization of longer to be implemented than planned. Much of the passports, as well as joint border operations by customs migration that occurs in the Mercosur region is outside and migration offices.13 formal channels. In December 2002, Mercosur leaders An interesting development took place in January 2008 signed an Agreement on Residency for Mercosur Nation- when ECOWAS adopted a common approach to migra- als aimed at giving migrants “equal civil, social, cultural, tion, clearly influenced by the European model (OECD and economic rights and freedoms� with the citizens of 2008). The approach consists of two parts, the first devoted the Mercosur country in which they are living, “particu- to the legal framework and key principles, and the second larly the right to work and to carry out any legal activity.� to a regional migration and development action plan. The related Agreement on Regulating the Migration of For countries that are also members of the West African Mercosur Citizens encouraged Mercosur governments to Economic and Monetary Union/Union Économique et legalize unauthorized nationals of Mercosur members Monétaire Ouest-Africaine (WAEMU/UEMOA), the treaty (World Bank 2010). confers the right of free movement of people, the right to In October 2003, ASEAN members raised their ambi- provide services, the right of establishment of persons car- tions from the formation of a free trade area to the cre- rying out an independent or salaried activity, and the right ation of an ASEAN Economic Community (AEC), in the Labor Mobility 287 Declaration of the ASEAN (Bali) Concord II, which was inward flows to specific areas of labor demand. For desti- subsequently endorsed by summits of ASEAN leaders (Soe- nation countries, the primary aim is to address skill gaps sastro 2005). Nevertheless, the liberalization of labor move- in the local labor market, whether for seasonal workers ment still has a very long way to go before achievement of (often in the agricultural sector) or low-skilled labor. this objective. Currently, ASEAN members have made Occasionally, BLAs also deal with higher-skilled workers only very modest commitments on mode 4 in their in areas of labor shortages such as health or information respective schedules of services commitments, which technology.15 have now undergone five rounds of negotiations under Bilateral labor agreements have had an interesting his- the ASEAN Framework Agreement on Services since that tory. They were popular in the United States and Europe pact went into effect in January 1996. Many of the mode in the 1960s but fell into disfavor in the 1970s and 1980s, 4 commitments go no farther than what is set out in affected by the adverse combination of inflation and high members’ WTO schedules. ASEAN has, however, made unemployment that came to be known as “stagflation.� more progress with the realization of mutual recognition For 22 years, beginning in 1942, the United States had a agreements (MRAs) than any other regional grouping, bracero program to admit temporary agricultural workers having signed six MRAs to facilitate the movement of from Mexico. Admissions under this program peaked at professional services suppliers through the recognition more than 450,000 a year but began to shrink because of of their professional accreditations. These agreements the enforcement of labor market regulations, combined cover engineering services, nursing services, architec- with technological changes. Nonetheless, the program tural services, medical practitioners, dental practitioners, continued to admit more than 200,000 Mexican tempo- and accountancy services. rary workers a year until it ended in 1964. In Western Europe, temporary worker programs were peaking when they were ended unilaterally in 1973–74. European tem- Bilateral Labor Agreements porary worker programs differed from the Mexico–U.S. Efforts to manage labor mobility among developing program in several important respects, including the locus countries at the regional level have been consequential, of employment (nonfarm manufacturing, construction, although progress has not matched aspirations. The most and mining, rather than agriculture), as well as in their notable efforts have taken place in agreements aiming for policies toward settlement. Unlike Mexicans who filled a common market; these generally go beyond simply seasonal U.S. jobs and were expected to return to Mexico managing labor mobility to encompass migration dimen- every year, migrants in Europe filled year-round jobs and sions. The examples reviewed above all concern countries earned rights to unify their families and settle with work that are regional neighbors with a significant history of and residence permits.16 population migration. It is worth highlighting some Several developed countries have entered into second- interesting initiatives, which include coverage of recent generation bilateral labor agreements (as of the turn of graduates, common migration policies, and common the millennium), although many of these were in the passports to facilitate border crossings. form of memoranda of understanding, rather than more As was noted above, the treatment of labor mobility in formal contractual arrangements. BLAs do not take any formal PTAs has focused overwhelmingly on skilled labor one set form; in fact, there is such a variety of agreements categories, with only a few recent agreements moving to that international organizations have developed a “Com- cover certain types of semiskilled workers. Against this pendium of Good Practice Policy Elements in Bilateral background, can other vehicles be used to promote labor Temporary Labour Arrangements� as a follow-up activity mobility? Bilateral labor agreements (BLAs) are alterna- to the first Global Forum on Migration and Development tives to the more legalistic and rigid PTAs and can serve (GFMD), held in 2007. both to promote and to regulate the flow of unskilled or BLAs have been increasing in number over recent years, semiskilled workers. but no single institution is responsible for collecting and maintaining information on them. Neither the ILO nor the IOM (International Organization for Migration) has Short History information on BLAs at the country level.17 It is therefore Bilateral labor agreements have provided a means for extremely challenging to collect these data, and what is employing seasonal and low-skilled foreign labor on a presented in this section will certainly be incomplete. temporary basis. They allow industrial countries that need Although many countries have entered into bilateral labor foreign labor to design labor exchange programs that steer agreements, others prefer to channel their temporary 288 Sherry Stephenson and Gary Hufbauer labor needs through their more formal immigration chan- agricultural workers. The approved requests are then com- nels. In the United States, most temporary admission pro- municated via Canadian network contacts in Mexico to pri- grams are open to citizens of all countries. The range of vate recruitment agencies in the participating Caribbean temporary visa programs includes both skilled profession- countries. Finding the workers to fill the required demand als (e.g., H-1B visas) and other kinds of temporary labor is then the responsibility of the countries of origin. In (e.g., H-2A temporary agricultural workers). 2000 about 7,300 Mexicans were among the 16,900 foreign In examining the panorama of bilateral labor agree- farmworkers admitted to Canada; the other workers were ments that we have been able to identify for this study and from Barbados, Colombia, Jamaica, Trinidad and Tobago, that are set out in table 13.1, it is interesting to note that and six other Eastern Caribbean islands.19 The BLA with such agreements have now been signed by countries in all Colombia is the result of demands by Canadian compa- regions of the world. In the Americas, Canada has been nies in Alberta and Manitoba for Colombian workers in very active in developing bilateral temporary worker pro- the food-packing industry. grams and has concluded agreements with Barbados, The BLAs concluded by Spain provide for a selection Colombia, Guatemala, Jamaica, Mexico, Trinidad and committee that is made up of representatives of the par- Tobago, and the countries of the Eastern Caribbean. In ticipating governments and is responsible for selecting Europe, the governments of Germany, Italy, Spain, and the best-qualified workers for existing job offers and for the United Kingdom have actively negotiated bilateral conducting training courses that may be needed. In these labor agreements with developing countries around agreements for regulating labor migration flows, the the world. In Africa, South Africa has pursued such Spanish authorities, through Spanish embassies in origin arrangements, mainly with neighboring countries. In Asia, countries, notify authorities in the origin country of the China has concluded BLAs with several developed- and number and types of workers needed. There is no set developing-country partners.18 quota. Origin countries in turn notify the Spanish authorities of the possibility of meeting this demand with their nationals willing to go to Spain. Terms of Coverage Spain’s bilateral agreement with Colombia covers agri- The coverage of the bilateral labor agreements varies. cultural workers who are selected to work temporarily in Canada’s agreements cover exclusively the agricultural sec- fruit harvesting in the Catalonia region. Within the frame- tor. The Seasonal Agricultural Workers Program (SAWP) work of the temporary and circular labor migration mode is based on bilateral memoranda of understanding and is that implements this agreement, the National Training managed by Human Resources and Skills Development Institute in Colombia designs training programs for the Canada (HRSDC). Canadian employers submit requests, labor migrants so that, on their return to their communi- which have to be approved by the HRSDC, for foreign ties of origin, they can transfer the skills and know-how Table 13.1. Bilateral Labor Agreements with Developing-Country Partners: Government Programs for Temporary Workers Region and country Developing-country partners Americas and the Caribbean Canada Barbados, Colombia, Guatemala, Jamaica, Mexico, Organization of Eastern Caribbean States, Trinidad and Tobago Europe France Mauritius Germany Bulgaria, Croatia, Czech Republic, Poland, Romania, Slovak Republic, Slovenia, Ukraine Greece Albania, Bulgaria Italy Albania, Moldova, Sri Lanka, Tunisia Spain Bulgaria, Colombia, Dominican Republic, Ecuador, Mauritania, Morocco, Philippines, Romania, Senegal United Kingdom India, Philippines, Spain Asia China Australia; Japan; Jordan; Korea, Rep.; Mauritius; South Africa; Spain; United Arab Emirates Africa South Africa Botswana; Cuba; Iran, Islamic Rep.; Lesotho; Malawi; Mozambique; Swaziland; Tunisia Source: ILO, IOM, and OSCE, “Compendium of Good Practice Policy Elements in Bilateral Temporary Labour Arrangements,� revised version, December 2, 2008. Note: ILO, International Labour Organization; IOM, International Organization for Migration; OSCE, Organization for Security and Co-operation in Europe. Labor Mobility 289 acquired in Catalonia. Under this BLA, less than 10 per- specific numbers of recruited workers. The number of Chi- cent of selected Colombian workers have failed to return nese citizens working as temporary laborers abroad has home. In Ecuador a Migration and Control Unit was cre- increased substantially, from 63,200 in 1987 to more than ated in 2002 within the Ministry of Foreign Affairs to half a million in 2004.20 receive job vacancy notices from Spanish enterprises and The BLAs in which China is a partner cover diverse match the job offers with the most appropriate candi- labor sectors. Chinese labor cooperation with the United dates through a large database. Spain has similar BLA Arab Emirates takes place in the areas of construction, programs with Bulgaria, the Dominican Republic, factories, medical care centers, and maritime activities. Mauritania, Morocco, Romania, and Senegal. Under the With Australia, the BLA centers on nursing and a few Spain–Philippines BLA, nurses and other Filipino work- other sectors, and an attempt is made to curb the exces- ers are allowed into Spain and are afforded the same pro- sive fees charged by the recruitment agencies by offering tections as Spanish workers. the alternative of government employment offices. Under The bilateral labor agreements signed by the United the BLA with Mauritius, Chinese workers may be recruited Kingdom with India, the Philippines, and Spain enable the only from companies that are on an established, govern- United Kingdom to recruit registered nurses and other ment-approved list. The BLA with Jordan concerns the tex- health care professionals (physiotherapists, radiographers, tile and construction sectors. An agreement with South occupational therapists, biomedical scientists, and other Africa was concluded in 2002 and was extended in 2006 to workers regulated by appropriate professional bodies in focus on human resources development and job creation both countries) for work on a temporary basis. The strategies, in addition to worker recruitment. A successful U.K.–Spain agreement provides for recognition of Spanish pattern for the bilateral labor agreement has been the nursing skills in the United Kingdom. agreement between China and Japan, under which more Greece has signed BLAs in the agriculture and fisheries than 30,000 Chinese trainees are sent to Japan every year in sectors. Under the agreements with Albania and Bulgaria, temporary labor (trainee) cooperation programs.21 Greek authorities assess the annual need for seasonal The advantages of BLAs or temporary worker pro- agricultural workers and grant residence and permits to grams, particularly for lower-skilled categories of work- workers from these countries according to demand from ers, are numerous. First, and most important, is the flexi- Greek employers. Under the BLA with Egypt, which cov- bility they offer with respect to the management of the ers the fisheries sector, temporary labor migrants are sub- labor market by the countries involved. Such agreements ject to specific regulations regarding the possibility of can be negotiated in response to the economic cycles of changing employers and the extension of their stay in the the market.22 As is seen in the examples cited, they can be country, and they are eligible for the transfer of social targeted to specific sectors and can even be firm based, if security rights and pensions on a mutual basis. necessary. Monitoring of such agreements can be carried South Africa has negotiated several bilateral agree- out on both sides as a joint responsibility, rather than ments with neighboring countries in response to its grow- putting the burden entirely on the destination country to ing labor crisis. The Joint Initiative for Priority Skills determine the legality of the worker. Guarantees can be Acquisition (JIPSA) Act of 2004 acknowledged that par- designed and written into the agreements in the form of ticular sectors require skills from outside the country. bonds or fines for noncompliance, to encourage respect South African mining companies fought hard to keep for the provisions by private parties. Incentives can their right to hire foreign contract workers, and the 2002 be built in on both sides of the agreement. Workers are Immigration Act was modified to accommodate this pres- more willing to respect the contract and return home sure. Bilateral agreements are focused on recruiting work- if there are prospects of an opportunity (based on per- ers from Botswana, Lesotho, Malawi, Mozambique, and formance and need) to go back to the host country for Swaziland to work in the mines and farms of South Africa. future employment. Most participants in the BLAs that The share of foreigners in the mines’ workforce rose from have functioned to date have found that the agreements 47 percent in 1990 to 60 percent in 2000, but this share has have fulfilled the expectations of both sides.23 declined recently in response to efforts to hire locally. The disadvantage of bilateral labor agreements is that, China has negotiated several BLAs with willing partners unlike PTAs, they are single-issue instruments. This limi- experiencing labor shortages, including Australia, Japan, tation means that developing-country partners do not Jordan, Korea, Mauritius, South Africa, Spain, and the have scope within a BLA to trade their “offensive� interests United Arab Emirates. These agreements are very diverse, in labor mobility for the “offensive� interests of their cover a wide range of topics on labor cooperation, and list developed-country trading partners. 290 Sherry Stephenson and Gary Hufbauer Multilateral and MFN Considerations be applied with “flexibility� when a developing country is a party to the economic integration agreement, under the It must be recognized that bilateral labor agreements repre- provisions of Article V(2). In our judgment, the Appellate sent an important derogation from the most favored Body would give great weight to Article V(2) in evaluating nation (MFN) principle that is the core of the world trad- a BLA. In fact, we think the “flexibility� provision would be ing system. The same is true of bilateral investment treaties decisive. (BITs) and double-taxation treaties (DTTs), of which there Third, there is the matter of negotiating history. As our are many thousands in the world today. None of these pre- discussion shows, BLAs, like BITs, have been around for a tend to treat all countries on equal terms; partners are very long time, predating the original GATT (signed in favored over nonpartners. Of course PTAs also discrimi- 1947) by decades. A strong argument can be made that if the nate between partners and outsiders, and those inside the Uruguay Round negotiators had meant to impose an MFN PTA receive more favorable treatment than those outside. requirement on these agreements, they would have said so in From the perspective of true multilateralism (“no discrim- very explicit terms. After all, important economic arrange- ination is the first-best policy�), these various forms of ments would have been upset by an MFN requirement. bilateral and regional agreements clearly occupy a second- Silence seems to indicate assent to the status quo ante. best world. The tension that characterizes negotiations over One of the main reasons that countries enter into BLAs is trade liberalization, as well as over investment and labor that these agreements are flexible and short term and appear mobility, is between a first-best multilateral approach, to escape the long-term contractual constraints of GATS. The which may be stalled because of lack of agreement among large majority of bilateral labor agreements cover a different countries worldwide, and a second-best regional or bilateral category of worker than do the formal services agreements approach that achieves liberalization between the partners (PTAs or GATS); they focus on unskilled (agricultural) or but creates discrimination against the rest of the world. A lower-skilled workers, whose movements governments have large and robust literature has developed around the debate not been willing to liberalize in the context of either GATS or as to which approach will engender the most liberalization regional services agreements. We conclude, from this brief and the greatest gains over the horizon of a decade or and sketchy review, that the Appellate Body would very likely longer. We will not rehearse the arguments here, since they respect the special status of BLAs if a claim were ever are familiar to most readers; we simply observe that the brought. As a practical matter, no WTO member has much issue is certainly not settled. For the past decade, however, interest in bringing a claim, and the possibility of litigation most governments have been “voting with their feet� by seems remote. placing more emphasis on bilateral and regional agree- ments. This trend seems very likely to continue. Conclusions What about the consistency of BLAs with GATS? This question has not been litigated in the WTO and is not This chapter has examined how recent preferential trade likely to be litigated any time soon, so a definitive answer agreements (those concluded since 1994 between devel- cannot be given. Three considerations, however, would oped and developing trading partners, as well as some probably have weight in the WTO Appellate Body’s rea- South-South PTAs and labor agreements) have dealt with soning, if a nonparty to a BLA did claim that its GATS labor mobility. It has shown that some of the most recent rights to labor mobility (mode 4) were violated by the PTAs have innovated in interesting ways to promote labor bilateral agreement. mobility, either by expanding the number of services sup- First, GATS Article II(1) establishes the MFN principle pliers accepted under particular categories (for example, for services, including mode 4. Under Article II(2), MFN without numerical quotas) or by creating new space for can be waived for a BLA, as for any other GATT or GATS specifically defined categories of labor, such as techni- obligation, but this requires a favorable vote by three- cians, nurses and health care workers, and sporting and fourths of WTO members. In any event, there are no extant cultural occupations. To date, however, all but a very few waivers for BLAs or BITs.24 PTAs that cover services focus on professional services Second, under GATS Article V(1), two or more coun- suppliers. A new generation of less formal temporary tries can enter into an economic integration agreement to worker programs is paying more attention to the needs of liberalize trade in services and thereby avoid the MFN lower-skilled and semiskilled temporary workers. Some requirement. The agreement should have “substantial sec- of the regional integration groupings among developing toral coverage� and should eliminate “substantially all dis- countries in Africa, Asia, and Latin America and the crimination� between the parties. These conditions are to Caribbean are making progress toward the opening of Labor Mobility 291 labor markets at the regional level to all categories of for firms and individuals that offer unique cultural talents workers, both for temporary movement and for perma- or specialized skills, as well as for some independent pro- nent settlement. Members of these groupings appear will- fessionals, and geographic or occupational niches of the ing to go farther in their ultimate objectives than is the industrial developed economies that suffer from labor case under the North-South PTAs, where the norm is to shortages. If developing countries wish to promote exports cover prescribed and limited, although often expanded, of services providers in the health services, this is certainly categories of workers. an area that offers a large potential for expansion. For this Thus, while the latest steps are positive and encouraging market, it might be advisable to develop local training pro- for developing countries, they leave much work for future grams for the specific skills required in the target market, negotiators. In our view, patience should be the watchword in the way that the Philippines has done and Indonesia is of negotiators based in developing countries. They should currently doing. take heart and guidance from the long experience of devel- oped and developing countries in crafting the liberaliza- 2. Concerning semiskilled and lower-skilled workers tion of trade in textiles and clothing. This was a supersensi- In the case of workers with lower skill levels and less formal tive industry as early as the late 1950s, when the educational training, the best vehicle for promoting greater Eisenhower administration in the United States negotiated labor mobility is not formal PTAs but the more flexible the first restraint agreement with Japan, and it remained instrument of temporary worker programs (TWPs). These sensitive for the next 50 years. Eisenhower’s accord with programs can be designed to promote circular migration Japan was followed by the Short-Term Cotton Agreement in a way that benefits the labor-sending and labor-receiv- and the Long-Term Cotton Agreement in the Kennedy and ing countries, as well as the workers themselves. TWPs are Johnson administrations, and then three generations of the extremely flexible in both design and execution and allow Multifibre Arrangement (MFA) under GATT auspices. the parties involved to design the clauses covering length of The complexity of bilateral textiles and clothing quotas stay, nature and place of employment, and appropriate under these agreements was truly bewildering and, from an guarantees. They also offer governments the possibility of economist’s viewpoint, highly distortive. But within this adjusting in a responsive manner to the cycles of their complex framework, over the span of five decades, trade in domestic labor markets. Such agreements must elicit the the sector was greatly liberalized and grew enormously. positive involvement of parties on both sides, making this a The secret, if there was a secret, was that negotiators of framework with buy-in, where all parties to the agreement good will, representing both developed and developing have an interest in seeing it succeed. Although these agree- countries, discovered niches of textiles and clothing ments have been successfully promoted so far by only a trade where the political costs of further liberalization, handful of countries, primarily China and the Philippines, combined with suitable safeguard mechanisms, were not there is tremendous scope for their further application in insurmountable. At every stage of this long process, the the world economy. economic gains from liberalization were enormous; the “magic,� if there was any magic, was to focus attention on 3. For developing-country governments and negotiators products and mechanisms that did not encounter over- Developing country governments and negotiators should whelming resistance in the developed countries. We think bear six precepts in mind: the same approach commends itself to labor mobility nego- • Developing-country negotiators should approach the tiations—a long, persistent, and patient search for niches in discussions of labor mobility with a positive attitude the labor markets of developed countries where greater and should emphasize the gains to the destination entry of migrants is not only tolerated but welcomed. country. The economic gains are invariably large, and On the basis of this overview, we offer four sets of rec- the political costs are often exaggerated, so it is useful ommendations. for negotiators from developing countries to research 1. Concerning professional workers particular labor markets and lay the facts on the table. When developing countries are able to define their inter- • To better serve their negotiators, developing-country ests well and are willing and able to pursue bilateral trade governments should conduct in-depth research on the agreements with the major developed trading partners labor markets of potential destination countries with reviewed in this study (other than the United States, at the aim of discovering promising niches. This will present), they should be able to obtain expanded market require the services of specialized officers or contractors access. Labor markets worth exploring are opportunities working in the destination countries. 292 Sherry Stephenson and Gary Hufbauer • Developing-country specialists should work with edu- respectful people. When adverse incidents happen, as cational and credentialing authorities in the developed they will, the government of the developing country countries to lay the groundwork for mutual recognition should cooperate as appropriate, through revocation of agreements for the benefit of their independent profes- visas and other measures. sionals and other highly skilled workers. • When multinational corporations seek to expand their 4. For developed-country governments and negotiators operations abroad, whether in a developed or a develop- Like developing countries, developed countries should ing country, government negotiators should team up with proactively search for labor market niches where addi- the corporations to ensure agreement on the requisite tional temporary workers will become valued members of number of visas for intracorporate transferees and con- the workforce and the community. Developed-country tractual services suppliers to support the new operation. officials must not surrender to arguments that the labor This needs to be done whether or not a PTA is in place. market is an undifferentiated mass, or succumb to • Developing-country negotiators should seek agreement the anti-immigrant voices of a vocal minority. They on the status of mode 4 workers, meaning their rights as should hammer home the distinction between permanent to visas, working conditions, social security contribu- immigration, which remains under sovereign control, tions, unemployment compensation, and ability to remit and temporary workers who are subject to negotiated funds. To some extent, these matters are covered in agreements. They should seek to build flexible responses TWPs, but important elements are often not addressed. not only into TWPs but also into the quota and time • Above all, senior officials in the developing country clauses of PTAs. Most important, they should put some must attend to the “image� of their migrants abroad— effort into seeking out and conveying positive messages doing whatever is possible to ensure that their migrants about the contributions and accomplishments of tempo- convey an impression of hard-working, law-abiding, rary workers. Annex Annex Table 13A.1. Quantitative Estimates of Gains from Increased Labor Mobility Gains Region Authors covered Assumption or situation Borjas (1999) United U.S. 1980–2000 immigration Large redistributive effect: return Small net gains for natives: States wave, with immigrants to capital, +2 percent of GDP; US$10 billion a year (0.1 representing roughly labor wages, –1.9 percent percent of U.S. GDP), or roughly 10 percent of the U.S. of GDPa 5 percent of average economic workforce growth over past 20 years Hamilton and World Elimination of all restrictions 60.1–204.6 percent of 1977 Whalley (1984) on labor mobility (1977 data) world GDP in 1977b Iregui (1999) World Elimination of all restrictions Nonsegmented labor market: If only skilled labor migrates: on labor mobility (between 15–67 percent of world GDP 3–11 percent of world GDP 37 and 53 percent of the labor Segmented labor market (skilled endowment of developing versus unskilled): 13–59 percent regions migrates) of world GDP Moses and World Elimination of all restrictions For 1977, with 100 percent For 1998, with 100 percent Letnes (2004) on labor mobility (1977 and elimination of wage differential: elimination of wage differential: 1998 data) US$0.34 trillion–US 11.27 trillion US$1.97 trillion–US$55.04 (1977 dollars) (more probably, trillion (1998 dollars) (more US$0.58 trillion); 4.3–111.6 probably, US$3.4 trillion); percent of 1977 world GDP 5.6–155 percent of 1998 world (more probably, 7.5 percent of GDP (more probably, 9.6 1977 world GDP) percent of 1998 world GDP) For 1977, with 10 percent For 1998, with 10 percent elimination of wage differential: elimination of wage differential: 22 percent of total potential 23 percent of total potential gain; gain; wages, +4.1 percent in wages, +11.4 percent in poorest poorest countries; +3.3 percent countries; +2.1 percent in in middle-income countries; middle-income countries; –2.5 percent in richest countries; –3.1 percent in richest countries; return to capital, –8.3 percent in return to capital, –21.0 percent poorest countries; –6.9 percent in poorest countries; –4.4 percent in middle-income countries; in middle-income countries; +5.7 percent in richest countries +7.2 percent in richest countries (continued next page) Labor Mobility 293 Annex Table 13A.1. (continued) Gains Region Authors covered Assumption or situation Walmsley and World Increase in migration from Total: +0.6 percent of world GDP Change in real wages of unskilled Winters (2002) developing countries to (US$156 billion in 2002, or workers: increase in developing high-income countries 1.5 times the expected gains countries (+0.7 percent in sufficient to increase labor from liberalization of all India); decrease in developed force in the host countries remaining goods) countries (–0.6 percent in the by 3 percent in 2002 Movement of unskilled workers United States) (accounting for most of the gains): Change in real wages of skilled +US$110 billion versus +US$46 workers: dramatic increase in billion for the movement of skilled developing countries workers (+4.5 percent in Mexico); Migrants’ welfare: +US$171 billion decrease in developed (+US$73 billion in the United countries (–0.8 percent in States, +US$25 billion in Japan, the United States) +US$68 billion in the EU) Change in rental price of capital: Resident welfare: net, –US$15 decrease in developing billion; developing countries, in countries (–0.4 percent in some cases, gain if remittances Mexico); increase in developed are high (+US$16 billion in India), countries (+0.8 percent in the but most lose (–US$7 billion in United States) Brazil); developed countries, small gains (+US$3.9 billion in EU) World Bank World Increase in migration from +0.6 percent of world GDP, (2006), 31 developing countries to (US$356 billion in 2025); high-income countries +0.4 percent of developed- sufficient to increase the country GDP; +1.8 percent of labor force in the host developing-country GDP countries by 3 percent by (including migrants’ income) 2025 (revision of Walmsley and Winters 2002) Source: Studies listed under “Authors�; see the bibliography for details. Note: EU, European Union; GDP, gross domestic product. a. Hatton and Williamson (1998) find similar results on wages when studying the 1870–1910 migration wage in the United States; they estimate that U.S. wages in 1910 would have been 11 to 14 percent higher in the absence of immigration after 1870. b. The large differences in estimates, both within and between studies, can be explained by differences in modeling frameworks (partial versus general equilibrium), production elasticities, productivity, cost of movement, or workforce size. Annex Table 13A.2. Agreements between the United States and Developing Countries Provision U.S.–Singapore U.S.–Chile U.S.–Morocco CAFTA–DR U.S.–Peru Entry into force January 1, 2004 January 1, 2004 January 1, 2006 March 1, 2006 February 1, 2009 Chapter on trade in Ch. VIII Ch. 11 Ch. 11 Ch. 11 Ch. 11 services Treatment of foreign services National treatment Yes (Art. 8.3) Yes (Art. 11.2) Yes (Art. 11.2) Yes (Art. 11.2) Yes (Art. 11.2) Most favored nation Yes (Art. 8.4) Yes (Art. 11.3) Yes (Art. 11.3) Yes (Art. 11.3) Yes (Art. 11.3) Local presence No (Art. 8.6) No (Art. 11.4) No (Art. 11.5) No (Art. 11.4) No (Art. 11.4) required Provisions on mode 4 Chapter Ch. 11 Ch. 14 None None None Committee Yes (Art. 11.7) Yes (Art. 14.5) Joint committee to Commission to — review the review the implementation of implementation the annex on of the annex on professionals professionals Dispute settlement Yes (Art. 11.8) Yes (Art. 14.6) — — — Transparency of Yes (Art. 11.5) Yes (Art. 14.4) — — — regulation (continued next page) 294 Sherry Stephenson and Gary Hufbauer Annex Table 13A.2. (continued) Provision U.S.–Singapore U.S.–Chile U.S.–Morocco CAFTA–DR U.S.–Peru Side letters Yes (professionals Yes (professionals None Yes (“No provision — must comply with will obtain visa shall impose any certain labor and through the obligation on a immigration laws U.S. H-1B party regarding its and have an program) immigration employer in the measures�) United States) Worker categories Investors, traders, Investors, traders, — — — covered intracorporate intracorporate transferees, transferees, professionals professionals Specification of length None None — — — of stay Provisions on professionals Annex on App. 11.A.2 Annex 14.3.D Annex 11.B Annex 11.9 Annex 11.B professionals Number of 2 (disaster relief 4 (disaster relief 0 (pledge to 0 (pledge to 0 (pledge to professional claims adjuster, claims adjuster, work on) work on) work on) categories management management covered consultant) consultant, agricultural manager, physical therapist) Specified quotas Singapore: no Chile: no numerical — — — numerical limit limit United States: 5,400 United States: 1,400 Postsecondary Yes: 4 years or more Yes: 4 years or more — — — degree required Specification of None None — — — length of stay Source: Authors’ compilation. Note: — = no provisions; CAFTA–DR, Dominican Republic–Central America Free Trade Agreement. Annex Table 13A.3. Agreements between Canada and Developing Countries Provision NAFTA Canada–Chile Canada–Colombia Canada–Peru Entry into force January 1, 1994 July 5, 1997 Signed November January 1, 2009 21, 2008 Chapter on trade in Ch. 12 Ch. H Ch. 9 Ch. 9 services Treatment of foreign services National treatment Yes (Art. 1202) Yes (Art. H-02) Yes (Art. 902) Yes (Art. 903) Most favored nation Yes (Art. 1203) Yes (Art. H-03) Yes (Art. 903) Yes (Art. 904) Local presence required No (Art. 1205) No (Art. H-04) No (Art. 905) No (Art. 907) Provisions on mode 4 Chapter Art. 16 Ch. K Ch. 12 Ch. 12 Side letters None None None None Committee Yes (Art. 1605) Yes (Annex K-05) Dispute settlement Yes (Art. 1606) Yes (Art. K-06) Yes (Art. 1206) Yes (Art. 1206) Transparency of Yes (Art. 1604) Yes (Art. K-04) Yes (Art. 1204) Yes (Art. 1204) regulation Worker categories Investors, traders, Investors, traders, Investors, traders, Investors, traders, covered intracorporate intracorporate intracorporate intracorporate transferees, transferees, transferees, transferees, professionals professionals professionals, professionals, technicians, spouses technicians (continued next page) Labor Mobility 295 Annex Table 13A.3. (continued) Provision NAFTA Canada–Chile Canada–Colombia Canada–Peru Specification of length None None None Peru: investors, 1 year; of stay traders, 90 days; intracorporate transferees, 1 year; professionals, 1 year; technicians, 1 year Canada: investors, 1 year; traders, 1 year; intracorporate transferees, 3 years; professionals, 1 year; technicians, 1 year Provisions on professionals Annex on professionals App. 1603.D.1 App. K-03.IV.1 App. 1203.D App. 1203.D Number of professional 63 (accountant, 72 (accountant, All categories of All categories except categories covered architect, medical architect, medical professionals except health, sports, art, professional, scientist, professional, health, sports, art, education, legal, and teacher, others) scientist, teacher, education, legal, and management services; others) management services; 50 technicians 50 categories of technicians (mechanical and avionics (mechanical and technician, construction avionics technician, inspector, food and construction inspector, beverage supervisor, food and beverage textiles processing, supervisor, textiles electrician, plumber, oil specialist, electrician, and gas well driller, chef, plumber, oil and others) gas well driller, chef, others) Specified quotas No numerical limit No numerical limit No numerical limit No numerical limit except for the United States: 5,500 Postsecondary degree Yes: 4 years or more Yes: 4 years or more Yes: professionals, 4 years; Yes: professionals, 4 years; required technicians, 2 years technicians, 1 year Specification of length None None None 1 year of stay Source: Authors’ compilation. Note: NAFTA, North American Free Trade Agreement. Annex Table 13A.4. Agreements between the European Union and Developing Countries Provision EU–Chile EU–CARIFORUM EU–Turkey EU–Morocco Entry into force February 1, 2003 December 29, 2008 December March 18, 2000 31, 1995 Chapter on trade in Title III, Ch. I Pt. II, Title II, Ch. 3 None Title III (pledge to work services on) Treatment of foreign services National treatment Yes (Art. 98) Yes (Art. 77) — — Most favored nation No Yes (Art. 79) — — Local presence required No (Art. 97) — — — Provisions on mode 4 Chapter Art. 101 Pt. II, Title II, Ch. 4 None None Side letters None None — — Committee Yes (Art. 100) Yes (Art. 85) — — (continued next page) 296 Sherry Stephenson and Gary Hufbauer Annex Table 13A.4. (continued) Provision EU–Chile EU–CARIFORUM EU–Turkey EU–Morocco Dispute settlement None Pledge to create one — — (Art. 87) Transparency of Yes (Art. 105) Yes (Art. 86) — — regulation Worker categories Investors, intracorporate Investors, intracorporate — — covered transferees, business transferees, business sellers, professionals sellers, professionals, graduate trainees Specification of length EU: professionals, Investors, 90 days; — — of stay 3 months intracorporate transferees, 3 years; business sellers, 90 days; independent professionals, contractual services suppliers, 6 months; graduate trainees, 1 year Provisions on professionals Annex on professionals Annex VII Annex IV None None Number of professional EU: 33 (engineer, EU: 29 (architectural, legal, — — categories covered accounting, accounting, engineering, construction, computer, management mining, computer, services) legal services, others) CARIFORUM: 0 Chile: 0 Specified quotas No numerical limit No numerical limit — — Postsecondary degree “Necessary academic University degree + 3 years — — required qualification and experience experience� Specification of length EU: 3 months EU: 6 months — — of stay Source: Authors’ compilation. Note: — = no provisions; CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; EU, European Union. Annex Table 13A.5. Agreements between Japan and Developing Countries Provision Japan–Mexico Japan–Chile Japan–Indonesia Japan–Philippines Entry into force April 1, 2005 Signed March 27, 2007 July 7, 2008 December 11, 2008 Chapter on trade in Ch. 8 Ch. 9 Ch. 6 Ch. 7 services Treatment of foreign services National treatment Yes (Art. 98) Yes (Art. 107) Yes (Art. 79) Yes (Art. 73) Most favored nation Yes (Art. 99) Yes (Art. 108) Yes (Art. 82) Yes (Art. 76) Local presence required No (Art. 100) No (Art. 109) No (Art. 78) No (Art. 72) Provisions on mode 4 Chapter Ch. 10 Ch. 11 Ch. 7 Ch. 9 Side letters None None None None Committee Yes (Art. 117) None Yes (Art. 96) Yes (Art. 113) Dispute settlement Yes (Art. 118) Yes (Art. 133) Yes (Ch. 14) Yes (Ch. 15) Transparency of Yes (Art. 116) Yes (Art. 132) Yes (Art. 95) Yes (Art. 111) regulation 4 Worker categories Investors, business Investors, business Investors, business Investors, business visitors, covered visitors, intracorporate visitors, intracorporate visitors, Intracorporate intracorporate transferees, transferees, transferees, transferees, professionals professionals professionals, professionals, professionals with professionals with “personal contracts,� “personal contracts,� nurses and care nurses and care workers workers (continued next page) Labor Mobility 297 Annex Table 13A.5. (continued) Provision Japan–Mexico Japan–Chile Japan–Indonesia Japan–Philippines Specification of length Japan: business visitors, Japan: business visitors, Japan: business visitors, Japan: business visitors, of stay 90 days; other 90 days; other 90 days; other 90 days; other categories, 3 years categories, 3 years categories, 3 years categories, 3 years Mexico: business Chile: business Indonesia: business Philippines: business visitors, 30 days; other visitors, 30 days; other visitors, 60 days; other visitors, 59 days; other categories, 1 year categories, 1 year categories, 1 year categories, 1 year Provisions on professionals Annex on professionals Annex 10 Annex 13 Annex 10 Annex 8 Number of professional Japan: 2 (engineer, Japan: 2 (engineer, Japan: 14 (legal and Japan: 10 (legal and categories covered specialist in specialist in accounting services, accounting services, humanities or humanities or engineer, specialist in engineer, specialist in international services) international services) humanities or humanities or Mexico: 42 (accountant, Chile: 41 (accountant, international services, international services’ engineer, lawyer, engineer, lawyer, nurse, health care Japanese university scientist, nurse, others) scientist, nurse, others) worker) graduate nurse, health Indonesia: 4 (mechanical care worker) and electrical engineer, Philippines: 4 (mechanical nurse, health care and electrical engineer, worker) nurse, health care worker) Specified quotas No numerical limit No numerical limit No numerical limit except — for nurses and health care workers Postsecondary degree 4 years or more 4 years or more Professionals, 4 years; Professionals, 4 years; required nurses and health care nurses and health care workers, public health workers, national health degree + 2 years work degree + 3 years work experience + 6 months experience + 6 months language training of training in the host country to pass the host-country certification exam Specification of length Japan: 3 years Japan: 3 years Japan: 3 years Japan: 3 years of stay Mexico: 1 year Chile: 1 year Indonesia: 1 year Philippines: 1 year Source: Authors’ compilation. Note: — = no provisions. Annex Table 13A.6. Agreements between Australia and New Zealand and Developing Countries Trans-Pacific SEP (Brunei Darussalam, Chile, New Zealand, ASEAN–Australia– Provision Singapore) New Zealand New Zealand–China Australia–Chile Entry into force May 28, 2006 Signed August 28, 2008 October 1, 2008 March 6, 2009 Chapter on trade in Ch. 12 Ch. 8 Ch. 9 Ch. 9 services Treatment of workers National treatment Yes (Art. 12-4) Yes (Ch. 8, Art. 5) Yes (Art. 106) Yes (Art. 9-3) Most favored nation Yes (Art. 12-3) Yes (Ch. 8, Art. 7) Yes (Art. 107) Yes (Art. 9-4) Local presence required No (Art. 12-7) No (Ch. 8, Art. 4) No (Art. 108) No (Art. 9-5) Provisions on mode 4 Chapter Art. 12-11 Ch. 9 Ch. 10 Ch. 13 Side letters None None None None Committee None None Yes (Art. 133) Yes (Art. 13-6) Dispute settlement None Yes (Ch. 9, Art. 9) Yes (Art. 134) Yes (Art. 13-7) Transparency of None Yes (Ch. 9, Art. 8) Yes (Art. 131) Yes (Art. 13-5) regulation (continued next page) 298 Sherry Stephenson and Gary Hufbauer Annex Table 13A.6. (continued) Trans-Pacific SEP (Brunei Darussalam, Chile, New Zealand, ASEAN–Australia– Provision Singapore) New Zealand New Zealand–China Australia–Chile Worker categories Professionals Business visitors, investors, Business visitors, investors, Business visitors, covered intracorporate transferees, intracorporate transferees, investors, contractual services contractual services intracorporate suppliers, installers suppliers, installers transferees, (to install purchased contractual services machinery—New Zealand suppliers, relatives only), spouses Specification of length No Indonesia: business visitors, China: business visitors, Australia: business of stay 60 days; investors, 6 months; investors, visitor, 1 year; 60 days; others, 2 years 6 months; intracorporate investors, 90 days; Australia: intracorporate transferees, 3 years; intracorporate transferees, 4 years; professionals, —; installers, transferees, 4 years; investors, 2 years; business 3 months professionals, 1 year visitors, 6 months; New Zealand: business Chile: — professionals, 12 months visitors, 3 months; New Zealand: business investors, 3 months; visitors, 3 months; intracorporate transferees, investors, 3 months; 3 years; professionals, intracorporate transferees, 3 years; installers, 3 years; installers, 3 months 3 months; professionals, 12 months Philippines: business visitors, 59 days; others, 1 year Vietnam: intracorporate transferees, 3 years; others, 90 days Provisions on professionals Annex on professionals Art. 12-11 Annex 4 Annexes 10 and 11 Annex 13-A Number of professional Pledge to “work on� Australia: 0 China: 5 (education, “Subject to national categories covered 6 categories New Zealand: 33 medical, translation, criteria� (engineer, architect, (engineering, legal, hotel, computer) geologist, taxation, veterinary, New Zealand: 6 (traditional geophysicist, computer, translation Chinese medicine, Chinese planner, accountant) services) chef, Mandarin teaching Indonesia: 13 (legal, tourism, aide, martial arts coach, restaurant, engineering, tour guide, skilled worker computer, R&D, “in category identified as maintenance services) being in shortage�) Philippines: all persons “who occupy a technical, advisory, or supervisory position� Vietnam: 2 (computer and engineering services) Singapore: 0 Specified quotas — Entry subject to national China: no numerical limit — rules New Zealand: traditional Chinese medicine, 200; Chinese chefs, 200; Mandarin teaching aides, 150; martial arts coaches, 150; tour guides, 100; skilled workers “in category in shortage,� 1,000 (continued next page) Labor Mobility 299 Annex Table 13A.6. (continued) Trans-Pacific SEP (Brunei Darussalam, Chile, New Zealand, ASEAN–Australia– Provision Singapore) New Zealand New Zealand–China Australia–Chile Postsecondary degree New Zealand: 3 years or China: “appropriate — required more + 6 years experience education level� Indonesia: “high + 2 years experience qualification� New Zealand: Philippines: “knowledge at “appropriate education an advanced level� level� + experience Vietnam: “university degree� + 5 years experience Specification of length — Vietnam: 90 days China: — Chile: — of stay New Zealand: 1 year New Zealand: 3 years Australia: 1 year Philippines: 1 year Indonesia: 2 years Source: Authors’ compilation. Note: — = no provisions; ASEAN, Association of Southeast Asian Nations; R&D, research and development; SEP, Strategic Economic Partnership. 300 Sherry Stephenson and Gary Hufbauer Annex Figure 13A.1. Provisions on Mode 4 in PTAs between the United States and Developing Countries 120 100 10 80 60 63 40 5.4 1.4 2 4 20 20 20 20 5 5 5 5 5 5 5 5 5 0 4) 00 re 00 ile 00 co 00 ru 00 R (2 A–D (2 po (2 –Pe 99 (2 Ch (2 roc 4) 4) 6) 6) 6) ga (1 – FT o S. S. –M in U. TA CA U. –S AF S. S. U. N U. quotas number of technician categories covered number of professional categories covered mode 4 categories covered annex on mode 4 chapter on services Source: Authors. Note: CAFTA–DR, Dominican Republic–Central America Free Trade Agreement; NAFTA, North American Free Trade Agreement; PTA, preferential trade agreement. Mode 4 refers to the movement of natural persons to supply services. The height of the bars indicates the degree of access that PTA provides for workers from developing-country partners. Values are assigned to each component of access as follows (not all components may be applicable to a particular agreement): Chapter on services in the PTA? If yes, 5 points. Annex on mode 4 service supply? If yes, 5 points. Mode 4 categories covered. Number of categories (shown on the bars) is multiplied by 5 to yield total points. Number of professional categories covered. Shown on the bars. Number of technician categories covered. Shown on the bars. Quotas. If uncapped, 10 points. Otherwise, shown as the total number of workers allowed under the quota, in thousands. Labor Mobility 301 Annex Figure 13A.2. Provisions on Mode 4 in PTAs between Canada and Developing Countries 200 10 10 50 50 150 10 100 10 94 (all but 6) 94 (all but 6) 72 63 50 30 25 20 20 5 5 5 5 0 5 5 5 5 4) 99 ile 00 ia 00 ru (2 mb (2 –Pe 99 (1 Ch 7) 8) 8) (1 o – da da ol TA na –C na AF Ca da Ca N na Ca quotas number of technician categories covered number of professional categories covered mode 4 categories covered annex on mode 4 chapter on services Source: Authors. Note: NAFTA, North American Free Trade Agreement; PTA, preferential trade agreement. For the method of deriving the values for the bars, see the note to annex figure 13A.1. 302 Sherry Stephenson and Gary Hufbauer Annex Figure 13A.3. Provisions on Mode 4 in PTAs between the European Union (EU) and Developing Countries 80 uncapped uncapped 70 10 10 60 50 29 33 40 30 25 20 20 10 pledge 5 5 5 5 5 0 EU–Morocco (2000) EU–Chile (2003) EU–CARIFORUM (2008) quotas number of technician categories covered number of professional categories covered mode 4 categories covered annex on mode 4 chapter on services Source: Authors. Note: CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; PTA, preferential trade agreement. For the method of deriving the values for the bars, see the note to annex figure 13A.1. Annex Figure 13A.4. Provisions on Mode 4 in PTAs between Japan and Developing Countries 100 80 10 10 10 60 10 44 43 18 14 40 30 30 20 20 20 5 5 5 5 5 5 5 5 0 Japan–Mexico Japan–Chile Japan–Indonesia Japan–Philippines (2005) (2007) (2008) (2008) quotas number of technician categories covered number of professional categories covered mode 4 categories covered annex on mode 4 chapter on services Source: Authors. Note: PTA, preferential trade agreement. For the method of deriving the values for the bars, see the note to annex figure 13A.1. Labor Mobility 303 Annex Figure 13A.5. Provisions on Mode 4 in PTAs between Australia and New Zealand and Developing Countries national 100 rules 80 China: 58 uncapped New Zealand: 60 1.8 10 national 40 11 rules pledge 30 25 25 20 6 5 5 5 5 5 5 5 5 5 0 00 fic 00 a– 00 – 00 a– (2 nd (2 li (2 rali (2 ci 6) 8) 8) 9) d stra a la P Pa ile ust in ea SE ns- an u Ch w Z al –A Ch A a Tr Ze AN e N ew ASE N quotas number of technician categories covered number of professional categories covered mode 4 categories covered annex on mode 4 chapter on services Source: Authors. Note: ASEAN, Association of Southeast Asian Nations; PTA, preferential trade agreement; SEP, Strategic Economic Partnership. For the method of deriving the values for the bars, see the note to annex figure 13A.1. 304 Sherry Stephenson and Gary Hufbauer Annex Figure 13A.6. Provisions on Mode 4 in PTAs between Developed and Developing Countries 200 180 160 140 120 100 80 60 40 20 0 e ile FT co R ru lo ile a ru CA C o FO le M o do ile ili sia es P in AN ile Z) ) TA us bi or SE D c ic RI hi RU in Ch a (N Pe Pe Ch Ch In Ch CA roc oc Ph ne AF A– ex m Ch ASE (A ap TP pp or N M o ng M M Co Si United States Canada European Union Japan Australia and New Zealand quotas number of technician categories covered number of professional categories covered mode 4 categories covered annex on mode 4 chapter on services Source: Authors. Note: ASEAN, Association of Southeast Asian Nations; CAFTA–DR, –Dominican Republic–Central America Free Trade Agreement; CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; NAFTA, North American Free Trade Agreement; PTA, preferential trade agreement; TPSEP, Trans-Pacific Strategic Economic Partnership. Notes the round resumed in 2007. A “signaling conference� was held at the request of interested ministers in July 2008, but it did not elicit much This chapter is a modified version of a chapter in Cattaneo et al. (2010). enthusiasm. Thibaud Delourme, a student at the Maxwell School of Syracuse Univer- 3. For an earlier discussion examining treatment of mode 4 in PTAs, sity, assisted with the research. see Nielson (2003). 1. Carzaniga (2008, 478) points out that foreigners working for a 4. The only exception is the FTA with Australia. No market access host-country company on a contractual basis as independent services provisions for labor mobility were included in the text itself, but a side let- suppliers (ISS) are covered by GATS (and, in general, by trade agreements ter was added after the conclusion of the negotiations that allowed an that include mode 4), whereas they would not be covered if they were annual quota of 10,500 Australian professionals to enter the U.S. market. employees of the company. What distinguishes their situation is the type This was done in 2002, and it proved to be the final straw for members of of payment received: the foreign employee receives a domestic currency the U.S. Congress. wage from the company in the host country, whereas the ISS is paid a fee 5. Sáez (2009) explains that within the EU, issues concerning trade in and the contractual services supplier is paid foreign wages. services do not fall exclusively under the competence of the community 2. By December 2005, the total number of services offers had reached because they go beyond Articles 113 and 238 of the treaty that accords 69, involving 93 WTO members. Of the 69, 31 were revised offers. There supranational treaty-making powers to the community on behalf of all has been very little change since; the number of initial offers has increased the member states. Thus, implementation of the services provisions and only to 71. Thus, one-third of WTO members have not put any initial obligations of a trade agreement must be approved by each EU member offer for services forward since the beginning of the negotiations. The state in accordance with domestic laws. Doha Round was suspended in July 2006, without any revised offers hav- 6. Article 101, on movement of natural persons, contains only ing been submitted. Services negotiations were not actively taken up when a review requirement: “Two years after the entry into force of this Labor Mobility 305 Agreement, the Parties shall review the rules and conditions applicable to migrant workers, in particular under the UN International Convention movement of natural persons (mode 4) with a view to achieving further on the Protection of the Rights of Migrant Workers and Their Relatives. liberalisation.� The memoranda do not include provisions for promoting labor mobility. 7. It should be mentioned that several EU member states have The Philippines has signed bilateral memoranda with many destination attached economic needs tests (ENTs) to their commitments on mode countries to cover the flows, rights, and obligations of its temporary 4 entry. Actual access provided, even under the expanded commitments, workers. A reciprocal temporary worker program agreed on by Argentina will depend on how these tests are interpreted and applied in practice. and Bolivia includes many of these protections. No definitions were supplied with the ENT entries, and some are applied 19. As part of Canada’s SAWP scheme with Caribbean countries and quite restrictively. Mexico, the HRSDC cooperates closely with private agencies, including 8. Indeed, a very recent PTA between the members of the European Foreign Agricultural Resource Management Services (FARMS) in Ontario Free Trade Association (EFTA)—Iceland, Liechtenstein, Norway, and and Nova Scotia and the Foundation of Enterprises for the Recruitment Switzerland—and Colombia, signed on November 25, 2008, does not of Foreign Labor (FERME) in Quebec, New Brunswick, and Prince even include an annex on professional service suppliers, and the body of Edward Island. Guest workers in Canada are employed in four provinces: the agreement contains no mention of the movement of natural persons, Ontario (two-thirds of the total), Quebec, Alberta, and Manitoba. other than their definition as mode 4. EFTA members are not facing labor Although the Mexican government tries to ensure that every worker shortages and, in the current hostile economic climate, did not feel any returns to Mexico, independent researchers estimate that 15 percent of the pressure to include liberalization of labor mobility in their agreement Mexicans fail to return home every year. See http://migration.ucdavis with Colombia. See http://www.sice.oas.org. .edu/RMN/more. 9. This approach is suggested by a paper by Henry Gao, “Report on 20. Gao, personal communication (see note 9). The data are from the China’s Export Interests in Services in Australia,� 2008, which was provided China Foreign Labor Cooperation Annual Report 2004, issued by the China to the authors. The strategy calls for collecting information on the labor International Contractors Association, Beijing. market and services export structure of the potential or current negotiating 21. See http://www.jitco.or.jp for details on this program. Informa- partner, carrying out surveys to identify those categories of labor with the tion on China’s bilateral labor agreements was provided to the authors by greatest potential for expansion following the removal of trade-restrictive Dr. Shu Bin, manager, Labor Department, China National Aero-Technol- barriers, and building on achievements of previous PTAs in the area. ogy Import & Export Corporation, during a workshop held in Beijing 10. Two members, Antigua and Barbuda, and Belize, were allowed a under World Bank auspices, May 21–22, 2009. five-year grace period to study the impact of free mobility for domestic 22. A recent study by Persin (2008) compares the United Kingdom’s helpers before adding them to their list. responses concerning its labor market and immigration policies in the 11. Protocol A/P.1/5/79, relating to free movement of persons, resi- context of the eastern enlargement of the EU with its willingness to pro- dence, and establishment, in application of Article 27 of the treaty estab- vide offers on mode 4 under the GATS. Persin finds that the government lishing ECOWAS. The successive texts complementing the free movement opted for managed migration through bilateral labor agreements and an regime are Supplementary Protocol A/SP.1/7/85, on the code of conduct employer-led system, rather than through more formal WTO commit- for the implementation of Protocol A/P.1/5/79; Supplementary Protocol ments. The author concludes that the more flexible “bilateral or regional A/SP.1/7/86, on the second phase (right of residence) of the aforemen- labor immigration schemes are preferred to a binding multilateral labor tioned protocol; Supplementary Protocol A/SP.1/6/89, amending and immigration scheme such as the GATS� because it is easier and less costly complementing the provisions of Article 7 of the aforementioned proto- under the former to agree on rules and procedures, as well as to solve any col; and Supplementary Protocol A/SP.2/5/90, on the implementation of problems jointly. the third phase (right of establishment) of the aforementioned protocol. 23. The Migration Policy Institute, based in Washington, DC, has 12. Supplementary Protocol A/SP.1/7/86. done considerable work on the benefits and challenges of temporary 13. Eight countries use the regional passport: Burkina Faso, Côte worker programs and circular migration schemes. See Batalova (2006); d’Ivoire, The Gambia, Ghana, Guinea, Niger, Nigeria, and Sierra Leone. Meyers (2006); Agunias (2007); Newland, Agunias, and Terrazas (2008). 14. Treaty Establishing the West African Economic and Monetary 24. Unlike BLAs and BITs, DTTs are specifically exempted from Union, 1996, Articles 4, 91–93. national treatment and MFN requirements by GATS Article XIV. 15. Host countries typically act unilaterally in determining whether temporary workers may bring their families and settle. Most European countries allow temporary workers whose work permits have been Bibliography renewed several times to obtain immigrant status after five years. 16. Between 1960 and 1973, the number of migrant workers in Agunias, Dovelyn Rannveig. 2007. “Managing Temporary Migration: Western Europe jumped from 2 million to 7 million, and the total for- Lessons from the Philippine Model.� Insight. Migration Policy Institute, eign population rose from 4 million to 12 million. Most of these work- Washington, DC. http://www.migrationpolicy.org/pubs/Insight_POEA ers came from geographically distant nations such as Turkey or _Oct07.pdf. Morocco, rather than neighboring countries. After the halt in temporary Batalova, Jeanne. 2006. “The Growing Connection between Temporary worker programs in the mid-1970s, the migrant workforce in Europe and Permanent Immigration Systems.� Task Force Insight 14. Migra- stabilized at around 5 million over the next decade. See Martin (2007); tion Policy Institute, Washington, DC. http://www.migrationpolicy see also Council of Europe (1996). .org/ITFIAF/TFI_Batalova.pdf. 17. The International Labour Organization (ILO) has developed a Borjas, George J. 1999. Heaven’s Door: Immigration Policy and the Ameri- multilateral framework on labor migration that constitutes a compre- can Economy. Princeton, NJ: Princeton University Press. hensive collection of principles, guidelines, and good practices on labor Carzaniga, Antonia. 2008. “A Warmer Welcome? Access for Natural migration programs, including bilateral labor agreements; see ILO, Persons under PTAs.� In Opening Markets for Trade in Services, ed. http://www.ilo.org/public/english/protection/migrant/areas/multilateral Juan A. Marchetti and Martin Roy, 475–502. London: Cambridge .htm. University Press. 18. A few memoranda of understanding on migratory and labor Cattaneo, Olivier, Michael Engman, Sebastián Sáez, and Robert M. Stern. cooperation have been signed recently by developing countries; they 2010. International Trade in Services: New Trends and Opportunities for include those between Peru and Chile (2006), Peru and Ecuador Developing Countries. Washington, DC: World Bank. (2006), and Peru and Mexico (2002). The aim of these memoranda is to Council of Europe. 1996. “Temporary Migration for Employment and provide for exchange of information and protection of the rights of Training Purposes.� European Committee on Migration (CDMG), 306 Sherry Stephenson and Gary Hufbauer Strasbourg. http://www.coe.int/t/dg3/migration/Documentation/Legal Migration Policy Institute, Washington, DC. http://www.migrationpol- _texts/CDMG_96_18e_Temporary_migration_employment_training icy.org/pubs/Insight-IGC-Sept08.pdf. _purposes_en.pdf. Nielson, Julia. 2003. “Labor Mobility in Preferential Trade Agreements.� In Hamilton, Bob, and John Whalley. 1984. “Efficiency and Distributional Moving People to Deliver Services, ed. Aaditya Mattoo and Antonia Implications of Global Restrictions on Labor Mobility.� Journal of Carzaniga, 93–113. Washington, DC: World Bank. Development Economics 14: 61–75. OECD (Organisation for Economic Co-operation and Development). Hatton, Timothy, and Jeffrey G. Williamson. 1998. The Age of Mass 2007. International Migration Outlook. Paris: OECD. Migration: Causes and Economic Impact. New York and Oxford, U.K.: ———. 2008. West African Mobility and OECD Migration Policies. Oxford University Press. Paris: OECD. http://www.oecd.org/document/29/0,3343,en_38233741 IMF (International Monetary Fund). 2010. Balance of Payments Manual, _38247095_41481445_1_1_1_1,00.html. Sixth Edition. Washington, DC: World Bank. Persin, Daniela. 2008. “Free Movement of Labor: UK Responses to the Iregui, Ana. 1999. “Efficiency Gains from the Elimination of Global Eastern Enlargement and GATS Mode 4.� Journal of World Trade Restrictions on Labor Mobility: An Analysis Using a Multiregional 42 (5): 837–64. CGE Model.� Estudios Economicos, Banco de la República, Bogota. Pritchett, Lant. 2006. Let Their People Come: Breaking the Gridlock on Global Karsenty, Guy. 2000. “Assessing Trade in Services by Mode of Supply.� In Labor Mobility. Washington, DC: Center for Global Development. GATS 2000: New Directions in Services Trade Liberalization, ed. Pierre Sáez, Sebastian. 2009. “The EU EPA Negotiations.� Presented at a World Sauvé and Robert M. Stern. Washington, DC: Brookings Institution Bank training course on services, Washington, DC, March 21. Press. Soesastro, Hadi. 2005. “Accelerating ASEAN Integration: Moving Beyond Martin, Philip. 2007. Towards Effective Temporary Worker Programs: Issues AFTA.� CSIS Economics Working Paper Series 091, Centre for Strategic and Challenges in Industrial Countries. International Migration Papers and International Studies, Jakarta. http://www.csis.or.id/publications 89. Geneva: International Labour Office. http://www.ilo.org/public/ _paper_view.asp?id=52&tab=1. english/protection/migrant/download/tempworkers_martin_en.pdf. Stephenson, Sherry. 2007. “Mobility of Service Providers in the Caribbean Martin, Susan, and B. Lindsay Lowell. 2008. “Examining Labor Mobil- Region.� Box 13.6 in L. Alan Winters, “The Temporary Movement of ity Provisions for the United States.� Presented at the seminar on Workers to Provide Services: GATS (Mode 4).� In A Handbook of Inter- “Demographic Change and International Labor Mobility in the national Trade in Services, ed. Aaditya Mattoo, Robert M. Stern, and Asia Pacific Region: Implications for Business and Cooperation,� Gianni Zanini, 535–37. Oxford, U.K.: Oxford University Press. organized by Korea National Committee for Pacific Economic Coop- ———. 2008. “Demographic Change and International Labor Mobility in eration (KOPEC) and the Pacific Economic Cooperation Council the Pacific Americas—Issues, Policies and Implications for Coopera- (PECC), Seoul, March 24–26. http://www.pecc.org/component/ tion.� In Labor Mobility in the Asia-Pacific Region, ed. Graeme Hugo and eventlist/details/121-kopec-demographic-change-and-international- Soogil Young, 171–201. Singapore: Institute of Southeast Asian Studies. labor-mobility-in-the-asia-pacific-region. Vis-Dunbar, Damon, and Henrique Suzy Nikiema. 2009. “Do Bilateral Mattoo, Aaditya, and Antonia Carzaniga, eds. 2003. Moving People to Investment Treaties Lead to More Foreign Investment?� Investment Deliver Services. Washington, DC: World Bank. Treaty News, April 30. International Institute for Sustainable Develop- Mattoo, Aaditya, and Pierre Sauvé. 2010. “Regionalism in Services Trade.� ment, Winnipeg, Manitoba, Canada. in A Handbook of International Trade in Services, ed. Aaditya Mattoo, Walmsley, Terri Louise, and L. Alan Winters. 2002. “Relaxing the Restric- Robert M. Stern, and Gianni Zanini. 2nd ed. Oxford, U.K.: Oxford tions on the Temporary Movement of Natural Persons: A Simulation University Press. Analysis.� CEPR Discussion Paper Series 3719, Centre for Economic Meyers, Deborah W. 2006. “Temporary Worker Programs: A Patchwork Policy Research, London. Policy Response.� Task Force Insight 12, Migration Policy Institute, Winters, L. Alan. 2008. “The Temporary Movement of Workers to Provide Washington, DC. http://www.migrationpolicy.org/ITFIAF/TFI_12 Services (GATS Mode 4).� In A Handbook of International Trade in _Meyers.pdf. Services, ed. Aaditya Mattoo, Robert M. Stern, and Gianni Zanini, Moses, Jonathon, and Bjorn Letnes. 2004. “The Economic Cost to Inter- 480–96. Oxford, U.K.: Oxford University Press. national Labor Restrictions: Revisiting the Empirical Discussion.� World Bank. 2006. Global Economic Prospects 2006: Economic Implications World Development 32 (10): 1609–26. of Remittances and Migration. Washington, DC: World Bank. Newland, Kathleen, Dovelyn Rannveig Agunias, and Aaron Terrazas. ———. 2010. Global Economic Prospects 2010: Crisis, Finance, and 2008. “Learning by Doing: Experiences of Circular Migration.� Insight, Growth. Washington, DC: World Bank. 14 Investment Sébastien Miroudot A distinguishing feature of the current wave of globalization control (direct investment). Bilateral rules on investment, is that trade and investment are more and more inter- by contrast, generally rest on a broader, asset-based defini- twined. Advances in technology and increased liberaliza- tion that extends to portfolio investment and various forms tion of trade in the second half of the twentieth century of tangible and intangible property (e.g., real estate). have reduced trade costs and allowed the “unbundling� of Following the example of the North American Free the production process (Baldwin 2006). Instead of produc- Trade Agreement (NAFTA) in 1994, most PTAs combine ing in a single country, firms can offshore specific tasks and provisions on the protection and promotion of investment, fragment production internationally, in vertically inte- as traditionally found in BITs, with provisions on the liber- grated chains (Jones and Kierzkowski 2001). A global pro- alization of foreign investment and comprehensive trade in duction network enables firms to cut costs by taking services disciplines. These agreements jointly liberalize advantage of differences in technology and factor prices trade and investment and are more advanced than multi- across countries, leading to productivity gains that can lateral rules. They take into account company strategies translate into higher income for all the countries involved that combine foreign direct investment (FDI) and trade to in the global supply chain. This fragmentation of produc- create global supply chains that maximize productivity tion requires liberalization of both trade and investment. through the distribution of production among a number As a consequence of these trends, countries are increas- of countries. ingly incorporating investment, traditionally dealt with in From a political-economy perspective, there are several separate bilateral investment treaties (BITs), into their pref- motivations for negotiating disciplines on trade and erential trade agreements (PTAs). Not surprisingly, the investment simultaneously. The first is that PTAs empha- number of new BITs concluded has been decreasing since size market access and investment liberalization more than 2001 (UNCTAD 2008), while the number of PTAs with they do investment protection and promotion. With the investment provisions has been rising (figure 14.1).1 exception of agreements signed by the United States and, in Another reason for the slowdown in concluding new BITs some cases, by Canada and Japan, BITs do not include is that more than 2,000 BITs are already in force, and most nondiscrimination provisions for the preestablishment bilateral relationships are already covered. stage. Moving from BITs to PTAs implies opening of mar- The inclusion of investment provisions in PTAs has kets and negotiations similar to those occurring in trade. given birth to a new type of trade agreement that has no Concessions on investment can be balanced by concessions equivalent at the multilateral level. No multilateral agree- on trade within the same agreement, and countries can ment on investment exists, and rules on investment at the deal in a more comprehensive way with market opening World Trade Organization (WTO) are limited to those in issues that today involve both trade and investment. the General Agreement on Trade in Services (GATS), on the Another difference between BITs and PTAs is that PTAs supply of services following an investment (commercial cover an indefinite period, whereas BITs have a limited life presence, mode 3), and in the Trade-Related Investment span—10 years, for example. Unlimited commitments on Measures (TRIMS) Agreement (box 14.1). GATS, moreover, trade and investment make PTAs more appealing for firms relies on an enterprise-based definition of investment and seeking opportunities in new markets. Policy makers may thus applies only to business or professional establishments also be interested in a long-term commitment, especially in which the investor has majority ownership or exercises where the ratification of trade and investment liberalization 307 308 Sébastien Miroudot Figure 14.1. Total Number of PTAs and Number with Investment Provisions, 1970–2009 200 180 160 140 120 number 100 80 60 40 20 0 70 72 74 19 6 78 80 82 19 4 86 19 8 90 92 19 4 96 98 00 02 04 06 2008 09 7 8 8 9 19 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 number of PTAs in force (active and notified to WTO) PTAs with investment provisions Source: World Trade Organization database and calculations by the author. Note: PTA, preferential trade agreement; WTO, World Trade Organization. Box 14.1. Rules on Investment at the WTO World Trade Organization (WTO) rules on investment are found in the General Agreement on Trade in Services (GATS), which covers the supply of services following an investment (mode 3, commercial presence) and in the Trade-Related Investment Measures (TRIMS) Agreement. GATS and mode 3 GATS includes schedules for specific commitments and lists of exemptions from most favored nation (MFN) treatment submitted by member governments. Only by reference to a country’s schedule and (where relevant) its MFN exemption list is it possible to know to which services sectors, and under what conditions, the basic GATS principles of market access, national treatment, and MFN treatment apply within that country’s jurisdiction. The schedules are complex documents in which the country identifies the services sectors to which it will apply the market access and national treatment obligations of the GATS, as well as any exceptions from those obligations it wishes to maintain. The commitments and limitations are in every case entered with respect to each of the four modes of supply that constitute the definition of trade in services in Article I of GATS: cross-border supply, consumption abroad, commercial presence, and presence of natural persons. Commercial presence (mode 3) describes the conditions under which foreign services suppliers may establish, operate, or expand a commercial presence, such as a branch, agency, or wholly owned subsidiary, in the member’s territory. TRIMS The TRIMS Agreement provides that no contracting party shall apply any trade-related investment measure inconsistent with Articles III (national treatment) and XI (prohibition of quantitative restrictions) of the General Agreement on Tariffs and Trade (GATT). An illustrative list of measures agreed to be inconsistent with these articles is appended to the agreement. The list includes measures which require particular levels of local procurement by an enterprise (local content requirements) or which limit the volume or value of imports that such an enterprise can purchase or use to an amount related to the level of products it exports (trade-balancing requirements). The TRIMS Agreement mandates notification of all nonconforming trade-related investment measures and requires their elimination within two years by developed countries, within five years by developing countries, and within seven years by least- developed countries. It establishes a committee that monitors the implementation of these commitments, among its other responsibilities. The agreement also provides for consideration, at a later date, of whether broader provisions on investment and competition policy should be added. Source: World Trade Organization. Investment 309 treaties is not politically easy. The initial political cost might containing substantive investment provisions. As discussed be higher for the PTA, but the issue does not have to be below (in “Three Types of PTA�), these PTAs fall into three reopened periodically, and once the agreement is con- categories: “NAFTA-inspired� agreements that tend to cluded, investors benefit from long-term protection and a reproduce the architecture of NAFTA; “GATS-inspired� long-term guarantee of nondiscrimination. Given the stale- agreements in which investment provisions for services are mate of the Doha Development Agenda, with no multilat- in a separate chapter influenced by GATS; and “hybrids� eral agreement on investment and no progress on the liber- that combine the other two approaches. The table also alization of services trade, it is through PTAs that countries highlights the wide range of bilateral investment measures can expect to improve disciplines on investment. that are found in these PTAs. Finally, because most of the remaining restrictions on investment are in the area of services, countries can find Typology of Investment Measures some advantage in dealing with bilateral investment liber- alization in PTAs that grant preferential treatment under We first review the typology of investment provisions, GATS Article V and for which no most favored nation examining, in turn, establishment, nondiscrimination, (MFN) exemption is needed. Countries thus ensure that investment regulation and protection, dispute settlement, their bilateral obligations are not multilateralized and that and investment promotion and cooperation. consistency between investment rules and WTO obliga- Establishment. Establishment is the most critical com- tions is guaranteed.2 ponent of market access for foreign investors. Concretely, The increase in the number of PTAs with investment establishment refers to either the creation of a new firm in provisions is thus driven by both economic and legal con- the host country (greenfield investment) or the acquisition siderations. The remainder of this chapter examines PTAs of an existing firm (through mergers and acquisitions). and how their provisions have changed the landscape of Provisions on establishment define the conditions under investment rules. Because PTAs are being compared with which foreign firms can invest, starting with the definitions multilateral rules in GATS and bilateral provisions in BITs, of “investor� and “investment.� The scope of the agreement the full overview of disciplines that affect investment is depends on these definitions; all subsequent rules will presented. The focus is on services and GATS because more apply only to investment covered in the PTA. than 60 percent of current investment is in services and Generally, PTAs do not give investors a generic and because GATS, although not explicitly an international unlimited right to set up a permanent presence in the host investment agreement, has influenced the architecture and country. Only some agreements signed by the European nature of bilateral investment measures found in PTAs. Union (EU) mention a right of establishment. In GATS- The next section looks at the nature and scope of invest- inspired agreements, provisions on establishment are ment provisions in PTAs. Key issues related to the extent of included under the market access principle. In NAFTA- investment liberalization are then discussed; they include inspired agreements, a reference to market access can also be the risk of an investment “spaghetti bowl,� the scope for found in some PTAs, but establishment is generally dealt MFN guarantees in this accumulation of competing rules with through nondiscrimination provisions relating to the and commitments, and the degree of investment protection preestablishment phase. Investors are granted national treat- in PTAs. The final sections examine the economic impact of ment and most favored nation treatment under the condi- investment provisions in PTAs, from both a theoretical and tions for establishment, as discussed later in the chapter. an empirical perspective, and present conclusions. From the point of view of investors, there is no signifi- cant difference between these approaches. All agreements acknowledge the right of governments to regulate and to Nature and Scope of Investment list reservations that allow countries to restrict establish- Provisions in PTAs ment in specific sectors or under specific circumstances. In Investment provisions in PTAs are generally deep and addition to general exceptions—including measures relat- comprehensive.3 There are instances of agreements that ing to the general public interest, such as national security cover investment marginally, in a few articles that contain and public health, order, and morals—agreements also no binding commitment but only refer in general terms to include industry-specific limitations. the promotion of investment between the parties. These Nondiscrimination. Nondiscrimination provisions limit agreements, however, are not representative of the recent the discretion of the host country in distinguishing among trend toward PTAs that cover investment through a full categories of companies in applying regulatory and restric- package of disciplines and liberalization commitments. tive measures. National treatment means that domestic and Annex table 14A.1 provides an overview of recent PTAs foreign-owned companies must be treated the same; most 310 Sébastien Miroudot favored nation treatment precludes discrimination among stances,� this is also the interpretation of any MFN clause foreign companies. National treatment in the preestablish- in international law: the more favorable treatment can only ment phase can be particularly effective in liberalizing be granted to the “same subject matter,� according to the investment because it permits foreign investors to set up ejusdem generis (“of the same kind�) principle (see OECD operations on an equal footing with domestic investors. 2005, ch. 4). The wording of national and MFN treatment Although MFN treatment is a core provision in any clauses varies among agreements and can matter for their international investment agreement, not all PTAs grant it implementation. In addition, PTAs list limitations to for investment. There are two types of MFN provisions in a national treatment, as well as exemptions from MFN treat- preferential agreement. First, the provision can state that an ment, at the industry level, and the detailed analysis of investor from a party or its investment should be treated no these reservations defines the extent of the preferential less favorably than an investor from another party to the treatment granted in the agreement. agreement. The provision ensures the same treatment for Investment regulation and protection. Bilateral invest- all parties within the agreement. This type of provision is of ment treaties traditionally contain provisions that limit the little relevance in the case of a bilateral agreement (and ability of governments to restrict the activities of investors most PTAs with investment provisions are bilateral). The or to expropriate their investments. The most extensive second type of MFN provision is called “nonparty� or package of such provisions can be found in the investment “third-party� MFN treatment and grants to investors of chapters of NAFTA and of subsequent NAFTA-inspired other parties or to their investment a treatment no less agreements. favorable than the one granted to any nonparty. The aim is A core provision is free transfer of funds, which allows to guarantee that better treatment offered to a third party, for the unrestricted flow of investment-related transactions through, for example, another PTA, is extended to the par- and capital movements. These transfers include all kinds of ties of the first agreement. This second type of MFN provi- fees and returns to investment, such as profits, dividends, sion can be seen as an efficient liberalization mechanism capital gains, royalty payments, management fees, technical that extends new commitments found in recent agreements assistance, payments related to the liquidation of the to parties of former PTAs. Omission of an MFN treatment investment, payments made pursuant to a loan, and pay- provision in a PTA is one way of preventing the multilater- ments arising out of dispute settlement. alization of investment obligations and can lead to invest- Fair and equitable treatment is another standard drawn ment diversion when some PTAs are more preferential than from customary international law. In contrast to national other investment agreements. (See the discussion of multi- treatment and most favored nation treatment, which are lateralization, below.) comparative standards based on the treatment afforded To the extent that domestically owned companies different groups of investors, fair and equitable treatment receive better treatment than foreign-owned ones, national is an absolute standard. treatment can be seen as the important discipline and A third set of rules pertains to expropriation. Because MFN treatment can be regarded as superfluous. Some the threat of expropriation or nationalization can discour- countries, however, provide investment incentives such as age investors, a core rule in investment agreements is that tax holidays or subsidies that target foreign investors any expropriation should take place on a nondiscrimina- specifically and are not offered to domestic companies. In tory basis and with adequate compensation. that case, MFN treatment is better than national treatment Two other types of provisions can have a positive for the investor. Accordingly, some agreements, such as impact on investors and help countries promote FDI. First, NAFTA, in what is termed “standard of treatment� provi- provisions prohibiting performance requirements prevent sions, specify the “better of the treatment� required by a country from imposing on investors measures that national and MFN treatment. would, for example, force them to source locally. Such local There are qualifications to the application of national content requirements can be highly trade restrictive, in and MFN treatment. To begin with, national and MFN addition to discouraging investment.4 In the context of treatment are generally granted “in like circumstances.� global production networks, these requirements are likely The foreign investor and the domestic investor (or another to leave a developing economy outside supply chains, as the foreign investor, in the case of MFN treatment) have to be objective of such networks is precisely to fragment the pro- in comparable situations, and whether more favorable duction process and to move inputs across countries. This treatment has been accorded is analyzed case by case. is an area in which trade rules and investment rules overlap Although some agreements (in particular, those signed by in the TRIMS Agreement and in its list of prohibited the United States) specifically mention “in like circum- performance requirements. Investment 311 Second, provisions on temporary entry and stay for key Investment promotion and cooperation. The last category personnel facilitate investment, as there is hardly any FDI of investment provisions focuses on the promotion of that does not imply movement of people, in addition to investment and the harmonization of certain rules. Rather movement of capital. Because these provisions touch on than lay down binding commitments or disciplines pro- sensitivities about migration laws, disciplines are very spe- tecting investors, these provisions tend to organize cooper- cific and are limited to investment-related temporary ation between countries through exchange of information migration. or the creation of specific bodies or commissions. The Dispute settlement. All PTAs include means for resolving agreement may mention a general objective of harmoniza- disputes between states concerning the interpretation and tion of investment rules and policies, but without specific implementation of the agreement. A smaller number of provisions on how to achieve harmonization. Such provi- agreements introduces an investor-state dispute settlement sions also include clauses foreseeing the future liberaliza- mechanism that enables investors to make claims and tion of investment. defend their interests directly. Such a mechanism can pro- vide an effective way of implementing investment provi- Three Types of PTA sions because it allows foreign investors to seek redress for damages resulting from host governments’ breaches of Although investment provisions in PTAs are diverse, two their obligations. The investor-state dispute settlement models can be identified, with very different approaches to mechanism relies either on ad hoc arbitration involving an the way investment commitments are scheduled and to the independent arbitrator, generally under the rules of the overall architecture of agreements. The first model is United Nations Commission on International Trade Law NAFTA, in which investment is dealt with in a single chap- (UNCITRAL), or on permanent arbitration, typically ter that covers both investment in goods and investment in through the International Center for the Settlement of services and that includes broad provisions on the regula- Investment Disputes (ICSID). tion and protection of investment, in addition to liberaliza- The investor-state dispute settlement provision comes tion commitments. In the second model, investment provi- from investment treaties and borrows its main elements sions are split between two chapters: an investment chapter from the system of commercial arbitration. In the universe that sets out disciplines relevant for all investments (in of trade agreements, by contrast, the paradigm is state- goods and services industries) and a services chapter that state resolution of disputes, as exemplified by the WTO contains liberalization commitments relevant for services, dispute settlement mechanism (Choi 2007). One of the as well as other provisions on trade in services. This second first PTAs to include direct claims by investors was NAFTA. model can be said to be GATS-inspired in that the provi- It is interesting that provisions which had been part of sions for services follow the approach of that WTO agree- investment treaties for a long time became controversial in ment. A third type of “hybrid� agreement combines the the context of the implementation of a trade agreement NAFTA and GATS approaches. with investment provisions. Although many PTAs now NAFTA-inspired agreements. The architecture of NAFTA contain an investor-state dispute settlement provision, is characterized by a clear separation between the invest- some agreements modeled after NAFTA lack a mechanism ment chapter and the chapter on cross-border trade in of this kind; an example is the Australia–United States Free services. Provisions relevant for investment in services and Trade Agreement (Dodge 2006). Even when the provision trade in services according to mode 3 (commercial pres- is present, negotiators have sometimes been reluctant to ence) are part of the investment chapter; provisions extend its coverage to new investment-related disciplines regarding financial services are dealt with in a separate found in trade agreements. In particular, in agreements chapter. The definitions of investor and investment are far- containing a separate chapter on services that covers mar- reaching. Investment is not limited to FDI but also covers ket access and national treatment provisions for invest- some types of portfolio investment, such as equity securi- ment in services, investor-state dispute settlement does ties, as well as property, such as real estate. not apply. NAFTA provides for national and MFN treatment in both So far, very few state-state disputes have been filed the preestablishment and postestablishment phases. Nondis- under PTAs; countries prefer to bring disputes to the WTO crimination disciplines apply to “the establishment, acquisi- (Morgan 2008). By contrast, the number of investor-state tion, expansion, management, conduct, operation, and sale disputes brought under international investment agree- or other disposition of investments.� In addition, the provi- ments has been rising, with a cumulative total of 255 sion on standard of treatment grants the better of national known cases at the end of 2006 (UNCTAD 2006). or MFN treatment, and the clause on minimum standard 312 Sébastien Miroudot of treatment provides for “fair and equitable treatment analogous nondiscrimination principles for services are and full protection and security.� The remainder of the found in a separate chapter. investment chapter covers all relevant protection disci- The treatment of investment is therefore influenced by plines, such as free transfer of funds, provisions on expro- services concepts that come from GATS. Whereas the priation and compensation, and investor-state dispute investment chapter includes an asset-based definition of settlement. investment, on the basis of which investment protection The intention of the agreement to be far-reaching can measures apply, the trade in services chapter does not also be seen in the way nondiscrimination commitments define investment but relies on the concept of commercial are treated. The approach is that of the negative list, presence (mode 3 in GATS). Commercial presence is the whereby national treatment and MFN treatment are supply of a service through a business or professional granted for all sectors, with two lists of reservations—one establishment in the territory of the country in which the for existing measures that are not conforming at the date of service is supplied. In the case of services, market access, the agreement, and another for future measures that gives national treatment, and MFN treatment apply only to the governments discretion to maintain or introduce commercial presence. restrictive measures after the agreement has entered into As in GATS, the MFN principle is a general obligation force. The negative-list approach is generally seen as the and applies to all services sectors covered in the agreement. more favorable to liberalization in the sense that commit- Exemptions can be enumerated in a negative list. For ments are made for investments in all sectors, and only the market access and national treatment, GATS-inspired exceptions listed at the time the agreement is signed can be agreements reproduce the format of GATS schedules of maintained. There is, nonetheless, some discretion left for commitments. There is a positive list of sectors for which governments because of the practice of listing reservations specific commitments are made, and limitations are listed for future measures. These reservations are thus not too for these commitments. Commitments and limitations are different from unbound commitments under the GATS listed by mode of supply, according to GATS typology. approach. In sectors in which reservations are made Some authors have pointed out that the GATS approach regarding future measures, there is a possibility of intro- toward the scheduling of commitments lacks transparency ducing new restrictive measures that are not listed in the (Hoekman 1996; Stephenson 2002; Mattoo 2005). For agreement. example, when sectors are unbound, there is no indication Empirically, negative-list PTAs are found to go further of the restrictive measures that apply. These sectors have to in their sectoral coverage (Fink and Molinuevo 2008). The be identified by deduction, as only the ones with specific proliberalization approach stems from another mechanism commitments are listed. The same criticism can apply to that complements the negative listing of restrictions. GATS-inspired PTAs. An advantage of these PTAs, how- NAFTA-inspired agreements not only lock in the invest- ever, is that they ensure the consistency of regional liberal- ment regime but also include as commitments under the ization with multilateral disciplines. When PTAs explicitly PTA any new measure taken unilaterally by the parties that refer to GATS articles to define the concepts that organize moves in the direction of liberalization. This creates a the liberalization of mode 3, it is easier to assess the prefer- ratchet effect because investment restrictions, once ential content of the PTA and to be sure that the provisions removed, cannot be reintroduced. are not “GATS-minus� (less favorable than in GATS). The Preferential trade agreements signed by Canada, Mexico, analysis is more complicated in the case of NAFTA- and the United States (the three parties to NAFTA) are, inspired agreements that are based on different concepts. not surprisingly, NAFTA-inspired. The approach has also The straightforwardness and familiarity of the GATS been exported to Asia, where a significant number of approach may explain why it has been popular among PTA bilateral agreements follow the NAFTA model. Examples negotiators. They are on accustomed ground when working include the Chile–Republic of Korea, Korea–Singapore, with GATS-like schedules of commitments, and they can and Singapore–Australia free trade agreements (FTAs).5 more easily strike a deal on the basis of existing multilateral GATS-inspired agreements. GATS-inspired agreements commitments augmented by preferential commitments. contain an investment chapter that covers all investments Hybrid approaches. Most agreements can be clearly and provides for the protection of investment through identified as GATS-inspired or NAFTA-inspired, but some disciplines as far-reaching as those in NAFTA-inspired depart from these two models. A very small number has agreements. The difference is that the provisions of the adopted a scheduling approach that is modeled after nei- investment chapter concerning market access, national ther. This is the case, in particular, for the agreements treatment, and MFN treatment apply only to goods. The signed in the context of the European integration process, Investment 313 such as EU enlargement agreements that are notified to the further elaborated. In particular, questions remain about WTO but are of a different nature from other PTAs. There the degree of liberalization achieved through the kind of are also PTAs without deep investment provisions in which disciplines found in PTAs. Moreover, the proliferation of investment promotion is mentioned only as a general international investment agreements and the coexistence objective, without any specific commitment. of bilateral, plurilateral, and multilateral rules are key Among PTAs with deep investment provisions, how- issues in the analysis of investment provisions in PTAs. ever, a significant number tends to combine the GATS and How is bilateral investment liberalized in PTAs? Is there an NAFTA approaches. These are designated as “hybrid� in investment “spaghetti bowl� comparable to the one for annex table 14A.1. First, most of the agreements signed trade? What exactly is the role of MFN provisions in this recently by Japan (with the exception of that with Chile) accumulation of competing rules and commitments? As deal with both commercial presence and investment in PTAs progressively replace BITs, will they offer the same services and include a GATS-like schedule of commit- degree of investment protection as do BITs? ments, as well as a negative list of nonconforming meas- ures. In coverage, they are comparable to NAFTA-inspired Liberalization of Bilateral Investment in PTAs agreements in that they employ a broad, asset-based defini- tion of investment that is extended to services with respect It is common to refer to the liberalization of investment to national and MFN treatment (unlike GATS-inspired when discussing bilateral investment measures. As a gen- agreements) and include the mechanisms of the negative-list eral objective, most PTAs explicitly mention in their pre- approach, such as the ratchet effect. At the same time, consis- ambles the liberalization, expansion, and promotion of tency with GATS commitments is improved. Restrictions on bilateral investment. Nevertheless, investment is liberalized investment in services appear twice in these agreements: in in these agreements mainly through the nondiscrimination the schedule of commitments for services under mode 3, principles described above. Countries make commitments and in the list of nonconforming measures in the invest- to grant the same treatment to foreign investors and ment chapter. The ratchet mechanism is transposed onto domestic investors, or to different foreign investors, but the GATS-inspired schedule of commitments through an they undertake no commitment to remove all existing bar- additional column in which Japan commits to bind any riers to investment. For example, countries can maintain new liberalization measure in specific subsectors. an investment screening process through which they Another hybrid approach can be found in the Australia– authorize or reject an investment, or they can impose spe- Chile FTA, which is mostly NAFTA inspired but in which a cific conditions before approving the investment. To the reservation on market access in the negative list reproduces extent that the same conditions apply to domestic and for- the market access column usually found in GATS sched- eign firms, there is no discrimination, yet the investment ules. The architecture is quite convoluted: the agreement regime can still be quite restrictive. In addition, countries first grants market access for all investments but then adds generally maintain discriminatory measures through a reservation for all services (negative listing) and an exemptions, reservations, and limitations with respect to exception for a positive list of subsectors in a GATS format national and MFN treatment. With the exception of provi- that includes reservations. sions in a few agreements for phasing out commitments, It is not certain that these agreements, with their com- these limitations are generally meant to stay, at least until plex structure, really afford the transparency that investors further negotiations occur. PTAs generally have mecha- require. Their intent, however, is to combine the best of nisms for “encouraging� the progressive removal of remain- two worlds: the advantages of the negative list of NAFTA- ing barriers but do not commit to their full elimination. inspired agreements, in which all restrictive investment One important achievement in PTAs, in comparison measures are listed, and the consistency of GATS-inspired with BITs, is that they include disciplines on market access.6 agreements with respect to multilateral commitments. In In most BITs, national treatment and MFN treatment cover the future, such a combination could be useful in multilat- only the postestablishment phase, and there are no nondis- eralizing PTAs and harmonizing the treatment of investors crimination obligations concerning the establishment of and investment. companies. Disciplines on market access are most impor- tant for the ability of foreign investors to enter the market and for the creation of a level playing field for domestic and Key Issues foreign companies. In NAFTA-inspired agreements, these Some of the issues arising from the existence of two sets of disciplines take the form of national and MFN treatment disciplines were touched on above, but they need to be obligations toward the “establishment, acquisition, and 314 Sébastien Miroudot expansion� of investments, whereas GATS-inspired agree- return them to the status quo found in GATS-inspired ments have a combination of market access and national agreements. For example, it is not uncommon to find a treatment commitments for services. The difference is that reservation for future measures saying that the country can market access in the sense of GATS Article XVI is defined adopt or maintain any measure related to new services. by means of a list of six different types of prohibited quan- (Japan tends to introduce such measures in its PTAs.) titative restrictions (see Delimatsis and Molinuevo 2008). Reservations for future measures give countries the oppor- Some of these restrictions (e.g., foreign equity limits) are tunity to avoid the disciplines of the agreement in as many discriminatory and are also prohibited by the national sectors as they wish. treatment principle, but others, such as quotas that apply This caution is certainly unavoidable if the agreements indifferently to domestic and foreign investors, are nondis- are to be signed and politically accepted. Legitimate con- criminatory. The latter restrictions are not covered by cerns related to public interest or national security also national treatment obligations because they do not involve lead to restrictions on foreign investment. It should be rec- discrimination between domestic and foreign companies.7 ognized, however, that agreements already include safe- In both types of agreement, countries can still maintain guards to deal with these concerns. reservations or carve out specific sectors, but in sectors in Restrictions on foreign investment are also generally which commitments are made, existing measures are found in services sectors that have traditionally been regu- locked in, creating a standstill. Countries cannot intro- lated and in which market imperfections can in some cases duce new restrictive measures on investment in these sec- justify the existence of restrictions. In addition, there are tors. This kind of guarantee is valuable for investors and many restrictions at the subfederal or subnational level that generally has a positive impact on investment decisions, are not always clearly identified.8 especially when the commitment is embedded in an inter- An approach more favorable to the liberalization of national agreement. investment would be to add, in addition to the standstill, In the case of NAFTA-inspired agreements, there is, in the rollback principle So far, only the Organisation for Eco- addition, the ratchet effect mentioned above. If a party nomic Co-operation and Development (OECD) Codes of decides to remove unilaterally some barriers to investment, Liberalisation of Capital Movements and Current Invisible the standstill of the agreement will apply to this new Operations include such a principle. The codes provide for regime, and the country cannot revert to the restrictive the periodic examination of remaining restrictive meas- measures—or, if it did, it would have to offer compensation ures: countries have to justify the continuation of their and renegotiate with the other parties. This is clearly a pro- measures, and other countries can try to persuade them to liberalization effect that is attached to NAFTA mechanisms remove these barriers or suggest less restrictive ways to rather than to the negative list. A similar ratchet effect can meet the objectives cited to justify the measures (OECD easily be incorporated into GATS-inspired agreements, as 2008). In the case of NAFTA, there is an element of rollback demonstrated in some of the PTAs signed by Japan. in the inclusion of phase-out commitments—a proviso An advantage specific to the negative listing is the auto- whereby nonconforming measures are in force for a lim- matic inclusion of new sectors under the disciplines of the ited number of years and then disappear. Although agreement. As technological innovations take place, it is NAFTA-inspired agreements include a ratchet mechanism, not unusual to see the emergence of new activities that do this mechanism is based on unilateral liberalization, with not fit into old classifications (such as new services pro- no obligation or commitment to achieve further progress. vided online), and there may be legal uncertainty about the GATS-inspired agreements often include clauses on future application of national treatment or MFN treatment to liberalization, with a commitment to review the agreement these innovations. In a negative-list agreement, these sectors (say, after three years), but in practice, there are few exam- are automatically covered, as the only reservations are ples of agreements in which commitments have been those listed when the agreement was signed. This is not the updated. case in GATS-inspired agreements, in which no specific From firms’ point of view, what ultimately matters is the commitment could have been made for these sectors, domestic investment regime. Countries can maintain very unless they clearly fit into an existing sector. restrictive measures such as public monopolies, discrimi- One should not overemphasize these differences natory licensing procedures, or foreign equity limits while between NAFTA-inspired and GATS-inspired agreements. still signing comprehensive PTAs. It is mainly through In principle, the former should be more favorable to the unilateral liberalization that firms can expect an improve- liberalization of investment. In practice, the NAFTA-style ment in market access for their investments. Sectors in agreements contain very broad reservations that often which investment is regulated are generally reformed not Investment 315 when a PTA is signed but, rather, when there is a domestic Rules of origin in PTAs define the national origin of consensus for reform. In that situation, all foreign coun- products traded in order to determine whether the prefer- tries are likely to benefit from new liberalization efforts, ential treatment of the agreement applies to these products. independent of PTAs. In the case of investment, rules of origin apply to juridical It is important to understand that PTAs with respect persons (the companies) and define the conditions under to investment offer only preferential binding and not which an investor from a third country can benefit from the always preferential treatment. To begin with, the domestic investment provisions contained in the PTAs signed by the investment regime can be more liberal than is suggested country where it has established. These rules are found by the limitations listed in the PTA.9 The limitations do either in the definition of the investor and the conditions of not mean that the country actually has investment incorporation or in clauses on denial of benefits that specify restrictions but only that it is allowed to maintain some who cannot benefit from the preferential treatment granted restrictions within the legal framework of the agreement. in the agreement. The schedules of commitments provide an indication of Rules of origin for investment are generally quite liberal the bound level of investment restrictiveness. Another (Beviglia Zampetti and Sauvé 2006; Fink and Nikomborirak reason that PTAs are sometimes not preferential in prac- 2007). Most PTAs follow the GATS approach (which is tice is that the investment regime is not discriminatory. also common in BITs), whereby the juridical person has For trade in goods, it is easy to discriminate between to be incorporated under the law of the member party and trade partners through the application of different tariffs engaged in “substantive business operations� in that party. (and the tariff already constitutes discrimination between This rule is part of the conditions set in GATS Article V for domestic and foreign companies). In the case of invest- PTAs to be allowed as derogations to the MFN obligation ment, barriers are generally the same for all foreign coun- of GATS. For this reason, there is little variation across tries; there are few exceptions to MFN treatment.10 The PTAs on the rules of origin for juridical persons. barriers apply to foreign-owned companies established There are no criteria defining substantive business oper- under domestic law—even though the rationale for dis- ations. According to Emch (2006), that requirement is crimination is not very sound, as the companies generate aimed at ensuring that the link of the company to the terri- jobs in the local economy. tory of the party is “genuine.� It is loose enough to not be For the reasons discussed above, the sectoral coverage interpreted by countries as giving them the opportunity to of preferential investment provisions is generally quite add strict criteria in order for business operations to be extensive (Fink and Molinuevo 2008; Marchetti and Roy regarded as substantive. 2008; Miroudot 2009), although there are exceptions in The consequence of these liberal rules of origin is that key sectors. In “Economic Impact: Theory and Evidence,� the impact of divergent provisions across PTAs is reduced. below, we will see that these provisions have been found to If the investment regime is more liberal in one country have a positive economic impact. because of a PTA, companies from third countries can invest in that market through any subsidiary established in a country belonging to the PTA. A concrete example is A “Spaghetti Bowl� of Investment Provisions? provided by Fink and Mattoo (2002), who describe how, According to the United Nations Conference on Trade and following the adoption of NAFTA in 1994, several Spanish Development (UNCTAD 2009b), there were, at the end of and Dutch banks established in Mexico through their U.S. 2008, a total of 2,676 BITs, 2,805 double-taxation treaties subsidiaries. Financial liberalization had just started in (DTTs), and 273 international agreements, other than BITs Mexico, and U.S. (and Canadian) investors were the first and DTTs, that contained investment provisions; this last to be able to invest in the Mexican banking industry. category includes PTAs. International investment agree- When the EU–Mexico agreement entered into force in ments are therefore proliferating even more than trade 2000, offering provisions similar to those in NAFTA for agreements, suggesting a giant “spaghetti bowl.� One financial services, some of these companies transferred the should, however, recall the exact meaning of this expres- ownership of their Mexican subsidiaries back to the parent sion introduced by Bhagwati (see Bhagwati, Krishna, and companies in Europe. This example illustrates how liberal Panagariya 1999). The “spaghetti bowl� is not the conse- rules of origin can compensate for the absence of MFN quence of the sheer number of agreements in force but of treatment. From the point of view of companies in third differences in the depth and patterns of sectoral liberaliza- countries, however, this is only a second-best situation tion in PTAs, combined with diverse sets of rules of origin because of the cost associated with investment through a that create high trade costs for firms. subsidiary. 316 Sébastien Miroudot Because of leaky rules of origin, the “spaghetti bowl� Unfortunately, most PTAs tend to put limits on the ben- problem is not too severe in the case of investment (Bald- efits of the application of the MFN principle to parties win, Evenett, and Low 2009).11 There are, nonetheless, when better treatment is offered to nonparties. First, some potential inconsistencies between the different types of PTAs (such as the Korea–Singapore FTA) simply omit international investment agreements when their provisions MFN treatment from their investment provisions. In other overlap. Inconsistency can also exist between provisions in agreements, there is an MFN clause, but other PTAs are PTAs and multilateral rules. In the case of investment, only excluded from the scope of the provision. Parties can grant services sectors have multilateral commitments in GATS. better treatment to third countries in another PTA without A certain number of PTAs are found to have GATS-minus any obligation to extend this better treatment to the parties commitments; they list nonconforming measures that have of the PTA that includes the MFN provision. This type of no equivalent in GATS. For example, in the economic part- provision is referred to as a regional economic integration nership agreement between the EU and the Caribbean organization (REIO) exception clause. It can be limited to Forum of African, Caribbean, and Pacific (ACP) States past agreements (as in the New Zealand–China FTA) or (CARIFORUM) that entered into force in 2008, measures extended to future agreements—the case with the FTA on subsidies are explicitly excluded from the scope of between the European Free Trade Association (EFTA) and application of market access and national treatment provi- Korea. The latter situation differs from an agreement sions. There is no similar carve-out in the GATS schedule without an MFN clause in that the other party has an of the EU, where only subsidies on research and develop- opportunity to negotiate the benefits granted in a new ment are excluded from national treatment with respect to PTA signed with a third country. When there is an REIO commercial presence.12 PTAs do not prevent GATS com- exception, the agreement does not prevent the new prefer- mitments from applying, and so these GATS-minus provi- ential treatment granted to third parties from being sions should have a limited impact, in practice. Neverthe- extended to the parties of the current PTA, but the exten- less, such inconsistencies create legal uncertainty and send sion is not automatic. It can be done through a request, a the wrong signals to investors. review of commitments, renegotiation, or a party’s unilat- eral decision, but the extension has to be agreed and made official in some kind of document complementing the Multilateralization of Investment Provisions original PTA. through MFN Clauses In NAFTA and in NAFTA-inspired PTAs, there is gener- The most favored nation clause is a pillar of international ally an exception to MFN treatment regarding commit- trade law and has historically been a powerful tool for ments made in PTAs signed earlier, but not for future multilateralizing bilateral commitments. For investment agreements. Instead of an REIO exception clause, the in services, there is a general MFN obligation that applies exception is listed as a reservation in the annex on future to all sectors in GATS and covers, as a consequence, all dis- measures or in a specific annex. Exceptions are also listed ciplines related to mode 3 trade in services. The MFN for future agreements. For example, in PTAs entered into principle is also found in bilateral investment treaties, but by the United States, four sectors are excluded from the its scope is limited to postestablishment when these bilat- application of the MFN principle to future agreements: eral agreements contain no provision on the liberalization aviation; fisheries; maritime matters; and telecommunica- of investment in the preestablishment phase. tions and transport services, including transport networks. A PTA is an exception to the MFN principle because it The scope of application of MFN can also be limited. grants preferential treatment to specific parties and breaks For example, PTAs often indicate that there is no require- the equality in the treatment of economic partners that is ment to extend dispute settlement procedures when these precisely the objective of MFN clauses. The inclusion of are more favorable in another PTA. MFN provisions within PTAs has two objectives. First, in Most BITs also contain an MFN exemption for PTAs the case of a regional or plurilateral agreement, it ensures (Adlung and Molinuevo 2008). In this case, parties to a that all parties to the PTA receive the same treatment. The BIT cannot expect to benefit from the more favorable relevance of MFN provisions is limited in the case of bilat- conditions granted by their partners in a PTA. There are, eral agreements, as there is only one partner, but MFN pro- however, BITs in which such an MFN exemption is not visions are useful with respect to nonparties (also called found. The question remains open as to whether commit- third parties). If a new PTA is signed, the parties to the for- ments made in the PTA should be extended to BIT par- mer agreement would like to benefit from any additional ties: the more favorable treatment offered to third parties preferential treatment found in the new PTA. can only be granted “in like circumstances� to the parties Investment 317 covered by MFN treatment, and differences in the scope the FTAs between the United States and Australia; Japan and objectives of the two types of investment agreement and Malaysia; and Korea and Singapore. could lead to interpretations limiting the extension of the 2. A BIT between the countries existed before the entry PTA disciplines to BIT parties. Another debate concerns into force of the PTA. Some countries decide to main- GATS Article II and whether the MFN clause in GATS tain the existing BIT alongside the PTA, especially when covers BITs. If it does, obligations incurred under any BIT the two sets of rules, in the PTA and in the BIT, tend to might have to be extended to all WTO members. There is, complement each other. The EFTA–Chile agreement however, no consensus about the interaction between BIT and most of the agreements signed by the EU follow this provisions and GATS, and there has been no dispute settle- approach. When the PTA is clearly superior to the exist- ment or case that could throw light on these issues (see ing BIT, with more comprehensive provisions, the latter Adlung and Molinuevo 2008). can simply be replaced. Either the BIT is terminated (as In conclusion, the role of the MFN clause in extending with the Australia–Chile FTA), or the PTA can suspend the benefit of PTAs to third countries seems limited. For part of the provisions of the BIT when it enters into some agreements, MFN mechanisms may apply, and new force (e.g., the U.S.–Morocco FTA). liberalization commitments can benefit nonparties, but this is not so in most cases. Moreover, studies show that When PTAs include an investment chapter, the provi- countries tend to negotiate similar commitments across sions on investment protection are generally equivalent to agreements (Fink and Molinuevo 2008; Marchetti and Roy the content of similar BITs. The fact that some agree- 2008; Baldwin, Evenett, and Low 2009) or to introduce ments are GATS inspired does not prevent investment in reforms unilaterally without discriminating between services from benefiting from the basic protections pro- investment partners. Fink and Jansen (2008) note that vided in the investment chapter. It should also be noted unilateral commitments on services (including invest- that, unlike the services chapter that refers to “commer- ment in services) are generally not discriminatory because cial presence,� the investment chapters of GATS-inspired countries seek to avoid the economic distortions associ- PTAs use a broader, asset-based definition of investment, ated with discrimination. and the provisions on the protection of investment apply There are, however, sectors in which governments are to this broad definition. “Commercial presence� is rele- found to discriminate; examples include air transport, vant in GATS-inspired PTAs only with respect to the level which is generally excluded from GATS and from PTAs, and of liberalization in services, as defined in the services financial services, where the scope of nondiscriminatory chapter and in the commitments annexed to the agree- disciplines in trade agreements is very limited. For these ment. On the basis of an analysis of 20 PTAs, Houde, sectors, the liberal rules of origin mentioned above might Kolse-Patil, and Miroudot (2007) show that there is no be more efficient than MFN clauses for circumventing dis- difference in the level of investment protection between criminatory policies. GATS-inspired and NAFTA-inspired PTAs. In conclusion, there is technically no impediment to the inclusion in PTAs of the provisions on investment protec- Investment Protection in BITs and in PTAs tion usually found in BITs, and countries have generally Provisions traditionally found in BITs now tend to be adopted the same kind of provisions in both types of included in trade agreements, and so a legitimate question agreement. The remaining differences between BITs and is whether PTAs provide the same degree of protection to PTAs result from the outcome of negotiations and the exis- foreign investors in terms of the extensiveness of the pro- tence of a BIT already in force before the PTA was negoti- tection provisions described in “Investment regulation and ated. Some countries have designed PTAs that can replace protection,� above. The answer depends primarily on the earlier BITs, while others maintain part of the guarantees type of PTA and the existence or absence of a BIT between offered to foreign investors in the BIT and have not dupli- the PTA parties. Two cases can be identified: cated the provisions in the PTA. 1. No bilateral investment treaty exists between the signa- Economic Impact: Theory and Evidence tories, and the PTA is their first investment agreement. The scope and depth of the rules introduced by the PTA The two preceding sections examined the architecture of depend on the outcome of the negotiation, but the PTA investment PTAs and the implications of their legal pro- can potentially include any of the protections that visions. Little was said about the economic impact of would be found in a BIT. Examples include NAFTA and bilateral investment measures. For a long time, this issue 318 Sébastien Miroudot has been overlooked, in large part because of lack of data costs can encourage trade rather than investment within and of methods for precisely assessing the impact of PTAs the region where the PTA is signed. Conversely, to the on flows of trade and investment. Recent developments extent that the PTA increases trade costs vis-à-vis third- in economic theory, as well as new empirical tools, have party countries, it can promote tariff-jumping investment to some extent enabled economic analysis to catch up. by these parties. An implication is that even an agreement There is now empirical evidence on the positive impact of with no investment provisions is likely to influence firms’ bilateral investment measures in PTAs. Studies have, decisions because of its impact on the trade-off between however, shown that not all PTAs improve the investment investment costs and trade costs. If, in addition, a PTA climate and that other determinants of FDI are important includes provisions that lower investment costs, it is likely as well and may condition the positive impact of PTA to further encourage FDI over arm’s length trade. provisions. Because of the link between trade costs and investment decisions, the concepts of trade creation and trade diver- sion can be transposed into the realm of investment. A Economic Benefits of PTAs with Investment PTA that increases trade barriers relative to third countries Provisions: Theory can be investment creating, by encouraging FDI from Trade and investment can be seen as two sides of market these third countries. The same PTA can be investment access. Firms have different means of serving foreign mar- diverting intraregionally because arm’s length trade is less kets; in particular, they can choose between exporting costly and firms liquidate former tariff-jumping FDI. (trade) and creating a subsidiary within the foreign econ- Restrictive rules of origin for goods within the PTA can omy for local production (international investment). The also lead to investment diversion, as manufacturers using recent literature on firm heterogeneity and global sourcing third-country inputs concentrate their production in the (Antràs and Helpman 2004; Helpman 2006) focuses on the country with the largest market and the lowest external choice between exports and FDI. Not all firms follow the tariffs (Estevadeordal and Suominen 2005). same path. Depending on their productivity, size, and The proximity-concentration trade-off is only one side structure of production, firms adopt different strategies of the complex relationship between trade and invest- that lead to different types of international activities such ment and is relevant only with respect to horizontal FDI. as offshoring, outsourcing, or vertical specialization. The In the case of vertical FDI and vertical specialization leading least productive firms tend to stay in the domestic market, to global value chains, trade and investment are comple- while more efficient companies can engage in international ments rather than substitutes. Vertical FDI is associated investment and become multinational enterprises (MNEs). with efficiency-seeking strategies whereby firms intend There are also differences across sectors in the way firms to benefit from locational advantages such as better or organize their production, based on product characteristics relatively cheaper factors of production or strategic and technologies. resources. In this context, a PTA is likely to increase FDI The “proximity-concentration tradeoff � (Brainard 1997) through both its trade and its investment provisions. In describes a substitution effect between trade and invest- particular, the reduction of barriers to imports of inter- ment. Market-seeking MNEs face trade costs when they mediate goods and services can encourage FDI (Ferran- export (all those costs related to sales overseas) but can tino and Hall 2001). save on production costs because of scale economies, as all The picture is further complicated by network effects the production is done in the home country. When they and third-country effects that arise because firms choose invest abroad and manufacture locally, MNEs no longer their production location from among several countries incur trade costs, but production costs can increase that belong to different regional integration schemes. because production is now split between the home coun- When economic integration is deepened in a group of try and the host, diminishing scale economies. The out- countries, trade liberalization can have a redistributive come of this trade-off depends on the relative sizes of effect on intraregional investment patterns; one country trade costs and investment costs. High border barriers might attract more FDI and other countries, less. such as tariffs may encourage the company to produce In brief, whether a PTA eventually increases or reduces close to the consumers in the foreign economy—that is, to FDI flows is an empirical question. The answer depends engage in tariff-jumping FDI. on the relative strength of the decreases in trade costs and In this theoretical framework, PTAs affect firms’ strate- investment entry costs, as well as FDI motives (efficiency gies at several levels. First, the trade liberalization provi- seeking or market seeking). Whether FDI is positive for sions in PTAs have an impact on trade costs. Reduced trade the host country, and for developing countries specifically, Investment 319 is a broader question that is beyond the scope of this chap- used in a gravity model to assess the role of certain types of ter. The answer is partly empirical, and an abundant liter- provisions that characterize new regionalism and, in partic- ature has discussed the gains (or lack of gains) from ular, investment provisions. Lesher and Miroudot (2007) FDI. Overall, however, FDI has consistently been found use a similar methodology to analyze the economic impact to be beneficial to developing countries, in particular of North-South PTAs and find that the more extensive the because it transfers knowledge as well as capital, produc- provisions on investment are, the higher are trade flows ing a long-term impact on economic growth.13 FDI also and, to an even greater extent, investment flows. leads to integration into world markets and is key to devel- More recent studies have tried to improve the theoret- oping economies’ participation in global value chains. ical framework by introducing a model that can better account for firms’ decision to trade or to invest. Miroudot (2009) tests an index of investment provisions in FDI and Empirical Studies trade equations derived from the knowledge capital Although economic theories have recently shed light on model, which introduces horizontal and vertical MNEs, firms’ decisions to invest or to export, few empirical studies as well as consideration of the skill endowments of coun- have investigated the role of PTAs in fostering trade and tries. The index of investment provisions is further investment. A wider literature exists on the impact of BITs refined by taking into account the sectoral coverage of on FDI flows. Annex table 14A.2 summarizes the results of investment provisions and the extent to which they go selected recent studies. beyond GATS in the case of services, in order to measure The earliest study to look at the impact of trade liberal- the preferential content of the agreement. The impact of ization on investment is that by Jeon and Stone (2000). PTA investment provisions is lower than in previous Focusing on the Asia-Pacific region, they find that the studies but is still positive and economically significant. Association of Southeast Asian Nations (ASEAN) has Dee (2006) proposes a model of “complex� FDI that also increased intraregional trade but that its impact on takes into account vertical and horizontal motivations for intraregional investment is insignificant. This study uses a FDI and introduces third-country effects so that the bilat- gravity model in which PTAs are included as dummy vari- eral investment relationship is influenced by the trade ables to identify differences between intraregional and and investment regime in other countries, in particular, extraregional bilateral investment flows. The dummy vari- neighboring countries. In the case of Asia, Dee argues able takes the value of 0 when there is no agreement and a that patterns of investment are already explained by fun- value of 1 when countries have signed a PTA. Using a simi- damentals and do not need the investment provisions of lar methodology, Hufbauer and Schott (2008) show that PTAs. When FDI and trade are not driven by size, income, the most prominent PTAs, including NAFTA, the ASEAN and other market characteristics, the impact of PTA pro- Free Trade Area, the EU, and EFTA, have increased bilateral visions is less obvious, but the author still finds positive investment stocks between their members. One exception effects for some economies. is the Canada–U.S. FTA, which preceded NAFTA and The most recent studies thus introduce nuances in the which seems to have led to decreased bilateral investment role of investment PTAs. Some of these PTAs are signed by stocks. This could be interpreted as the unwinding of for- countries that are “naturally� inclined toward increased mer tariff jumping between the United States and Canada bilateral FDI flows, and the provisions of the PTA may play once trade liberalization took place. a minor role. For some other countries, the economic envi- When using a dummy variable to account for the exis- ronment is not favorable to FDI, and the investment provi- tence of the PTA, it is difficult to assess the exact role of sions of PTAs may not be enough to change this situation. investment provisions. The trade provisions, as we have seen, Nonetheless, these provisions are solidly associated in are likely to play an important role in FDI decisions, and the empirical analysis with increased FDI flows; the studies agreement itself, independent of its provisions, could increase include control variables such as gross domestic product bilateral investment—for example, by advertising to poten- (GDP) or GDP per capita that account for differences in tial investors opportunities in the partner country. Further- growth and income. more, the dummy variables cannot distinguish between The literature yields mixed results on the effect of BITs. ambitious agreements with deep investment provisions Following Hallward-Driemeier (2003), several studies find and PTAs that only marginally address investment. Adams weak evidence of a positive impact of BITs on investment. et al. (2003) were the first to control for this problem by Gallagher and Birch (2006) point out that BITs signed introducing an index of liberalization that measures the by Latin American countries with the United States do breadth and depth of PTA provisions. The index is then not seem to attract U.S. investment, while the number 320 Sébastien Miroudot of BITs signed by a country with other countries is posi- regional dimension of international production networks tively correlated with the total amount of inward FDI. (see, for example, Inomata et al. forthcoming). In Europe Aisbett (2007) identifies an endogeneity issue, in that BITs and North America, it is through comprehensive economic are first signed with countries with which there is already integration agreements such as NAFTA and the European an important bilateral investment relationship. When the Economic Agreement that global supply chains have selection in BIT participation is controlled for, the correla- emerged. In Asia, production networks preceded the intro- tion between BITs and investment flows is not robust. duction of PTAs, but countries are now engaged in multi- Two studies provide evidence of a positive impact of ple negotiations to ensure that policies can cope with the BITs on FDI flows, in particular, to developing countries. rapid growth of trade and investment. Some authors have Egger and Merlo (2007) propose a dynamic panel estima- pointed out that the distribution of intra-ASEAN invest- tion and estimate that in the short run, BITs produce a ment flows has remained unchanged in the last decade, a 4.8 percent increase in FDI. The long-run impact is esti- situation perhaps related to the late adoption of the mated at 8.9 percent. The authors stress that failure to ASEAN Investment Agreement (AIA) and to its relatively account for the dynamic nature of FDI leads to biased loose disciplines (Jarvis et al. 2009). results. Busse, Königer, and Nunnenkamp (2010) control Because of the importance of the relationship between for the endogeneity of BIT adoption and for unilateral trade and investment (whether as substitutes or comple- investment liberalization and find a positive impact ments), there is a rationale for introducing investment of BITs on FDI flows. Further investigation of the rela- provisions in trade agreements, and it is not surprising to tionship shows that BITs can be a substitute for weak see more and more PTAs including deep investment com- domestic institutions but not for unilateral capital mitments. PTAs generally do not improve on existing BITs account liberalization. with respect to protection of investment, but by adding Recent studies on the impact of BITs on investment the market access dimension and by regrouping trade and have gone further than PTA empirical studies by intro- investment provisions under the same agreement signed ducing a dynamic framework and by better taking into for an indeterminate period, they offer a better package of account the endogeneity issue. They have, however, not disciplines for investors. BITs influence the policy deter- sufficiently distinguished the different types of BITs, in minants of FDI, but PTAs also improve the economic particular, those that include liberalization provisions in determinants and have been found to have a stronger addition to protection disciplines—the case of some BITs impact on investment (UNCTAD 2009a). signed by the United States, Canada, and Japan. An excep- Several lessons for developing countries emerge from the tion is a study by Berger et al. (2010), who look at the recent wave of trade and investment agreements. First, there impact of specific types of provisions. Although, in the are important economic gains in the negotiation of bilateral case of PTAs, the authors find that guarantees of market investment measures, and empirical studies show that these access and state-investor dispute settlement mechanisms gains are higher for North-South PTAs. Developed coun- have a positive impact on FDI flows, these provisions do tries already have among themselves instruments that grant not play a significant role in the case of BITs. Investors national treatment and MFN treatment to investors and respond quite indiscriminately to BITs. their investments (in particular, the OECD Codes of Liber- The fact that most BITs do not cover preestablishment alisation of Capital Movements and Current Invisible Oper- could explain part of the mixed empirical results and the ations, and the companion National Treatment Instru- difference observed between BITs and PTAs. Moreover, ment). With no multilateral agreement on investment, and about 2,500 BITs are in force, and the more common it is with GATS negotiations stalemated, PTAs are so far the only for two countries to share a BIT, the less one can expect the option for increasing market access for investors, for both variable to explain differences in bilateral investment flows. inward and outward investment. Investment protection could be achieved for developing countries through BITs, but disciplines limited to postestablishment are not enough Conclusions to realize the gains from the internationalization of produc- PTAs are more and more becoming “PTIAs�—preferential tion. Only BITs covering preestablishment and comple- trade and investment agreements. Dealing with investment mented by a PTA that liberalizes trade can offer the same when opening markets becomes increasingly relevant in attractiveness to foreign investors. the context of the fragmentation of world production and North-South PTAs with investment provisions are global supply chains. Studies on globalization highlight the especially relevant when thinking about global value Investment 321 chains consisting of a parent company established in the Finally, even if agreements with less ambitious provi- North and subsidiaries created in the South to supply sions have been found to have a positive impact because of inputs or to perform final assembly of the product. In the the signal they send investors, agreements with wide sec- context of vertical specialization, the prohibition of per- toral coverage and broad disciplines are associated with formance requirements, such as local content require- higher flows of FDI. The signaling effect is often related to ments, is particularly important because FDI is motivated a commitment to negotiate further, and the positive impact by global production rather than being aimed at the local will be sustained if the countries effectively build on the market. To the extent that the foreign affiliate does not initial agreement and expand the scope of investment compete with local companies, market access, not just commitments. nondiscrimination, is crucial in this context because the There is little evidence of asymmetric treatment in foreign investor will look for the lowest investment costs North-South PTAs: provisions tend to be the same for all in all potential locations. As demand and global value parties. A difference can be seen in the sectoral coverage chains shift to the South (Kaplinsky and Farooki 2010), of agreements for investment in services. The developed the experience of successful North-South PTAs with economy generally has more subsectors with commit- investment provisions should now be reproduced in ments, but fewer preferential commitments. The opposite South-South PTAs. So far, there are fewer South-South holds for the developing economy: typically, fewer sectors PTAs with deep investment provisions, but their number are scheduled in the PTA than for the developed-country has been increasing in Asia (Miroudot 2009). partner, but the commitments are more often preferen- The second lesson is that because of liberal rules of tial. This difference reflects the fact that GATS schedules origin, not all developing countries have to sign a PTA of commitments for mode 3 are more extensive for devel- with the main trading economies. Regional initiatives oped countries. In terms of additional liberalization, the and the multilateralization of PTAs can guarantee that in effort can be bigger for the developing country. This is the future all developing countries will become part of at in the country’s self-interest, and an asymmetric treat- least one regional agreement offering access to the main ment that would restrict the liberalization commitments markets through the “spaghetti bowl� of investment PTAs. of the developing country would limit its potential gains What is important is to promote multilateral-friendly pro- (Heydon and Woolcock 2009). Through symmetric pro- visions that can ensure the consistency of provisions visions, investment PTAs can achieve a pro-development across agreements. In particular, third-party MFN rules outcome. can help countries benefit from agreements signed by For the least-developed economies, a concern that can other countries and can be efficient in minimizing the be addressed through PTAs is the risk of facing costly costs associated with distortions in the preferential treat- compensation under commercial arbitration and having ment granted. For developing countries that may lack the legitimate policy considerations overridden by direct resources to negotiate multiple PTAs and may be afraid of claims of foreign investors. Although an investor-state dis- bilateral negotiations with larger and more developed pute settlement mechanism is in the end desirable, in that economies, these mechanisms should be emphasized and it offers the highest guarantees to foreign investors, some further strengthened. authors have argued that such a mechanism could be pre- The third lesson is that despite promotion of the liberal- mature for the least-developed economies.14 Moreover, ization of investment, these agreements have all preserved there is no multilateral institution for administering BITs the right of countries to regulate. Countries can be reluc- and resolving disputes about the interpretation and appli- tant to grant full national treatment and market access to cation of investment treaties (Adlung 2009). By referring foreign investors, but in practice, agreements allow for to WTO definitions and legal instruments (in particular, many exceptions and reservations that can lead to progres- GATS), and by resorting to state-state dispute mechanisms sive liberalization, as well as the reservation of activities for in the tradition of trade agreements, PTAs can offer a bet- public purposes. The economic gains measured in empiri- ter option for developing countries that are not ready for cal studies are found in agreements that maintain noncon- the commercial arbitration of BITs. In the future, however, forming measures in key sectors and that have carved out the convergence of dispute mechanisms in international certain sectors. For economic gains to be maximized, trade and economic law will be necessary (Choi 2007), restrictive investment measures should be minimized, but and improvement of investor-state dispute settlement this does not generally entail forgoing policies that may be mechanisms could better serve the interests of developing desirable for development. economies, as well. Annex 322 Table 14A.1. Overview of Recent PTAs Covering Investment Investment protection Nondiscrimination Year Temporary Fair and Free Expropriation Investor-state in Establishment or National MFN entry for key equitable transfer and dispute PTA force Type market access treatment treatment personnel treatment of funds compensation settlement Australia–Chile FTA 2009 Hybrid Positive list Negative list Negative list Yes Yes Yes Yes Yes CAFTA–DR FTA 2006 NAFTA-inspired Negative list Negative list Negative list No Yes Yes Yes Yes Canada–Peru FTA 2009 NAFTA-inspired Negative list Negative list Negative list Yes Yes Yes Yes Yes Chile–Colombia FTA 2009 NAFTA-inspired Negative list Negative list Negative list Yes Yes Yes Yes Yes China–Singapore FTA 2009 GATS-inspired Positive list Positive list No Yes Pending the conclusion of the ASEAN–China Investment (services) Agreement EU–CARIFORUM EPA 2008 GATS-inspired Positive list Positive list Negative list Yes No No No No EFTA–Korea, Rep. FTA 2006 GATS-inspired Positive list Positive list Negative list Yes Yes Yes Yes Yes (services), (services), negative list negative list (goods) (goods) Japan–Brunei Darussalam EPA 2008 Hybrid Positive list Positive list Negative list No Yes Yes Yes Yes (commercial (commercial presence), presence), negative list negative list (investment) (investment) Japan–Chile SEP 2007 NAFTA-inspired Negative list Negative list Negative list Yes Yes Yes Yes Yes (positive list (positive list for financial for financial services) services) Japan–Indonesia EPA 2008 Hybrid Positive list Positive list Negative list Yes Yes Yes Yes Yes (commercial (commercial presence), presence), negative list negative list (investment) (investment) Japan–Malaysia EPA 2006 Hybrid Positive list Positive list Negative list Yes Yes Yes Yes Yes (commercial (commercial presence), presence), negative list negative list (investment) (investment) Japan–Philippines EPA 2008 Hybrid Positive list Positive list Negative list Yes Yes Yes Yes To be (commercial (commercial negotiated presence), presence), negative list negative list (investment) (investment) Japan–Thailand EPA 2007 GATS-inspired Positive list Positive list Through a Yes Yes Yes Yes Yes request Korea, Rep.–Singapore FTA 2006 NAFTA-inspired Negative list Negative list No Yes Yes Yes Yes Yes (positive list (positive list for financial for financial services) services) New Zealand–China FTA 2008 GATS-inspired Positive list Positive list Positive list Yes Yes Yes Yes Yes (services), (services), (services), negative list negative list negative (goods) (goods) list (goods) Panama–Singapore FTA 2006 NAFTA-inspired Negative list Negative list Negative list Yes Yes Yes Yes Yes U.S.–Morocco FTA 2006 NAFTA-inspired Negative list Negative list Negative list No Yes Yes Yes Yes U.S.–Oman FTA 2009 NAFTA-inspired Negative list Negative list Negative list No Yes Yes Yes Yes U.S.–Peru PTA 2009 NAFTA-inspired Negative list Negative list Negative list No Yes Yes Yes Yes Source: Author’s compilation. Note: ASEAN, Association of Southeast Asian Nations; CAFTA–DR, Dominican Republic–Central America Free Trade Agreement; EFTA, European Free Trade Association; EPA, economic partnership agreement; EU, European Union; CARIFORUM, Caribbean Forum of African, Caribbean, and Pacific (ACP) States; FTA, free trade agreement; GATS, General Agreement on Trade in Services; MFN, most favored nation; NAFTA, North American Free Trade Agreement; PTA, preferential trade agreement; SEP, strategic economic partnership. The table includes only PTAs that entered into force between 2006 and 2009 and that contain substantive investment provisions covering both goods and services. In the case of the EFTA–Republic of Korea FTA, there is a separate agreement on investment in which Norway is not a party. In the China–Singapore FTA, provisions on investment are those of the ASEAN–China Investment Agreement when it enters into force. The term “positive list� means that the provision is included in the agreement for a positive list of sectors; “negative list� means that the provision applies to all sectors, except those specified in a list in which reservations are made by the parties. 323 Table 14A.2. Selected Empirical Studies on the Impact of Bilateral Investment Measures 324 Empirical study Data coverage Empirical model Results Investment provisions in preferential trade agreements Dee (2006) Bilateral stocks of outward Model of “complex� FDI with third-country effects and When FDI and trade are sufficiently driven by fundamentals foreign direct investment an index of investment provisions for 38 preferential (as in Asia), the resulting network patterns of investment do (FDI) for up to 32 parent trade agreements (PTAs). not need to be boosted by investment provisions of PTAs. countries and 109 hosts When this is not the case, the investment provisions of PTAs (1988–97). signed with source countries have mixed effects. (A positive effect is found for PTAs signed by large economies.) Lesher and Bilateral trade and investment Gravity regressions for trade and FDI flows with an index Investment provisions are positively associated with trade and, Miroudot (2007) flows of 51 reporter for investment provisions in 24 North-South PTAs. to an even greater extent, investment flows. Results are countries with 180 partner insignificant for bilateral investment treaties countries (1990–2004). Hufbauer and Schott Bilateral FDI inward stocks for Gravity regressions with dummy variables for nine Most PTAs have increased bilateral investment stocks between (2008) about 170 countries prominent PTAs. their members. The Canada–U.S. FTA has decreased bilateral (1976–2005). FDI stocks (interpreted as the unwinding of former tariff- jumping FDI) Miroudot (2009) Bilateral trade and investment Gravity regressions derived from the knowledge-capital Preferential treatment for investment is associated with higher stocks of 18 Asian model for trade and FDI flows with an index for inward and outward investment stocks. The impact is lower economies with 190 partner investment provisions in 22 Asian PTAs. (The index than in previous studies but economically significant (about countries (1990–2006). includes data on sectoral coverage.) 14 percent more FDI for the most extensive investment provisions over the least extensive ones) Investment provisions in bilateral investment treaties (BITs) Gallagher and Total FDI inflows and FDI from Regression on total FDI inflows with a variable indicating BITs with the United States do not independently attract U.S. Birch (2006) the United States in 24 the number of BITs with other countries and gravity FDI, but a higher number of BITs with other countries is Latin American countries regression for FDI inflows from the United States with associated with more FDI inflows (1980–2002). a dummy variable for BITs. Egger and Merlo (2007) Bilateral stocks of outward FDI Dynamic panel estimation of an FDI equation derived The short-run impact of BITs on FDI is estimated at 4.8 percent for 24 home and 28 host from the knowledge-capital model; dummy variable and the long-run impact at 8.92 percent. Accounting for countries (1980–2001). for BIT ratification and signature. the dynamic nature of FDI leads to less biased results Aisbett (2007) Bilateral investment outflows Model of host government’s decisions whether to The correlation between BITs and investment flows is not of 24 OECD countries to participate in a BIT and foreign investor’s decision robust when selection into BIT participation is controlled for, 28 developing countries whether to invest in the host, controlling for and there is no evidence for the claim that BITs signal a safe (1980–1999). endogeneity of BIT adoption. investment climate Busse, Königer, and Three-year averages of Gravity regressions on the share of bilateral FDI in total BITs do promote FDI flows to developing countries and may Nunnenkamp (2010) bilateral FDI flows from FDI outflows, with a dummy variable for BITs ratified even substitute for weak domestic institutions, although not 28 source countries into 83 and control variables for unilateral FDI liberalization; for unilateral capital account liberalization developing host countries instrumentation of BIT and generalized method of (1978–2004) moments (GMM) estimation to control for the endogeneity of BIT adoption. Investment provisions in BITs and PTAs Berger et al. (2010) Three-year averages of FDI Gravity regressions on the share of bilateral FDI in total FDI reacts positively to PTAs only if they offer liberal admission flows from 28 source FDI outflows, with dummy variables for specific types rules. Dispute settlement mechanisms appear to play a minor countries into 83 developing of investment provisions; GMM estimation to control role. The reactions to BITs are less discriminate, with foreign host countries (1978–2004) for the endogeneity of trade and investment treaties. investors responding favorably to the mere existence of BITs Source: Author’s compilation. Note: OECD, Organisation for Economic Co-operation and Development. Investment 325 Notes Adlung, Rudolf, and Peter Morrison. 2010. “Less Than the GATS: ‘Nega- tive Preferences’ in Regional Services Agreements.� Journal of Interna- This chapter draws on Houde, Kolse-Patil, and Miroudot (2007); Lesher tional Economic Law 13: 1103–43. and Miroudot (2007); and Miroudot (2009). The author thanks Rudolf Aisbett, Emma Kate. 2007. “Bilateral Investment Treaties and Foreign Adlung and Jean-Christophe Maur for their helpful comments and sug- Direct Investment: Correlation versus Causation.� CUDARE Working gestions. The views expressed are those of the author and do not represent Paper 1032, University of California Berkeley, Berkeley CA. those of the Organisation for Economic Co-operation and Development Antràs, Pol, and Elhanan Helpman. 2004. “Global Sourcing.� Journal of (OECD) or its member countries. Political Economy 112 (3): 552–80. 1. In this chapter, we define PTAs with investment provisions as Baldwin, Richard. 2006. “Globalisation: The Great Unbundling(s).� agreements that include provisions on the liberalization or protection Contribution to the project “Globalisation Challenges for Europe and of investment. In most PTAs, promotion of investment is among the Finland,� Economic Council of Finland, Helsinki. http://www.vnk.fi/ objectives mentioned in the preamble or in a specific article on invest- hankkeet/talousneuvosto/tyo-kokoukset/globalisaatioselvitys-9-2006/ ment, but only agreements with substantive provisions for encouraging artikkelit/Baldwin_06-09-20.pdf. bilateral investment (through liberalization or protection) are taken Baldwin, Richard, Simon Evenett, and Patrick Low. 2009. “Beyond Tariffs: into account. Multilateralizing Non-Tariff RTA Commitments.� In Multilateralizing 2. Whether GATS MFN obligations apply to WTO members with Regionalism: Challenges for the Global Trading System, ed. Richard no MFN exemption for their bilateral investment treaties under the Baldwin and Patrick Low, 79–145. Cambridge, U.K.: Cambridge GATS remains an open question. See Adlung and Molinuevo (2008) for University Press. a discussion. Barth, James R., Juan A. Marchetti, Daniel E. Nolle, and Wanvimol 3. For a recent stocktaking of investment provisions in PTAs, see Sawangngoenyuang. 2006. “Foreign Banking: Do Countries’ WTO Kotschwar (2009). Commitments Match Actual Practices?� WTO Staff Working Paper 4. On the negative impact of local content requirements, see Moran, ERSD-2006-11, World Trade Organization, Geneva. Graham, and Blomström (2005, ch. 11). Berger, Axel, Matthias Busse, Peter Nunnenkamp, and Martin Roy. 2010. 5. On the spread of specific NAFTA provisions to countries outside “Do Trade and Investment Agreements Lead to More FDI? Account- the Americas, see Baldwin, Evenett, and Low (2009). ing for Key Provisions inside the Black Box.� WTO Staff Working 6. As noted, BITs signed by the United States and some BITs signed by Paper ERSD-2010-13, World Trade Organization, Geneva. Canada and Japan cover the preestablishment phase and therefore deal Beviglia Zampetti, Americo, and Pierre Sauvé. 2006. “Rules of Origin for with market access. Services: Economic and Legal Considerations.� In The Origin of 7. The possibility of de facto discrimination should be taken into Goods: Rules of Origin in Regional Trade Agreements, ed. Olivier Cadot, account. The fact that the same measure applies de jure to both domestic Antoni Estavadeoral, Akiko Suwa Eisenmann, and Thierry Verdier, and foreign investors does not mean that the impact of the measure is the 114–46. Oxford, U.K.: Oxford University Press. same for the two. The treatment of the foreign company can still be Bhagwati, Jagdish, Pravin Krishna, and Arvind Panagariya, eds. 1999. regarded as “less favorable.� Trading Blocs: Alternative Approaches to Analyzing Preferential Trade 8. In NAFTA, for example, these restrictions are “grandfathered� but Agreements. Cambridge, MA: Massachusetts Institute of Technology. are not explicitly listed. Blonigen, Bruce A. 2005. “A Review of the Empirical Literature on FDI 9. Studies suggest that this is often the case. See, for example, Barth Determinants.� NBER Working Paper 11299, National Bureau of et al. (2006), on the banking industry. Economic Research, Cambridge, MA. 10. Some exemptions can, however, have an important impact on Brainard, S. Lael. 1997. “An Empirical Assessment of the Proximity- FDI in sectors such as financial services, audiovisual services, and profes- Concentration Trade-off between Multinational Sales and Trade.� sional services. American Economic Review 87: 520–44. 11. Fink and Jansen (2008) and Baldwin, Evenett, and Low (2009) Busse, Matthias, Jens Königer, and Peter Nunnenkamp. 2010. “FDI study rules of origin for services providers, but their analysis regarding Promotion through Bilateral Investment Treaties: More Than a Bit?� juridical persons is relevant for all investors. The rules of origins are the Review of World Economics 146 (1): 147–77. same for goods and services in the case of established companies. Choi, Won-Mog. 2007. “The Present and Future of the Investor-State 12. This example and others can be found in Adlung and Morrison Dispute Settlement Paradigm.� Journal of International Economic Law (2010). 10 (3): 725–47. 13. On the determinants of FDI, see Blonigen (2005). On FDI Dee, Philippa. 2006. “Multinational Corporations and Pacific Regional- spillovers and the benefits of FDI for development, see Moran, Graham, ism.� Pacific Economic Papers 358, Australia–Japan Research Center, and Blomström (2005). Australian National University, Canberra. 14. See UNCTAD (2008). The increased complexity of investment Delimatsis, Panagiotis, and Martin Molinuevo. 2008. “Article XVI: Market agreements and the larger number of arbitration cases necessitate capacity- Access.� Max Planck Commentaries on World Trade Law, WTO—Trade building initiatives for developing economies. in Services 6: 367–426. Dodge, William S. 2006. “Investor-State Dispute Settlement between Developed Countries: Reflections on the Australia–United States Free References Trade Agreement.� Vanderbilt Journal of International Law 39 (1): 1–38. Adams, Richard, Philippa Dee, Jyothi Gali, and Greg McGuire. 2003. “The Egger, Peter, and Valeria Merlo. 2007. “The Impact of Bilateral Investment Trade and Investment Effects of Preferential Trading Arrangements: Treaties on FDI Dynamics.� World Economy 30: 1536–49. Old and New Evidence.� Australian Productivity Commission Staff Egger, Peter, and Michael Pfaffermayr. 2004. “The Impact of Bilateral Working Paper, Australian Productivity Commission, Canberra. Investment Treaties on Foreign Direct Investment.� Journal of Com- Adlung, Rudolf. 2009. “Trade in Healthcare and Health Insurance Serv- parative Economics 32: 788–804. ices: The GATS as Supporting Actor?� WTO Staff Working Paper Emch, Adrian. 2006. “Services Regionalism in the WTO: China’s Trade ERSD-2009-15, World Trade Organization, Geneva. Agreements with Hong Kong and Macao in the Light of Article V(6) Adlung, Rudolf, and Martin Molinuevo. 2008. “Bilateralism in Services GATS.� Legal Issues of Economic Integration 33: 351–78. Trade: Is There Fire Behind the (BIT-) Smoke?� Journal of Interna- Estevadeordal, Antoni, and Kati Suominen. 2005. “Rules of Origin in the tional Economic Law 11: 365–409. World Trading System and Proposals for Multilateral Harmonization.� 326 Sébastien Miroudot In Quantitative Methods for Assessing the Effects of Non-Tariff Measures World Economy, ed. Sven W. Arndt and Henryk Kierzkowski, 17–34. and Trade Facilitation, ed. Philippa S. Dee and Michael J. Ferrantino, Oxford, U.K.: Oxford University Press. 337–410. Singapore: Asia-Pacific Economic Cooperation (APEC) Kaplinsky, Raphael, and Masuma Farooki. 2010. “What Are the Implica- Secretariat and World Scientific Publishing. tions for Global Value Chains When the Market Shifts from the North Ferrantino, Michael J., and H. Keith Hall. 2001. “The Direct Effects of Trade to the South?� Policy Research Working Paper 5205, World Bank, Liberalization of Foreign Direct Investment: A Partial Equilibrium Washington, DC. Analysis.� USITC Office of Economics Working Paper 2001-10-B, U.S. Kotschwar, Barbara. 2009. “Mapping Investment Provisions in Regional International Trade Commission, Washington, DC. Trade Agreements: Towards an International Investment Regime?� In Fink, Carsten, and Marion Jansen. 2008. “Services Provisions in Regional Regional Rules in the Global Trading System, ed. Antoni Estevadeordal, Trade Agreements: Stumbling Blocks or Building Blocks for Multilateral Kati Suominen, and Robert Teh, 365–417. Cambridge, U.K.: Liberalization?� In Multilateralizing Regionalism: Challenges for the Cambridge University Press. Global Trading System, ed. Richard Baldwin and Patrick Low, 221–61. Lesher, Molly, and Sébastien Miroudot. 2007. “Analysis of the Economic Cambridge, U.K.: Cambridge University Press. Impact of Investment Provisions in Regional Trade Agreements.� Fink, Carsten, and Aaditya Mattoo. 2002. “Regional Agreements and Aussenwirtschaft 62 (2): 193–232. Trade in Services.� Policy Research Working Paper 2852, World Bank, Marchetti, Juan A., and Martin Roy. 2008. “Services Liberalization in Washington, DC. the WTO and in PTAs.� In Opening Markets for Trade in Services: Fink, Carsten, and Martin Molinuevo. 2008. “East Asian Preferential Trade Countries and Sectors in Bilateral and WTO Negotiations, ed. Juan Agreements in Services: Liberalization Content and WTO Rules.� A. Marchetti and Martin Roy, 61–112. Cambridge, U.K.: Cambridge World Trade Review 7: 641–73. University Press. Fink, Carsten, and Deunden Nikomborirak. 2007. “Rules of Origin in Mattoo, Aaditya. 2005. “Services in a Development Round: Three Goals Services: A Case Study of Five ASEAN Countries.� Policy Research and Three Proposals.� Policy Research Working Paper 3718, World Working Paper 4130, World Bank, Washington, DC. Bank, Washington, DC. Gallagher, Kevin P., and Melissa B. L. Birch. 2006. “Do Investment Agree- Miroudot, Sébastien. 2009. “Impact of Investment Provisions in Asian ments Attract Investment? Evidence from Latin America.� Journal of RTAs.� In Expansion of Trade and FDI in Asia: Strategic and Policy World Investment and Trade 7 (6): 961–74. Challenges, ed. Julien Chaisse and Philippe Gugler, 186–211. London: Hallward-Driemeier, Mary. 2003. “Do Bilateral Investment Treaties Routledge. Attract FDI? Only a Bit . . . and They Could Bite.� Policy Research Moran, Theodore H., Edward M. Graham, and Magnus Blomström, Working Paper 3121, World Bank, Washington, DC. eds. 2005. Does Foreign Direct Investment Promote Development? Helpman, Elhanan. 2006. “Trade, FDI, and the Organization of Firms.� Washington, DC: Peterson Institute for International Economics. Journal of Economic Literature 44: 589–630. Morgan, David. 2008. “Dispute Settlement under PTAs: Political or Heydon, Kenneth, and Stephen Woolcock. 2009. The Rise of Bilateralism: Legal?� In Challenges to Multilateral Trade: The Impact of Bilateral, Comparing American, European, and Asian Approaches to Preferential Preferential and Regional Agreements, ed. Ross P. Buckley, Vai Io Lo, Trade Agreements. Geneva and New York: United Nations University and Laurence Boulle, 241–63. The Hague: Kluwer Law International. Press. OECD (Organisation for Economic Co-operation and Development). 2005. Hoekman, Bernard. 1996. “Assessing the General Agreement on Trade in International Investment Law: A Changing Landscape. Paris: OECD. Services.� In The Uruguay Round and Developing Countries, ed. Will ———. 2008. OECD Codes of Liberalisation: User’s Guide 2008. Paris: Martin and L. Alan Winters, 88–124. Cambridge, U.K.: Cambridge OECD. University Press. Sauvant, Karl P., and Lisa Sachs. 2009. The Effect of Treaties on Foreign Houde, Marie-France, Akshay Kolse-Patil, and Sébastien Miroudot. 2007. Direct Investment: Bilateral Investment Treaties, Double Taxation “The Interaction between Investment and Services Chapters in Selected Treaties, and Investment Flows. Oxford, U.K.: Oxford University Press. Regional Trade Agreements.� OECD Trade Policy Working Paper 55, Sauvé, Pierre. 2007. “Investment Regulation through Trade Agreements: Organisation for Economic Co-operation and Development, Paris. Lessons from Asia.� In Towards Coherent Policy Framework: Under- Hufbauer, Gary, and Jeffrey Schott. 2008. “Fitting Asia-Pacific Agreements standing Trade and Investment Linkages, Economic and Social Com- into the WTO System.� In Multilateralizing Regionalism: Challenges for mission for Asia and the Pacific (ESCAP), 27–64. Studies in Trade and the Global Trading System, ed. Richard Baldwin and Patrick Low, Investment 62. Geneva and New York: United Nations. 554–635. Cambridge, U.K.: Cambridge University Press. Stephenson, Sherry M. 2002. “Regional versus Multilateral Liberalization Inomata, Satoshi, B. Meng, N. Yamano, and S. Zhu. Forthcoming. “Globaliza- of Services.� World Trade Review 1: 187–209. tion and Regional Integration: A Study of International Production UNCTAD (United Nations Conference on Trade and Development). Networks.� Institute of Development Economics, Japan External 2006. Latest Developments in Investor-State Dispute Settlements. IIA Trade Organization (IDE-JETRO), Chiba, Japan. Monitor 4. New York and Geneva: United Nations. Jarvis, Darryl, Chen Shaofeng, Tan Teck Boon, and Lee Kuan Yew. 2009. ———. 2008. International Investment Rule-Making: Stocktaking, Chal- “Investment Liberalization in the Association of Southeast Asian lenges and the Way Forward. UNCTAD Series on International Nations: Progress, Regress or Stumbling Bloc?� In Expansion of Trade Investment Policies for Development. New York and Geneva: United and FDI in Asia: Strategic and Policy Challenges, ed. Julien Chaisse and Nations. Philippe Gugler, 138–85. London: Routledge. ———. 2009a. The Role of International Investment Agreements in Attract- Jeon, Bang Nam, and Susan F. Stone. 2000. “Foreign Direct Investment ing FDI to Developing Countries. UNCTAD Series on International and Trade in the Asian-Pacific Region: Complementarity, Distance Investment Policies for Development. New York and Geneva: United and Regional Economic Integration.� Journal of Economic Integration Nations. 15: 460–85. ———. 2009b. World Investment Report 2009: Transnational Corpora- Jones, Ronald W., and Henryk Kierzkowski. 2001. “A Framework for tions, Agricultural Production and Development. New York and Fragmentation.� In Fragmentation: New Production Patterns in the Geneva: United Nations. 15 TRADE FACILITATION Jean-Christophe Maur Regional cooperation on customs and the facilitation of each type of cooperation? What types of trade facilitation trade goes hand in hand with preferential trade liberaliza- reform should be undertaken regionally? tion. Preferential regimes require specific customs arrange- ments and some degree of cooperation between partners’ Regional Trade Facilitation Initiatives: border agencies if trade creation between the partners to Background the agreement is to be achieved. Cooperation between neighboring countries to facilitate international trade goes Regional trade facilitation initiatives are numerous and back to antiquity. Regional trade facilitation efforts are varied. This variation is expressed in the scope of the agree- thus not at all a new idea, but they have acquired in recent ments negotiated and in the institutions created to manage years a renewed dimension that invites further examina- regional cooperation. Historically, cooperation on trade tion. Two concurrent dynamics are in play: the spread of facilitation has taken place not so much in PTAs as through preferential trade agreements (PTAs), and the rise of an various kinds of bilateral cooperation arrangements, such international consensus on the need for modern border as transit corridor agreements. management tools. Recently negotiated PTAs tend to incorporate, in addi- Scope tion to liberalization of trade in goods, numerous provi- sions on reciprocal regulatory reform and cooperation. There is no common definition of trade facilitation in Trade facilitation is one of these aspects, as shown in agree- PTAs. An overview of selected agreements shows that trade ments on customs procedures, simplification, harmoniza- facilitation is generally defined by the scope of the meas- tion, and cooperation. This represents a significant change ures covered in the agreement rather than by a specific from earlier practice, when trade facilitation provisions definition (box 15.1). This is not to say that there are no were nearly absent from PTAs. elements of commonality between the various definitions. Since the late 1990s, new international efforts have been The common elements have to do with procedures related made to promote faster and easier movement of goods to the importation and exportation of goods (e.g., cus- across borders, in order to reap the full benefits of the lib- toms, standards, and technical barriers to trade) and with eralization of policies affecting international trade. Notable their enforcement. in this respect are the initiatives undertaken in the World Trade facilitation provisions in PTAs vary mainly by Trade Organization (WTO), starting in 1998 with a sym- the amount of detail and the scale of their aspirations. posium on trade facilitation, and the work of the Asia- Certain agreements, such as APEC, are broad in scope Pacific Economic Cooperation (APEC) in this area.1 and include, for instance, services trade. Detail and ambi- The study of regional trade facilitation efforts is instruc- tion seem also to have increased over time, as is shown by tive and offers useful lessons from local successes regarding recent agreements signed by the United States and the the approach to reform and the implementation of European Union (EU). reforms. This chapter presents what we know about these As a first step toward understanding the scope of trade issues. It asks the following questions: What are the factors facilitation in trade agreements, we can examine the trade behind the increase of facilitation provisions in regional facilitation negotiations in the WTO Doha Round. The focus PTAs? What are the specific advantages and drawbacks of in the WTO is mainly on the revision of three articles of the 327 328 Jean-Christophe Maur Box 15.1. Definition and Scope of Trade Facilitation in Selected PTAs The excerpts that follow are from the texts of the respective preferential trade agreements (PTAs). Asia-Pacific Economic Cooperation (APEC), 1999 Trade facilitation is defined as “the use of technologies and techniques which will help members to build up expertise, reduce costs and lead to better movement of goods and services.� APEC, 2002 “Trade facilitation generally refers to the simplification, harmonization, use of new technologies, and other measures to address procedural and administrative impediments to trade.� Common Market for Eastern and Southern Africa (COMESA), 1994; Southern African Development Community (SADC), 1996; East African Community (EAC), 2004 The three agreements use the same language: “‘Trade facilitation’ means the co-ordination and rationalisation of trade procedures and documents relating to the movement of goods from their place of origin to their destination�; “‘trade procedures’ means activities related to the collection, presentation, processing and dissemination of data and information concerning all activities constituting international trade.� European Union (EU)–Chile, 2003 No definition is given, but the objective of the agreement is “the facilitation of trade in goods through, inter alia, the agreed provisions regarding customs and related matters, standards, technical regulations and conformity assessment procedures, sanitary and phytosanitary measures and trade in wines and spirit drinks and aromatised drinks.� Caribbean Forum of African, Caribbean, and Pacific (ACP) States (CARIFORUM)–EU Economic Partnership Agreement (EPA), 2008 “Customs and trade facilitation� (not defined) covers customs legislation, procedures and administrative cooperation, including the establishment of a joint committee. Detailed provisions address, among other things, joint initiatives, additional facilitation measures for traders with high level of compliance; nondiscrimination and reasonableness principles; application of the EU’s single administrative document; risk assessment and simplified measures; binding rulings; use of information technology; transparent and nondiscriminatory licensing of customs brokers; no mandatory use of preshipment inspection; facilitation of transit movement; reduction and simplification of documentation; rapid release of goods; right of appeal; standards of integrity; customs valuation; and relationships with business communities. Dominican Republic–Central America Free Trade Agreement (CAFTA–DR), 2004 “Customs administration and trade facilitation� is not defined but covers the publication of customs laws, regulations, and general procedures; simplified procedures and rapid release of goods; automation; risk management; cooperation on exchange of information and advance information of regulation changes; express shipment; review and appeal procedures and customs penalties; advance ruling, and capacity building. Trade facilitation measures are only mentioned as they relate to future joint work in the field of standards, cooperation on regulatory issues such as convergence or equivalence of standards, international harmonization, and reliance on supplier’s declaration of conformity.a Canada–Costa Rica, 2002 “With the objectives of facilitating trade under this Agreement and cooperating in pursuing trade facilitation initiatives on a multilateral and hemispheric basis, Canada and Costa Rica agree to administer their import and export processes for goods traded under this Agreement on the basis that: (a) procedures be efficient to reduce costs for importers and exporters and simplified where appropriate to achieve such efficiencies; (b) procedures be based on any international trade instruments or standards to which the Parties have agreed; (c) entry procedures be transparent to ensure predictability for importers and exporters; (d) measures to facilitate trade also support mechanisms to protect persons through effective enforcement of and compliance with national requirements; (e) the personnel and procedures involved in those processes reflect high standards of integrity; (f) the development of significant modifications to procedures of a Party include, in advance of implementation, consultations with the representatives of the trading community of that Party; (g) procedures be based on risk assessment principles to focus compliance efforts on transactions that merit attention, thereby promoting effective use of resources and providing incentives for voluntary compliance with the obligations to importers and exporters; and (h) the Parties encourage cooperation, technical assistance and the exchange of information, including information on best practices, for the purpose of promoting the application of and compliance with the trade facilitation measures agreed upon under this Agreement.� Source: Author’s compilation. a. The scope of “customs administration� provisions in the U.S.–Chile (2004) and U.S.–Peru (2009) agreements is nearly identical (under the title of “customs administration�) to that in the CAFTA–DR agreement, except that capacity building is not mentioned. General Agreement on Tariffs and Trade (GATT)—dealing, aspects and services related to international trade are briefly, with freedom of transit, fees and formalities, and pro- either excluded or are dealt with in different negotiations. cedures—and the emphasis is on rules and regulations, A second approach is to refer to other international rather than on procedures (see box 15.2). Infrastructural instruments, such as the Revised Kyoto Convention of the Trade Facilitation 329 circumstances are regulations optimal, given market Box 15.2. Trade Facilitation and the WTO failures affecting the supply of trade facilitation serv- ices and regulations? Because markets are imperfect, Trade facilitation became a topic of discussion for the World governmental intervention is sometimes needed to Trade Organization (WTO) at the Singapore Ministerial Conference in December 1996, when members directed the deliver the optimal social outcome. When market fail- Council for Trade in Goods “to undertake exploratory and ures cannot be remedied at the national level, address- analytical work . . . on the simplification of trade procedures in order to assess the scope for WTO rules in this area� ing them becomes a transnational issue, and collective (Singapore Ministerial Declaration, para. 21). action is needed. Regional solutions should be sought After several years of exploratory work, in July 2004 when the failing markets correspond to a well-defined WTO members formally agreed to launch negotiations on trade facilitation on the basis of modalities contained in set of nations. Annex D of the so-called “July package.� Under this 2. Determine which of these trade facilitation policies are mandate, members are directed to clarify and improve best implemented at the regional level. When do General Agreement on Tariffs and Trade (GATT) Article V (freedom of transit), Article VIII (fees and formalities regional approaches offer better and more cost-effective connected with importation and exportation), and Article X prospects than other means (regional, unilateral, or (publication and administration of trade regulations). The multilateral) for carrying out trade facilitation reform? negotiations are also aimed at enhancing technical assistance and capacity building in this area and improving A subsidiarity test needs to be applied: actions to effective cooperation between customs and other achieve a given policy objective should be taken at the appropriate authorities on trade facilitation and customs compliance issues. lowest level of government capable of effectively To date, members have submitted a great number of addressing the problem at hand (Sauvé and Zampetti proposals under the mandate, and these provide the basis 2000). Ideally, this level of action should correspond to for the ongoing negotiations. The negotiations are to be completed according to the overall Doha Development the level affected by the need for the regional good— Agenda timeline. that is, the political jurisdiction should match the eco- Source: World Trade Organization. nomic domain of benefits. Thus, the most appropriate participants will partake in the provision of regional trade facilitation, and transaction costs will be econo- mized (Arce and Sandler 2002). World Customs Organization (WCO) and the standards of various specialized agencies and international organi- Cooperation for Regional Trade Facilitation zations.2 A third approach is to avoid all a priori defini- tions and to embrace whichever definition is provided in Trade facilitation reform is often a regional issue. By defini- a given agreement. tion, the crossing of borders involves two trading partners, In this chapter, trade facilitation encompasses all and the improvement of transport conditions often measures (often, but not always, enforced at the border) requires some form of regional infrastructure hub. Thus, that can reduce the cost of policing international trade. a wide variety of regional cooperation efforts for trade Governments need to control international trade in vari- facilitation exist, as described next. ous ways in order to manage externalities or meet core Transit corridor management. Many regional organiza- policy objectives such as consumer protection, national tions have been set up to guarantee the smooth and rapid security, and revenue generation. The economic challenge flow of goods from gateway ports to the hinterland. Corri- is to ensure that such policies are enforced in the most dor management agreements typically deal with technical efficient way. As discussed in the next section, this does standards for vehicles, mutual recognition of drivers’ not necessarily imply the removal of trade protection, but licenses, and market access by transport services. The man- it does involve the minimization of costs that are directly agement structures require cooperation among private and related to the operations of international trade supply public sector stakeholders operating in several countries chains. and on issues that range from streamlining of regulatory In the context of regional integration, deciding on trade requirements to improvement of infrastructure. An example facilitation reform policies involves the following two of successful corridor management is the Trans-Kalahari steps: corridor in southwestern Africa (box 15.3). Sanitary and phytosanitary protection. Coordinated 1. Identify the most cost-effective policies for regulating action by countries is required to prevent the spread of dis- imports and exports and ways of implementing these eases borne by agents that easily cross borders. For exam- policies. From an economic perspective, under what ple, Tanzania and its neighbors in the Southern African 330 Jean-Christophe Maur Box 15.3. The Trans-Kalahari Corridor The Trans-Kalahari Corridor (TKC) is a road route between South Africa’s Gauteng province and Namibia’s Walvis Bay, via Botswana. The corridor, which opened in 1998, replaced a longer route through western South Africa. In 1999, through the efforts of the Walvis Bay Corridor group—a public-private partnership that aims to make Namibia’s Walvis Bay Port an international gateway to and from the South Africa Customs Union (SACU)—a rehabilitation project was carried out. The route already had fairly good connectivity and required a relatively small amount of capital investment. The main work program of the TKC Management Group—consisting of representatives from transport operators, infrastructure and transport authorities, port and customs authorities, freight forwarders, and other interested businesses—thus focused on facilitating agreements between the three member countries that would promote the simplification of border-crossing procedures. To this end, the TKC Management Group established partnerships with the customs agencies of Botswana, Namibia, and South Africa. In August 2003 the group embarked on the pilot phase of a program to replace all existing transport documents with a single administrative document (SAD). In November 2003 the ministers of transport of Botswana, Namibia, and South Africa signed a broad memorandum of understanding that formally binds the signatories and the private sector to a program to deal with cross- border transport and trade issues such as border management, customs control, traffic regulation, and road transport policies. Since these agreements were put in place, border processing times in the corridor have been cut from an average of 45 minutes to 10–20 minutes. Cost savings from the reduction in border delays are estimated at approximately US$2.6 million per year. Operators now move approximately 620,000 tons annually along the corridor, representing about 65 percent of expected capacity. By contrast, in 1999, before the improvements, only 15 percent of the route’s capacity was utilized. In February 2008 the Namibian roads authority announced that it would invest up to 310 million Namibian dollars in upgrading road infrastructure along the portion of the corridor that extends from Okahandja and Karibib. The work entails widening the road and was expected to take four years to complete. On February 8, 2008, the governments of Namibia and Botswana signed a memorandum of understanding concerning the creation of a new dry port facility at Walvis Bay that will be operated by Botswana. The acquisition of land in Namibia by Botswana under a 50-year lease is being carried out as part of the regional integration initiative of the Southern African Development Community (SADC). The initiative will increase use of the Trans-Kalahari Corridor, and it would probably not have been economically feasible before the regulatory overhaul of the corridor. Source: Author’s compilation from the following sources: M. Madakufamba, “Towards Seamless SADC Transport Corridors,� The Namibian, March 3, 2008, http://www.namibian.com.na/index.php?id=28&tx_ttnews[tt_news]=48259&no_cache=1; A. Shilongo, “NAM, Bots Sign MoU,� New Era, February 11, 2008; C. Tjatindi, “N$310-M Earmarked for Trans Kalahari Corridor Rehabilitation,� New Era, February 7, 2008; U.S. Agency for Interna- tional Development (USAID), “Trans-Kalahari Corridor Exemplifies Collaboration,� USAID, Washington, DC., 2006; Arnold 2006; World Bank 2005; Adzibgey, Kunaka, and Mitiku 2007. Development Community (SADC) have agreed on a five- for the failure of national organizations to set up such year program of vaccination, surveillance, and control of systems, hampered as they are by the small scale of opera- animal movements across borders to combat highly conta- tions, underdeveloped national financial services, and the gious bovine diseases that persist in Tanzania (Tanzania unwillingness of international insurers to face the politi- 2005). cal and commercial risks of developing markets. In Regional standards and accreditation bodies. Accredita- Uganda the cost of customs bonds is estimated to amount tion bodies ensure that standards certification laboratories to 4 percent of import and export costs. Uganda’s are able to assess conformance with standards. In many Integrated Framework diagnostic study (Uganda 2006) countries, the national standards infrastructure is not well recommends a regional approach toward reducing the developed, and the economy is too small to support such incidence of these costs. institutions. To overcome this problem, Brunei Darussalam One-stop border posts (OSBP). Joint management of (to give one illustration) has concluded an agreement with border posts on each side of the border can be the source of the Singapore Accreditation Council. many benefits. In southern Africa, TradeMark Southern Guarantees. Regional guarantee and insurance mecha- Africa is providing support for the establishment of several nisms enable transporters to reduce costs by avoiding OSBPs along the North-South corridor, a major transit duplication and the need to make cash deposits. The Com- project.3 Countries can align their procedures so as to mon Market for Eastern and Southern Africa (COMESA) ensure streamlined movement of goods across the border. has introduced a “yellow card,� a third-party regional Harmonization of some procedures, data exchange, unified motor vehicle insurance scheme that allows traders to pur- documentation, and mutual recognition of findings allow chase insurance covering transport in the region (Arvis the elimination of cost duplication. The sharing of facilities 2005). COMESA launched in 2010 a regional customs such as scanners and weighbridges is a potential additional bond guarantee system that will be initially operational in source of cost saving. Finally, joint operations may enhance Burundi, Kenya, Rwanda, and Uganda. Regional guarantees the overall efficiency of border agencies (information (to secure the payment of duty and taxes) can compensate sharing, for example, improves intelligence) and may foster Trade Facilitation 331 regulatory changes that will facilitate trade, such as allow- and results in inefficiencies. In other cases, the complexity ing forms of extraterritoriality or the adoption of interna- and lack of transparency of administrative processes indi- tional norms. rectly favor the emergence of operators that can work the system and charge fees for facilitation services. The existence of rents may create opposition to reform. Economic Dimensions of Trade Facilitation For example, in a small project funded by the World Bank, The economic impact of facilitation of trade flows at the an international mail service that offers highly simplified regional level is twofold. First, preferential trade facilita- procedures for small exporters, Exporta Facil, was set up in tion measures have static efficiency effects because of bet- Peru. The exporter using this service does not need to use a ter allocation of factors. Second, they have more systemic customs agent, logistics agent, or freight forwarder or to and dynamic effects that are associated with imperfect consolidate the merchandise, and even the packaging is competition settings. Schiff and Winters (2002) note that provided for. The exporter only needs to go to a post office “in the presence of economies of scale or inter-country and complete the export declaration for the tax agency, externalities, market solutions are generally sub-optimal, using the Internet (Toledano and Ansón 2008). According and failing to cooperate can be very costly. However, to the project team, private customs agents put up some regional cooperation is not the same as regional integra- resistance to the initiative. A compromise was reached that tion, and, indeed, there is generally rather little connection set the maximum value of the packages to be exported at between them.�4 US$2,000, but with the proviso that the maximum value would be revised a year later. (The plan is to raise the max- imum to US$10,000.) Static Effects The other effect of domestic rents is well known from Assuming for now that trade facilitation efforts are con- the economic analysis of discriminatory trade liberaliza- ducted on a preferential basis, what would be the trade and tion: it can lead to trade diversion and to negative welfare welfare effects of liberalization? The answer hinges on effects if the loss of domestic rents to foreign exporters is whether the implicit protection afforded to domestic not compensated by the benefits from lower prices result- industries by higher trade costs prior to liberalization gen- ing from the liberalization. In many situations, however, erates domestic rents. there are no such domestic rents, or they are small, and so The first type of rent to be considered is tax revenue, the liberalization effects of trade facilitation, even on a and here different situations prevail according to the coun- preferential basis, will necessarily be welfare enhancing. try. Sometimes border fees such as consular or transit fees This is an important difference from tariff liberalization, are imposed with a revenue objective in mind—that is, fees where the risks of trade diversion are much higher. Trade are higher than what would be strictly necessary to cover facilitation lowers trade costs and benefits the consumers the costs of services such as maintenance of transit corri- of imported goods because larger and more affordable dors or the operational costs of border agencies. The Arab quantities of goods are available. Republic of Egypt abolished its consular fees in the context Under a preferential facilitation regime (e.g., a mutual of its regional agreements with the European Union (EU), recognition agreement on procedures, or a specific author- COMESA, and the Greater Arab Free Trade Area (GAFTA). ized traders’ regime that gives some form of preferential Although no cost estimates were made of the revenue that treatment to exporters from the PTA partner country), had been generated by these fees, Egypt acknowledged that the sources of supply may shift from the most efficient the costs of services rendered had been only a fraction of world supplier (if there were no discrimination) to the the fee levied.5 most favored one (the PTA partner). Trade facilitation, The second type of rent is that captured by vested pub- however, is often nondiscriminatory, eliminating the risk lic and private interests. Lack of facilitation provides scope of trade diversion. for such interests to levy surcharges on importers and Even though the overall effects of trade facilitation are exporters or to provide poor-quality services, resulting in positive, the reforms can produce losers: the external trade delay, loss of goods, and corruption. In some instances, partners that may be excluded from the preferential rents are directly created by the public sector: corrupt offi- regime; the economic actors that had benefited from high cials use the complexity and uncertainty of rules to extract trade costs; domestic producers that were protected from “speed money� from traders. Exclusive licensing of some international competition; and private sector interests that operations at the border, as illustrated by compulsory use were able to extract rents from the complexity of the trade of customs brokers, stifles competitive provision of services environment. 332 Jean-Christophe Maur Dynamic Effects • Sharing of facilities such as border posts and, possibly, gateway facilities such as airports and ports. There are two main dimensions to the benefits of transna- tional cooperation, corresponding to the specific market or An important caveat is that some aspects of these institutional failures faced by countries: (a) the realization cooperation mechanisms are not necessarily regional or of economies of scale through, for example, elimination of bilateral and may be achieved multilaterally. For instance, duplication and increased competition, and (b) the avoid- harmonization is not necessarily optimal at the regional ance of negative externalities and the creation of positive level; more broadly shared international standards may externalities among neighbors. be more efficient. Still, a large part of these extra costs may be better addressed at the regional level, by a limited Economies of Scale number of countries because of political-economy con- siderations and the complexity of arrangements such as Given the many interventions and parties in the interna- mutual recognition. tional transport of goods, and perhaps the need to cross Economies of scale can also be realized on administra- several borders, there is plenty of scope for trade operators tive procedures and on the private services that deliver to encounter cost duplication. Because an important por- trade facilitation. Because the procedures and services that tion of these costs is fixed, eliminating duplication will facilitate trade can involve large fixed, and possibly sunk, enable efficiency gains for firms and will allow smaller- costs, full economies of scale in administrative procedures scale operators to access export markets—an important and services for international trade transactions may not consideration for developing countries.6 be realized at the country level, especially in small and poor Duplication arises because similar requirements must countries. It is unclear to what extent there is a scale barrier be met repeatedly, but also because national rules differ, to efficient border administration that employs modern which increases search costs and associated uncertainties practices such as single windows and risk management. and creates further opportunities for rent seeking and cor- The extent to which economies of scale can be realized ruption.7 COMESA has reduced duplication through a depends on the presence and size of fixed costs in the man- regional license for carriers, obviating the need to pay for agement of trade transactions. If these fixed costs are low, it multiple licenses (Schiff and Winters 2003). Inspection of is unlikely that regional economies of scale will be large goods, if carried out in different places on each side of a enough to justify complex regional cooperation mecha- border, also delays trade, and different national regulatory nisms. Customs reform projects have been shown to be requirements force traders to meet two standards instead sustainable at the country level (Moïsé 2005), but when of one. To illustrate, in Tanzania, registration requirements fixed costs are important, it may be more cost-effective to for agrochemical pesticides are burdensome and subject to share them regionally rather than incur them alone. It is high fees, yet Tanzania’s market for such pesticides is small, then important to identify where the fixed costs lie. and meanwhile, equivalent and more efficient products For Finland, Norway, and Sweden, a motivation for have already been registered and tested in neighboring entering into cross-border cooperation agreements (start- Kenya (Tanzania 2005).8 ing in 1960) was division of labor; that is, sharing of the Methods of reducing duplication costs may involve: cost of managing the 1,630-kilometer-long border between • Forming a common market within a customs union. Norway and Sweden and the 739-kilometer-long border Intraunion borders are removed, and external borders between Norway and Finland.10 Another indication that are managed on behalf of the union. small administrations may not be able to afford all the • Harmonization through the use of identical templates material and infrastructure necessary is seen in the current and information fields for documentation. (Simplifi- WTO negotiations; several members have called on “small, cation of documentation, often pursued along with vulnerable economies� to undertake a regional approach harmonization, is another measure for reducing dupli- toward implementing certain expected WTO commit- cation, although it is not intrinsically regional.) ments that will require capacity building and have made a • Mutual assistance among authorities, through meetings specific submission regarding regional trade facilitation of experts, exchange of information and data, and assis- enquiry points.11 tance with extraterritorial investigations.9 Standards and backbone services. Beyond customs opera- • Mutual recognition of rulings (e.g., on customs control tions, there are perhaps more acute problems in the imple- measures in transit operations) and of certification and mentation of standards. Many developing countries are too testing. small to offer the full range of conformity assessment for Trade Facilitation 333 standards, partly because of problems with access to accred- more, on the liberalization of regional transport services itation. Small countries should thus benefit from regional (through, e.g., relaxation of cabotage restrictions or expan- integration (World Bank 2005). sion of air traffic rights) than on the availability of infra- Setting up regional certification and accreditation bod- structure. Indeed, transport hubs tend to be geographically ies, or opening regional markets for such bodies, could be mobile, suggesting the secondary importance of infra- a way to provide cheaper and better testing, building on structure as a determinant of their location. scale economies and comparative advantage. Regional approaches can make sense for countries facing serious Regional Cross-Border Externalities shortages in technical skills, which can be an issue for mod- ern trade facilitation techniques.12 For instance, Cambodia, Regional agreements can serve as a policy coordination the Lao People’s Democratic Republic, and Vietnam are all mechanism to help prevent individual countries from deficient in standards infrastructure, and lack of certification adopting national strategies that fall short of optimal and accreditation is a common problem (Haddad 2008). global outcomes. Lack of coordination is in some instances Once again, solutions to national constraints are not linked to externality problems—problems that affect the necessarily regional; the opening of certification and welfare of individual countries but cannot be handled by accreditation markets to international companies may be these countries alone. Although the distinction between enough to remedy the absence of national providers of positive and negative externalities is not really important such services. In some instances, however, market openness from a strict analytical point of view, it may be useful from may not be enough, and access to regional providers a policy viewpoint because each may raise different imple- (implying open access to testing and certification in neigh- mentation challenges. Positive (or negative) externalities boring countries and perhaps some form of formal arise when actions by one or more countries create benefits arrangement) might be needed. (impose costs) that are not taken into account in the deci- Another sector in which size can be an important con- sion to perform that action. In other words, the private cost straint is the backbone services that are crucial to trade to the country that originates the action does not equal its transactions: finance and insurance, transport and logis- social cost. tics, handling, measurement services, and communication Negative externalities. Countries that carry on transit services. The supply of these services for trade transactions trade can be tempted to use trade-restricting policies such can require a scale of production that extends beyond as setting revenue-maximizing fees for transit, imposing national borders.13 Insurance and financial services (letters compulsory transit routes and checkpoints, limiting access of credit, guarantees, insurance, and the like) constitute key to foreign transporters, or requiring securitized convoys. inputs for the capacity to trade internationally. National Fees and requirements may exceed the cost of the services operators in developing countries may not be able to pro- provided (use of roads, provision of security, and so on) or vide these services, or only at noncompetitive prices. go beyond the measures strictly necessary for secure tran- According to the European Bank for Reconstruction sit.17 In the worst cases, the motives behind these policies and Development (EBRD 2003), national banking systems are protectionist; often, countries simply fail to consider do not pool enough capital to underwrite trade transac- the negative externalities imposed on neighbors.18 The risk tions.14 Small firms often lack access to the financial guar- of negative spillovers is particularly important when alter- antees for payments that would allow them to export.15 native transit routes are few, as is frequently the case in Fixed costs and geographic factors confer natural Africa.19 Domestic transport infrastructure constraints monopoly characteristics on some modes of transport, often have regional implications, justifying, from an eco- particularly rail and maritime at the national level, that can nomic perspective, regional ports or airport hubs. Land- be mitigated or even eliminated in a broader regional set- locked countries depend on the quality of their neighbors’ ting. For instance, regional transport hubs help realize infrastructure. economies of scale and create extra competition.16 For To illustrate, Uganda’s most important transport and freight transport, the emergence of multimodal hubs, gen- trade facilitation issues are outside the country’s direct con- erally located near important existing air, sea, or rail infra- trol (Uganda 2006). Tanzania and Kenya, its coastal neigh- structure, generates important economies of scale, through bors, offer poor trade facilitation—the Kenyan port of higher utilization of infrastructure, as well as efficiency Mombasa, which handles 95 percent of Uganda’s external gains (compared with point-to-point routes) through trade traffic, is congested; transit bond regimes are finan- competition between modes of transport (Müller-Jentsch cially burdensome; and rail transit does not offer a compet- 2003). Transport hubs depend as much, and probably itive alternative to poor road transport and expensive 334 Jean-Christophe Maur pipeline transport.20 As Schiff and Winters (2002) note, Trans-European Networks on Transport. Similar aspira- this type of externality is often asymmetric: landlocked tions are seen in trade action plans in developing countries, countries stand to gain a great deal from better transit, but such as Uganda’s recent Diagnostic Trade Integration the benefits for the coastal partner (i.e., improved access to Study (DTIS), which emphasizes the use of electronic data the internal market) are much smaller. In practice, land- interchange (EDI) at the regional level; the development of locked countries have not gained much from participation a regional cargo-tracking system; and the interconnection in regional trade agreements, probably because important of the East African Community’s customs electronic sys- trade obstacles have persisted (Yang and Gupta 2005). tems (Uganda 2006). Transport hubs, mentioned above, Standards and phytosanitary (SPS) measures provide a also create positive externalities, such as access to multi- further example of possible market failure. As was dis- modal transport platforms. cussed earlier, weak or absent enforcement of SPS in one Positive externalities—beyond the realization of country can mean that negative consequences spill over to economies of scale—also arise from the provision of inter- neighbors. national finance and insurance. International provision of Positive externalities. Like the elimination of negative such services offers the possibility of mutualizing risks externalities, the creation of positive externalities such as across a region and contributes to positive network effects, network effects may justify regional intervention. Trans- such as linking banks that do not usually do business with port, electronic, and other information networks play an each other and diffusing skills through the network. The increasingly important role in trade facilitation reform. We principle of mutualization is applicable to other trade- distinguish here between two main forms of positive exter- related financial instruments that are specifically relevant to nalities: the establishment of shared facilities, and the cre- trade facilitation and transit, such as guarantees for pay- ation of networks. ment of taxes and insurance. COMESA, for example, plans There are benefits for neighboring countries in joining to set up a regional transit bond scheme. Regional guaran- existing networks rather than developing their own sys- tees (to secure the payment of duty and taxes) can address tems or multiplying bilateral channels of communication the failure of national organizations to set up such systems. and information exchange. Southern African countries, for In Uganda, the cost of customs bonds is estimated to add up example, are planning to exchange electronic information to 4 percent to import and export costs, and one recom- on transit cargo on the basis of a system called Asycuda++. mendation of the Integrated Framework diagnostic study is (Mozambique, however, has a proprietary system that is to use a regional approach to reduce the incidence of these incompatible with its neighbors’ software; see Mozambique costs (Uganda 2006). Arvis (2005) argues that the lack of a 2004.) At the heart of network externalities lie the notions regional customs guarantee explains why transit initiatives of adoption of common standards for operation and in developing countries to replicate the success of the TIR interoperability. By joining networks, countries may gain transit system have failed so far (see also Arvis 2010 for a access to several markets while having to pay the cost of detailed discussion of transit systems and the TIR).21 plugging into this network only once. This is a considera- tion in favor of adopting international standards, which Systemic Effects of Regional Trade Facilitation will be accepted by all other countries belonging to the same network of standards. Trade facilitation is not an end in itself; rather, it is a means The adoption of standards for border procedures may of fostering regional integration, and it can play a critical lead to regional practices that differ from internationally complementary role in regional integration strategies. accepted norms. A good practice therefore is to base Trade facilitation as complementary to regional integra- regional standards on international ones. Europe is push- tion. PTAs create a potential loophole for tax evasion if ing for the use of the single administrative document producers and products outside the preferential agreement (SAD), but the SAD itself is based on an internationally are able to take advantage of the exemption regime. It is accepted standard, the UN layout key. There is a solidly therefore in the nature of PTAs to include provisions established culture of international customs and product related to customs implementation and cooperation. First, standards that provides strong incentives for adopting regional PTAs require the establishment of a specific these in regional trade agreements. customs regime for the processing of goods benefiting The European Union has developed several regional from the preferential treatment. This requires, among networks facilitating trade: the New Computerized Transit other things, the provision of additional information and System (NCTS) for electronic transmissions, the Galileo documentation by traders and additional verification work satellite system for global navigation (GALILEO), and the by customs. The most prominent implementation feature Trade Facilitation 335 of regional trade agreements is seen in the criteria set out information technology yields benefits for all traders. In in rules of origin.22 Second, enforcement of customs control theory, it is always possible to manage access to such facili- at border is likely to become stricter as the trade tax base is ties, by, for example, charging different access fees, but in reduced by the liberalization and because of the need to practice, this not done—it is probably pointless and may combat the incentives for tax evasion. run afoul of WTO disciplines. Paradoxically, PTAs may complicate rather than facilitate Some modern techniques of trade facilitation rely on trade, in that they create additional administrative require- special and differential treatment for specific categories of ments for exporters who wish to benefit from the preferen- traders. This is particularly the case in the areas of risk tial regime. The costs of compliance are nontrivial and management and authorized trader approaches to border could amount to a significant share of trade. For instance, management. Security is a growing concern in interna- the compliance costs of rules of origin were estimated at tional trade, and PTAs may be used to set up specific around 2 percent of trade value for the North American regimes not accessible to all traders. Recent PTAs negoti- Free Trade Agreement (NAFTA) and 6.8 percent for the EU ated by the United States, for instance, contain obligations (Cadot et al. 2006). for more favorable treatment of express shipments, espe- It could be argued that in the more complex environ- cially with respect to expedited clearance times. These ment of regional trade, PTAs create incentives for more measures bear a distinct risk of creating some form of dis- efficient enforcement of customs procedures. To maxi- crimination that excludes countries or traders that do not mize market access, the authorities of the exporting have the capacity or possess the correct criteria to access country have an incentive to help the importing country these simplified regimes. In particular, it may be difficult manage its border controls efficiently. This can include for companies in developing countries to receive preferred mutual assistance in investigations and exchange of trader status. information. Although efforts to further simplify and speed border The existence of complementarities between regional crossing should be encouraged, when special regimes are integration and trade facilitation is even more evident required, there is a need to be careful about the implica- when a broader view of trade facilitation is taken. The nar- tions of creating incentives for rent seeking and for exclu- row focus on border management enforcement was the sion of certain operators. In the specific case of developing main approach adopted in the older generation of PTAs countries, firms may have more problems in complying (Moïsé 2005). Modern PTAs go beyond the promotion of with the requirements of authorized traders because their intraregional trade; they are intended to further trade lib- shipments are less regular or because they cannot put eralization in general. together the documentation necessary for accreditation. Interplay between regional and multilateral trade facilita- Similarly, risk management profiles may single out ship- tion. Two concerns about PTAs are whether they are wel- ments from particular origins or in certain industries in a fare enhancing and whether they represent a natural path way that may be detrimental for countries exporting these toward broader liberalization. In both cases, the gold stan- goods. Parties to a PTA could also discriminate against dard is multilateral liberalization. What is the role of third parties by granting preferential access to their most regional trade facilitation provisions in the broader trade efficient screening facilities to PTA members. In fact, how- liberalization context? To what extent do regional trade ever, there is little evidence in existing PTAs of preferential facilitation provisions discriminate against third parties? treatment in trade facilitation. Do they contribute to the “spaghetti bowl� problem? On balance, it appears that trade facilitation provisions As noted earlier, trade facilitation provisions in PTAs are in PTAs are likely to impart additional momentum to likely to be nondiscriminatory, and any reform undertaken reform and to benefit not only the PTA members but other in the context of a specific agreement would be expected to traders, as well. In some dimensions of trade facilitation, benefit the rest of the world. The nondiscriminatory aspect nevertheless, discrimination remains a distinct possibility. of trade facilitation is apparent in the language of most For example, certain features of trade facilitation regimes agreements, which generally outline universal reforms. In could discriminate by type of operator or source of trade any case, many dimensions of trade facilitation are, in the and thus undermine broader regional or multilateral liber- terminology of public goods, nonexcludable: for instance, alization. There seems, however, to be no evidence of this once simplification or transparency is achieved, it does not happening on a large scale now. Regional trade facilitation make sense (and may not be feasible, in most cases) to provisions therefore do not seem to pose a risk of under- exclude other countries from benefiting from it. Similarly, mining multilateral efforts, and PTAs could, indeed, com- the improvement of border agencies’ equipment through plement the multilateral process. 336 Jean-Christophe Maur Balancing the Benefits and the Costs of Implementation Regional cooperation yields mutual (“social�) benefits that exceed the individual (“private�) benefits that countries The costs associated with trade facilitation reforms and proj- would derive by acting alone. Even in the rare cases in ects are not confined to the costs of regional infrastructure which a given country would not benefit or would lose (which are not intrinsically different from those in a nonre- from the cooperative action, it should be possible to com- gional setting) but also include the cost of regional coordi- pensate that country with part of the extra benefits gener- nation. The discussion focuses on the latter. ated by solving the externality problem. There are, how- Regional coordination is not cost free. Anecdotal evi- ever, situations in which this may not be possible—for dence suggests that the transaction costs of coordination example, where there are important transaction costs, as rise sharply with the number of countries participating in a mentioned above, or where information is imperfect and PTA. A review of transit corridors (Arnold 2006), for countries may not fully realize the size of the gains from instance, finds that corridor management in large regional trade facilitation or may not be able to agree on the exact agreements has been difficult to achieve and has sometimes scope of reform to be carried out jointly. led to more cumbersome rules, compared with (probably Going some way toward solving the imperfect informa- more nimble) cooperation agreements and bilateral trade tion problem are simple measures that may sow the seeds agreements. For trade facilitation, there may be many of future cooperation through generation and exchange of stakeholders involved, as well as numerous agencies on information. Such measures include raising awareness each side of the border that may need to be coordinated. among groups of stakeholders to strengthen coalitions of The heterogeneity of national interests is often an addi- reformers; creating formal or informal regional bodies tional complicating factor. Border controls pursue multiple (e.g., expert committees for government officials; regional objectives—collecting revenues, guaranteeing the security trade associations) where stakeholders can meet and of the home territory, protecting consumers, and, of exchange ideas; and supporting research and analysis of course, facilitating trade. The nature of these objectives and issues to inform policy makers. the priority accorded them varies by country. The nature of cooperation will be dictated by the type of Some types of coordinated trade reform might make it regional trade facilitation policy needed. Depending on the more difficult to accommodate such particularities.23 For type of policy, each participating country’s level of contri- instance, adoption of a strong harmonization rule in rela- bution will differ. The first and most common way of pro- tion to customs information systems might bind all partic- ducing a regional good is by the summation of countries’ ipating countries to use identical technology, leaving less efforts: each country needs to contribute a similar amount space for adaptation to individual circumstances. to the joint effort. A regional agreement to harmonize bor- Finally, and linked to all these points, some forms of der procedures and adopt common documentation is an coordination can limit regulatory innovation at the coun- example. The more countries that engage in harmonizing, try level. Given the changing business and economic con- the higher will be the benefits for the entire region because text within which international trade transactions take of the resulting access to a broader overall market. A varia- place, it is by no means clear that what looks like an opti- tion of this situation is when countries’ contribution levels mal regulation today will necessarily be seen as such in five are different. Facilities for which economies of scale are years’ time.24 Governments can change national regula- important, such as regional guarantees or testing laborato- tions relatively easily in the light of new circumstances, but ries, also need contributions from all countries in the the same is not true of regional or multilateral rules: con- region, but these may be uneven, related to the capacity of sensus among all parties is usually required. International each country to contribute. coordination of reforms can therefore run the risk of intro- A second type of regional public good is produced by ducing too much rigidity and regulatory “stickiness.� This tackling the failures arising from the “weakest link.� The would reduce the overall amount of regulatory experimen- inability of one country to adopt trade facilitation meas- tation and could retard the rate of learning and discovery. ures may have negative spillovers on neighbors, which then Flexible forms of regional cooperation may help lower the have an interest in assisting their weaker neighbor to costs of future renegotiations and amendments. upgrade its capacity. An example is the enforcement of rules of origin and other customs regulations in a regional PTA. Traders will rapidly “port-shop� to determine which Delivering Regional Public Goods customs authorities are the more lenient and whether this The existence of externalities points to the need to coordi- leniency justifies the added cost of rerouting goods nate actions and ensure the delivery of public goods. through a specific border. Even relatively small differences Trade Facilitation 337 in the level of implementation can create sizable losses for Transit corridor agreements are ad hoc regional agree- the region in the form of forgone tax revenue, lowered ments that have the single objective of managing interna- standards, and increased threats to health or security. Even tional trade corridors. They take very different forms, and, worse, when border tax revenues are an important source as shown in Arnold (2006), no single model prevails. The of government revenue, there may be an incentive to prac- corridor arrangements are very much works in progress tice beggar-my-neighbor tactics by attracting imports from and have been influenced by the problems they were meant partner countries in a PTA. to address. In particular, the role of governments in corri- Finally, some public goods important for regional facil- dor arrangements varies greatly, from that of sole parties to itation do not need to be supplied by each country; gate- being only one actor. The active involvement of the private ways such as ports, pipelines, and airport hubs can be used sector is an interesting feature of such arrangements. by all countries in the region but are not needed in each A distinction can be made between shallow and deep country. This does not have to be major infrastructure— PTAs. The former contain only very limited provisions specific equipment may also be shared among a group of related to customs enforcement of preferences, whereas the countries. In such cases the challenge is for countries to latter tend to increasingly include extensive trade facilita- agree not to duplicate efforts and instead to concentrate on tion provisions. The big differences between them and ad a best-shot effort. hoc cooperation agreements are that in deep PTAs other In summary, the promotion of regional trade facilita- sectors are liberalized in parallel and the institutions tion requires a careful analysis of the types of regional behind the PTA tend to be more complex—providing, for efforts that are needed. The PTA provisions will need to example, for dispute settlement—and to be closer to polit- reflect the various forms of cooperation that are required ical decision centers. to supply the regional public goods. In some instances, Customs unions enable a more advanced form of trade coordination of policies will have to be the prime objective; facilitation cooperation. The fundamental tenet of a cus- in other circumstances, implementation will have to be toms union is the uniformity of the external tariff. This has carried out in only a subset of countries; and in still others, implications for customs management: the quality of bor- countries will have to make tangible contributions involv- der enforcement has to be identical across the customs ing cross-border transfers of expertise, and funds will have union to ensure that the weaknesses of one party do not to be designated. Finally, for some regional efforts, all of undermine the tax collections of others (through the diver- the above actions will be needed. sion of imports to the border where enforcement is the least stringent) or hinder the fulfillment of other border control objectives, since not all countries in the union may Institutional Arrangements for Facilitation be concerned about the same issues. of Regional Trade Finally, there are single common markets, such as the The question of regional public goods raises implicitly the EU. Although the experience of the EU is far removed from question of what institutional format would be best suited some of the preoccupations of developing countries, it for delivering them. In fact, there is a varied array of shows that even in the context of a common market, where regional institutional arrangements designed to tackle internal borders are dismantled, the management of exter- trade facilitation problems. nal borders remains challenging, and the integration of bor- Bilateral cooperation agreements in the fields of cus- der procedures is accompanied by significant reforms. toms, standards, and transit are common. These agree- We next examine in more detail two forms of regional ments aim primarily at improving specific areas of arrangements that are most relevant for regional trade enforcement and generally have narrow objectives. Some facilitation reform: transit corridors and PTAs. recent agreements push new forms of cooperation. For instance, the United States and China have an agreement Transit Corridors permitting U.S. officers stationed in China to perform inspections related to standards. This concept is similar to Access to global trade is an important development chal- that of the container security initiative, under which U.S. lenge, both for individual countries and for neighbors customs officers are stationed in participating ports. In the wishing to access their markets. This challenge is especially case of the EU, customs cooperation agreements seem to important for landlocked countries and land-based cross- be confined to those between large trading partners and ing of borders. Experience shows that trade facilitation developed or emerging partners, and their aim is essen- projects and reforms are complex to design and carry out tially to develop advanced forms of cooperation. and that they are not always successful because they require 338 Jean-Christophe Maur the implementation of a wide range of consistent measures include the signing of international agreements and the in several sectors and countries. establishment of the regulatory framework necessary for According to Arnold (2006), the source for much of the the provision of transport and logistics services across bor- information in this section, ders (see box 15.4). Provision of infrastructure concerns the physical facilities that make up the corridor—mainly, So far, one of the most effective approaches is to focus on transport and border infrastructure. Operationalization trade corridors. Indeed, one salient feature of overland includes the maintenance of the physical infrastructure, trade, since the most ancient times, is its organization the monitoring of corridor operations, and stakeholder along linear corridors. The trade corridor is the natural management. entity to identify problems, bring together policy-makers and stakeholders, and implement concrete facilitation These components can be delivered by various institu- measures or investment. Corridors also bring together tional structures, separately or together. In southern Africa, national reforms and regional policies. (Arnold 2006, vii) some corridor management structures are public-private partnerships and others are entirely government driven. Although transit corridors are generally relatively well Adzibgey, Kunaka, and Mitiku (2007) point out the variety defined geographically, there is no standard for their oper- of legal instruments employed in Africa: treaties (Northern ations in practice or for institutional and regulatory Corridor), multilateral agreements (Central Corridor), arrangements. Different modes of transport may be memoranda of understanding (Trans-Kalahari), constitu- involved, and the number of routes can vary. Institutional tion (Dar es Salaam), and company registration (Maputo; arrangements for corridors range from very informal Trans-Kalahari). The diversity of international corridor arrangements, to official government management, to gov- arrangements is interesting in that it seems to mirror the ernment bilateral agreements and sophisticated institu- various degrees of formality in the levels of international tions and decision-making processes. cooperation. It would seem that a memorandum of under- The scope of corridor management covers trade and standing does not carry the same legal weight as a constitu- transit agreements, infrastructure and facilities, transport tion. Formality can, however, come at a cost, as Adzibgey, and logistics services, standards, regulations and proce- Kunaka, and Mitiku (2007) note: the constitution of the dures, security, and overall performance of the corridor Dar es Salaam corridor had not yet been signed by Zambia (table 15.1). Participants in transit corridors thus represent at the time of their writing. In some instances, there is no many different sectors, and the interface between the pub- management structure at all.25 lic and private sectors is particularly important. Recent studies of transit corridors have insisted on the The management of transit corridors is a relatively com- central and important role of the private sector, whose plex affair that requires three key ingredients: legal arrange- dynamism has been seen as crucial for pushing important ments, the provision of physical infrastructure, and the changes (Arnold 2006; Adzibgey, Kunaka, and Mitiku 2007; operationalization of the transit itself. Legal arrangements Arvis, Raballand, and Marteau 2010). Table 15.1. Functions of Selected Corridors Corridor Functions ASEAN Trade and transit agreements; standards and procedures Asian Highway Trade and transit agreements; infrastructure and facilities; standards and procedures Canada–Mexico (NAFTA) Infrastructure and facilities ECOWAS Trade and transit agreements; standards and procedures GMS Corridors Trade and transit agreements; standards and procedures Maputo Infrastructure and facilities; transport and logistics services; overall performance Mercosur Trade and transit agreements; standards and procedures Northern Corridor Infrastructure and facilities; transport and logistics services; overall performance Pan-American Infrastructure and facilities SAARC Corridors Infrastructure and facilities TEN Trade and transit agreements; infrastructure and facilities; standards and procedures TRACECA Trade and transit agreements; infrastructure and facilities; standards and procedures Trans-Kalahari Standards and procedures; overall performance Source: Arnold 2006. Note: ASEAN, Association of Southeast Asian Nations; ECOWAS, Economic Community of West African States; GMS, Greater Mekong Subregions; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement; SAARC, South Asian Association for Regional Cooperation; TEN, Trans-European Networks; TRACECA, Transport Corridor Europe-Caucasus-Asia. Trade Facilitation 339 or relocating border functions (when, for instance, cus- Box 15.4. Mercosur’s Transit and Cross-Border toms officials are located in the ports of trading partners). Transport Agreement Simplification and harmonization of border and clearance procedures. Many actions can be carried out by governments An agreement among the countries of the Southern Cone Common Market (Mercosur, Mercado Común del Sur) in unilaterally—including simplification of documentation Latin America has introduced a uniform legal regime for requirements and adoption of modern border management international transport by authorized transport companies. techniques—but coordination of border operations can go It provides for the development of an international transport document, sets up a customs regime modeled a long way toward improving transit. Cooperative activities after Europe’s TIR Convention (although it is not as may include the adoption of common standards for docu- complete because it lacks a guarantee mechanism), and binds the participants to implement an international third- mentation; transit agreements; cooperation agreements party liability scheme. The agreement provides for between border agencies on mutual assistance, exchange of standards regarding carrier responsibility for lost, damaged, data, and joint expert groups; and joint border posts and and delayed goods. Among the general provisions is the opening of transport markets. According to the agreement, synchronization of border operations. Interoperability through adoption of processes that enable Each contracting Party undertakes to give, on the basis of reciprocity, national treatment to the transporters to operate across borders. The principles are transport companies authorized by other Parties to similar to those of border cooperation. Governments must carry out international transport under the terms of the Agreement. Such transport companies from agree on common sets of standards for transport that avoid other countries can also be given exemption from transshipment of cargo from one operator to another. In domestic taxes duties and rates on a reciprocal basis. some instances, international and regional standards The agreement contains provisions that apply to both already exist. Certification that standards are being met bilateral and transit road transport and that provide for is also necessary; here, the greatest challenges concern reciprocity in the allocation of passenger and goods traffic between the parties. It also distinguishes between goods trucks and inland and coastal shipping, the standards for carried on own account and those carried by third parties. which are often determined by local bodies without any Finally, it allows for temporary use of the vehicles of third countries by an authorized transport company. mutual recognition of neighboring countries’ ability to certify such standards. Source: Arnold 2006. Promotion of market access, private sector participation, and competition. Transport services and ancillary services such as logistics, customs broking, freight forwarding, trade finance, and insurance are not always able to compete in foreign mar- A basic function of transit corridor management is to kets. Air and maritime transport are particularly subject to provide a large amount of coordination, and an essential restrictions on the operations of foreign operators in domes- dimension of that coordination is with neighboring coun- tic transport. Some countries mandate the use of national tries. Another dimension is international advocacy, customs brokers and freight forwarders. The rationale is through regional groupings and associations of stakehold- often the poor quality of documentation submission by ers. Finally, regional cooperation is required for joint traders, but the requirement ends up creating inefficiencies— efforts—the provision of common facilities, but also trans- and rents for local operators. Although in these situations, fers of resources across countries to ensure better opera- multilateral reform (services liberalization) or unilateral tion of transit. reform (elimination of monopolies and regulatory rents) In most corridors, the regional coordination function is would seem more appropriate, regional competition may be likely to be essential for the delivery of legal, physical, and a step toward more cost-effective services if trade partners operational inputs into the management of the transit cor- possess services industries that can provide effective competi- ridor. Where it matters most is in the legal and regulatory tion. Allowing the private sector to operate transport infra- aspects, as the effectiveness of transit corridors depends on structure, or even some border management functions, may trade and transit agreements between the countries partic- be a way of introducing modern, efficient management. ipating in the corridor. The scope of such agreements Extension of cost-effective and efficient guarantees against should be large. The list here is inspired by Arnold (2006): liabilities. Air and maritime transport tend to have access Interconnection; removal of transit bottlenecks. Many to regional or worldwide guarantees, but road transport actions, such as the elimination of unnecessary administra- often has to be insured by local suppliers in each country of tive requirements for transit, can be taken by governments transit. Regional insurance systems are one solution. alone, but in some instances, joint efforts may be Another type of liability is that arising from nonpayment required—for example, in establishing joint border posts of customs duties and taxes. 340 Jean-Christophe Maur Improvement of corridor capacity. Gateways, roads, and involved in border control to work toward a common border crossings require important infrastructure for tran- objective of facilitating trade. Coverage in the same trade sit. A hurdle to investment is that border crossings for tran- agreement of the policy areas implemented by these agen- sit may be located in remote regions that receive low prior- cies (such as SPS measures and standards) could provide a ity in government funding, as is the case with the useful venue for agreeing to common aims linking these Pan-American Highway and border crossings to land- policies, and the specific political-economy incentive locked countries in Africa (Arnold 2006). In Asia, the behind each set of policies may be superseded by the over- Greater Sub-Mekong Region has been established partly to arching objective of the PTA itself. provide for joint regional infrastructure in transport and Conceivably, all the facets of trade facilitation reform telecommunications (Brooks 2008). could be included in specifically designed cooperation agreements on, for instance, customs cooperation or tran- sit. PTAs, however, generally offer scope for covering a wide Institutional Arrangements in PTAs spectrum of policies across which various concessions, PTAs represent one of the fastest developing forms of including noneconomic ones, could be traded off (Devlin regional cooperation and incorporate more and more pro- and Estevadeordal 2004). Although this broadening does visions, in particular in regulatory areas (Horn, Mavroidis, increase the complexity of negotiations, PTAs can guaran- and Sapir 2010). PTAs have so far not led to much trade tee better commitments. In theory, any attempt to deny a facilitation reform (Moïsé 2005), but they are able to per- trade facilitation concession by imposing other trade barri- form several functions that may be necessary for regional ers (e.g., tariffs) should be more difficult because use of trade facilitation. We mentioned earlier the importance of these measures is regulated by the agreement. By the same coordination and transfer of resources—both functions token, enforcement of trade facilitation measures will be that PTAs could assume. In addition, the newer PTAs offer guaranteed by the possibility that partners will withdraw prospects for reform beyond the WTO agenda, they incor- other concessions. porate possible institutional mechanisms for assisting with PTAs are not infrequently complemented by resource implementation, and they gather under one umbrella dis- sharing and redistribution mechanisms among partner ciplines that can complement and reinforce each other. countries, which may include financial and technical assis- Deep integration agreements offer the possibility of tance. Trade facilitation reforms can be demanding, in both pursuing comprehensive trade facilitation, involving expertise and material. Regional efforts offer, in addition to reforms in several sectors of the economy that can be other support, the possibility of benchmarking (as, for incorporated into the new generations of PTAs. These example, in the regional program for Trade and Transport agreements afford the opportunity to deal with sectors that Facilitation in Southeastern Europe; see De Wulf and Sokol are not well covered multilaterally, while providing effi- 2005) and the sharing of good practice. The EU devotes cient enforcement mechanisms. The wider remit of PTAs considerable money and effort to assisting neighbor coun- compared with multilateral approaches is reflected in the tries with which it has association agreements (OECD definitions reviewed in box 15.1, which are much broader 2006). Similarly, APEC has developed a program of techni- for regional agreements than at the multilateral level. A cal assistance under which members have drawn up both good illustration is the adoption of flexible and harmo- collective and individual country assistance plans, covering nized policies on visas and the opening of services— 16 subject areas.28 Most of these plans are carried out indi- dimensions that are usually out of reach of a multilateral vidually by the countries. Participation is voluntary but is agreement but that can be part of regional discussions. conducted in coordination with other APEC members, and (The Economic Community of West African States closely related objectives are pursued in this way. Progress [ECOWAS], for instance, has suppressed visas between is monitored at the APEC level. Several funds financed by member countries.) The political economy of PTAs makes individual country members, such as Japan’s Trade and dealing with migration issues easier because countries can Investment Liberalization Fund (TILF), provide limited exchange commitments on movements of natural persons— support to members. something not possible in the General Agreement on Trade Regional capacity building also takes place in South- in Services (GATS).26 South agreements such as COMESA. The South African By tackling many dimensions of trade facilitation, PTAs Revenue Service (SARS) provides direct technical assistance may be able to exploit natural complementarities between to regional trade partners. In particular, when the provision the different elements of trade facilitation reform.27 A par- of better trade facilitation at the regional level is impaired ticular challenge for reform is to get all the agencies by the some members’ lack of capacity, with implications Trade Facilitation 341 for the regional system, a regional group can assist the deliv- constitutes such measures. For instance, Bin (2008), look- ery of joint assistance by acting as a coordination mecha- ing at Southeast Asia and using a broad definition that nism and sharing the costs among members. Even in the includes standards and technical regulations, as well as cus- absence of redistribution arrangements, PTAs potentially toms cooperation (now a staple of most regional trade create beneficial access to external financial resources by agreements), finds 34 agreements, out of a total of 134, that increasing the credibility of the regional group in offering contain trade facilitation measures. Wille and Redden loan collateral and enhancing its ability to do so (Devlin (2007) review some of the same agreements, employing a and Estevadeordal 2004). definition closer to what is being negotiated under the PTAs act as trust-building mechanisms, promoting WTO, with more of a focus on customs operations. interactions between officials and exchange of information Examples of agreements containing advanced provi- (Schiff and Winters 1998). Trust is a vital aspect of trade sions about border clearance and facilitation of trade are facilitation cooperation, as it helps mitigate risk through now relatively plentiful and can be found not only in PTAs increased confidence in shared information and systems. in which the EU and the United States take part but also in It thus contributes to reduction of physical constraints on agreements involving Japan and Southeast Asian countries the transport of goods in the form of, for example, inspec- and in several South-South agreements—the South Asian tions, or requirements to adhere to compulsory routes. Free Trade Area (SAFTA), COMESA, and the Malaysia– PTAs have a good track record in enabling trust build- Pakistan ECA. ing across partner countries’ administrations, but attempts to involve businesses through such means as public-private EU Agreements partnerships have been much less successful. For instance, the European Union has had mixed success with its efforts It is only recently that the EU has included more ambitious to build ambitious public-private partnerships in the con- trade facilitation provisions in its agreements. Before that, text of its European transport network policy.29 trade facilitation–related measures were found only in the Customs cooperation committees are often established in framework of customs cooperation (Fasan 2004), which PTAs to discuss enforcement issues and to help defuse dis- really had to do with enforcing trade rules. Since the con- putes (World Bank 2005, 89). More informal expert groups clusion of the association agreements with Mediterranean have also been established in a regional integration context, countries, provisions on trade facilitation have been such as the EU Florence process on infrastructure, which has expanded, but only in the agreements with Chile (2002) been influential in promoting reform (Rufin 2004). and Mexico (2000) were comprehensive and more ambi- Regional trade integration implies the building of tious objectives set. After a short moratorium on the nego- regional institutions that can promote certain policies on tiation of preferential agreements, a new generation of behalf of its members. PTAs offer a cost-saving institutional agreements is under way that includes the economic part- architecture (Devlin and Estevadeordal 2004; Sandler 2006) nership agreements with African, Caribbean, and Pacific through which the demand for regional public goods can be (ACP) countries. The EPAs completed so far include large more easily aggregated. Both the redistributive mechanisms trade facilitation chapters. discussed above and the cooperation mechanisms estab- The agreement with Chile is the first of its type to men- lished through PTAs contribute to the delivery of public tion specifically modern and ambitious trade facilitation goods by limiting free riding and helping to improve capac- measures. Publication on the Internet, advance rulings, the ity. It is also often thought that regional institutions are bet- single window, right of appeal, use of risk management ter placed to advance international harmonization agendas techniques, simplification of formalities, including the use (World Bank 2005). Finally, the pooling of scarce resources of the European single administrative document (SAD), can make regional institutions more efficient. automation, and mechanisms of cooperation are men- tioned in the agreement. The EPA agreements that have been signed also have a Trade Facilitation Provisions in PTAs broad scope, although they contain variations that may Trade facilitation provisions are relatively new elements in have resulted from the negotiation process. Stevens et al. PTAs, but they are expanding rapidly. Several existing (2008) note that levels of ambition on some provisions agreements involve major trading nations and contain vary somewhat across EPA agreements that have been sophisticated commitments. negotiated or are under negotiation. For example, some How expansive trade facilitation measures in PTAs are is EPAs, such as those with the Caribbean Forum of African, difficult to gauge, as it depends on the definition of what Caribbean, and Pacific (ACP) States (CARIFORUM) and 342 Jean-Christophe Maur Pacific ACP countries, contain provisions for the use of the rather aspirational nature of the agreements rather than single administrative document; others do not. The latter laying down binding disciplines. This is the result of the include the agreements with the Economic and Monetary nature of trade facilitation measures, which tend to be Community of Central Africa (CEMAC, Communauté context-specific and procedural rather than easily trans- Économique et Monétaire de l’Afrique Centrale), with latable into precise and uniform legal language. It may SADC, and with Côte d’Ivoire and Ghana.30 This difference also reflect the newness of negotiations on trade facilita- occurs even though the backbone structure of the agree- tion among countries. Finally, one can question the rela- ments remains the same. The reason remains unclear: does tive absence of language relating to technical assistance, at it reflect a unilateral evolution in the EU position, particu- least in the EPA agreements, given the need for technical, lar aspects that needed more emphasis because of specific legal, and management upgrading to support trade facili- characteristics of the partner country (including its level of tation reform. development), or a truly negotiated outcome? Several general remarks can be made about the EU U.S. Agreements agreements. First, the PTAs reflect the general ambitions of the EU in the WTO trade facilitation negotiations. The EU As with the EU agreements, recent agreements signed by the is among those WTO members that broadly support the United States have greatly extended the trade facilitation current agenda of negotiations, with comparatively few nar- provisions, in line with U.S. multilateral negotiating strategy. row offensive interests. Its preferential trade facilitation The increased emphasis on customs procedures started agenda is therefore very similar to that prevailing at the with the PTA with Australia (2003). Since then, U.S. PTAs multilateral level. The agreements also refer explicitly to have followed a relatively uniform template, with both international standards and organizations, suggesting that developing and developed countries. The customs admin- the EU is seeking complementarity between the different istration chapter of early agreements (e.g., with Chile and processes. Perhaps this complementarity could be enhanced Morocco) has been renamed “customs administration and by identifying exactly which areas need to be pursued trade facilitation� in recent agreements, but the content has jointly and as a matter of priority. (This is something that remained substantively the same. This is indicative of an the committee established by the agreement might do.) agenda that is driven largely by the United States and less Second, agreements are used to help promote a by the trading partner. The most recent agreement, with European-sanctioned vision of trade facilitation. This is Peru, displays very few variations from the agreement with evident in the stated objective of developing common posi- CAFTA plus the Dominican Republic (CAFTA–DR). tions in international organizations, the diffusion of the Overall, the provisions of U.S. agreements seem slightly SAD standard (which is based on an international stan- more binding than comparable provisions in European dard), and strong institutional mechanisms. agreements. For instance, regarding the expedited release Third, advanced agreements provide for the establishment of goods, the agreements contain disciplines about release of formal institutional mechanisms for cooperation, with at point of arrival and separation of release and border special committees on customs and trade facilitation. Such clearance. The provisions are quite specific, indicating a committees have a broad remit: exchange of information, desirable release time of no more than 48 hours, or 6 hours notification, monitoring, joint development of standards for express shipment. Similarly, dispositions on the use of and policy positions, facilitation of the provision of techni- electronic submission and computer-based systems are cal assistance, and mutual assistance on enforcement. In more prescriptive than the EU agreements, which suggest addition, for rules of origin an agreement of this kind pro- only cooperation. vides a consultative forum for dispute settlement and for The agreements also focus on a subset of disciplines that amendment of the agreement’s provisions. are of interest to the United States. Advance rulings and In recent negotiations of EPA agreements, cooperation is express shipment stand out. In both cases, the agreements becoming more coercive, with the possibility of retaliation provide for detailed rules that are not present in non-U.S. in case of failure to comply with administrative coopera- free trade agreements (FTAs). They also reflect current U.S. tion. The importance of cooperation mechanisms may be proposals in the WTO negotiations.32 linked to a fourth aspect of the agreements: in most cases, the language of provisions remains unspecific, giving the Other Agreements partner country the opportunity of choosing how to reform its procedures and, in most cases, of defining the objective The trade facilitation agenda, unlike other subjects in PTAs of the reform.31 The provisions thus correspond to the such as intellectual property, is not entirely driven by the Trade Facilitation 343 parallel efforts of the two hubs, Europe and the United effort concerns business mobility and the extension of the States. There is a strong drive to promote trade facilitation APEC business travel card scheme, which facilitates immi- in some countries in Asia and the Pacific. We review here gration procedures and eases entry for business travelers. some of what we believe are the most notable agreements, A fourth characteristic is the tendency to refer fre- drawing largely on Wille and Redden (2007). quently to international and regional initiatives and stan- Asia-Pacific Economic Cooperation (APEC). In 2001 dards, suggesting that the plans aim at promoting an open APEC adopted the Shanghai Accord, which stresses the liberalization agenda. importance of trade facilitation. The accord endorsed the ASEAN Free Trade Agreement (AFTA). Like APEC, APEC Principles on Trade Facilitation as part of a collec- AFTA is nonbinding.34 As noted by Wille and Redden tive action plan and set a target of a 5 percent reduction in (2007), some elements of trade facilitation were present cross-border trade transaction costs by 2006.33 A trade early on in the 1992 AFTA, although they were not explic- facilitation action plan (TFAP) followed in 2002. A second itly identified as trade facilitation. Progress on these trade facilitation action plan (TFAP II) was agreed for the remained limited. The real focus on trade facilitation period 2007–10, with the target of a further 5 percent started in earnest with the ASEAN Vision 2020 plan, initi- reduction in costs. The APEC Trade Facilitation Principles ated in 1997. This was followed by a number of initiatives, encompass a broad agenda: transparency; communication the most recent of which is the Vientiane Action Programme and consultation; simplification; practicability and effi- of 2004. Among the actions that have been undertaken, the ciency; nondiscrimination; consistency and predictability; ASEAN single window, agreed in December 2005, is harmonization, standardization, and recognition; modern- demonstrating some vision. Yet overall, ASEAN actions are ization and the use of new technology; due process; and fragmented, framed by a succession of initiatives and decla- cooperation. rations and sometimes only aspirational. Individual coun- Several characteristics of the APEC approach are worth tries have, however, made good progress on some fronts, highlighting. The first is that on the implementation side, such as reforming customs procedures in the spirit of the actions are strictly voluntary, and therefore implementa- WCO revised Kyoto convention and standardizing infor- tion has not been uniform. Wille and Redden (2007) report mation requirements (Wille and Redden 2007). that 62 percent of the TFAP was completed in 2006. It is Australia–Singapore FTA. The FTA between Australia also unclear whether implementation was a direct conse- and Singapore was signed in November 2002 and came quence of APEC efforts, given the overlapping membership into force in July 2003. An interesting aspect of this agree- of various APEC members in other agreements, including ment is that it involves two advanced nations. Trade facili- NAFTA and ASEAN, as well as decisions to reform trade tation measures are not grouped under a single heading facilitation unilaterally. and are not identified as such, but several provisions of the The second point is that the menu of actions listed in agreement nevertheless have trade facilitation as an aim. the trade facilitation action plans is much more detailed These include provisions on paperless trading, risk man- and ambitious than in EU, U.S., or other binding agree- agement, and publication of information. There are also ments. One measure of the ambition of the action plan is interesting complementary provisions on business mobil- the reference to agendas that go beyond customs to ity and sharing of best practices. include, for instance, services, business mobility, and secu- rity, as under the 2002 Secure Trade in the APEC Region Conclusions (STAR) initiative. Even within customs, the actions sug- gested are deep in nature, with detailed plans for paperless There is a tendency toward greater sophistication in pro- trading, single windows, and sophisticated risk manage- visions concerning trade facilitation in regional trade ment techniques, including advance lodging of informa- agreements. The focus remains very much on customs pro- tion, scanning, and simplified clearance for authorized cedures, even if the language used sometimes seems to relate economic operators. to border agencies in general and in some rare instances is The TFAP II goes even further, exhibiting a third char- complemented by, for example, liberalization measures in acteristic of APEC cooperation—the recent emphasis on services. Most often, provisions in PTAs remain relatively multicountry reform initiatives known as collective unspecific. In particular, clear and measurable objectives are actions/pathfinders. TFAP II lists the APEC single window largely absent. APEC is an exception in this respect. as the joint effort for customs procedures under which a The approach followed by agreements led by the United framework and lessons from country initiatives are to assist States and the EU are closely related to these actors’ posi- individual efforts to build single windows. Another joint tions in the WTO negotiations. This is not necessarily a bad 344 Jean-Christophe Maur thing, as PTAs can be complementary to multilateral the most advanced trade facilitation provisions involve efforts (although they could also be used to circumvent developed and developing countries that are distant from these efforts). It is surprising that PTAs are not more each others, whereas agreements among developing coun- diverse. As outlined earlier, there are specific areas where tries are not deep enough to cover such aspects satisfacto- PTAs could serve to further trade facilitation reform, but it rily. Furthermore, many transit dimensions have been is not obvious that the agreements reviewed here are focus- managed outside trade agreements. Finally, cooperation on ing on this potential. technical assistance, although present in some agreements, There are two paths for reaching agreement on trade remains largely underdeveloped. facilitation provisions in PTAs. The first is through the classic mercantilist approach to negotiations. It would seem that developing countries have much to gain from Notes improved facilitation measures and may wish to negotiate Comments from Toni Matsudaira are gratefully acknowledged. these in exchange for concessions in other domains. This 1. For the symposium, see WTO Secretariat Background Note advice, however, should be qualified by the absolute need G/C/W/80 Rev.1 and Symposium Report G/C/W/115. to ensure that any commitment on trade facilitation provi- 2. Among these bodies are the Confederation of British Industry (CBI); the International Civic Aviation Organization (ICAO); the Inter- sions is much more specific than is currently the case. The national Maritime Organization (IMO); the International Road Trans- second path is through reciprocal benefits arising from port Union (IRU); the United Nations Centre for Trade Facilitation and cooperation. Here, the scope seems more promising for Electronic Business (UNCEFACT); and the United Nations Economic Commission for Europe (UNECE). Agreements may refer to some agreements among developing countries, and probably of these standards; for example, the Association of Southeast Asian among geographically close trade partners. The broad pre- Nations (ASEAN) Free Trade Area refers to the United Nations/Electronic cept behind such mutual cooperation is that it generates Data Interchange For Administration, Commerce and Transport(UN/ EDIFACT); CAFTA–DR cites the World Customs Organization (WCO) positive externalities. Of course, these two characteriza- Customs Data Model; and the EC–CARIFORUM EPA refers to the tions must not be taken wholesale; trade-offs can arise in Revised Kyoto Convention. South-South agreements (for instance, regarding transit 3. TradeMark Southern Africa website, http://www.trademarksa.org/. between landlocked and coastal countries), and coopera- 4. On the substantive differences between regional trade and cooper- ation agreements, see Devlin and Estevadeordal (2004). tion can take place in North-South agreements, on such 5. Communication from the Arab Republic of Egypt to the WTO matters as security issues and cooperation between cus- Negotiating Group on Trade Facilitation, TN/TF/W/75, November 15, toms authorities. 2005, http://docsonline.wto.org/. 6. World Bank (2005, 85) notes, for instance, that the cost of certify- We found little evidence that existing trade facilitation ing organic nut production in Moldova for export to Germany can come provisions contradict nondiscrimination principles. On to US$18,000 per year, a not insignificant amount for firms in poor the contrary, PTAs often tend to reaffirm such principles countries. 7. Arvis, Raballand, and Marteau (2010) examine the large impact and promote international standards. In some instances, associated with uncertainty along the supply chain because of nonharmo- however, PTAs are used to further specific preferred mod- nized regulations. els in such areas as express shipment provisions (in U.S. 8. The Tanzanian certification and testing agency for pesticides agreements) and use of European standards. “charges relatively high fees to register an agro-chemical and also requires three years of field testing. It does not recognize the testing done and reg- Nonbinding agreements such as APEC have gone much istration of chemicals in neighboring countries, including Kenya. Hence, further in promoting trade facilitation reform than have there are a broad range of newer, more effective and safer chemicals which binding agreements. This is partly a reflection of the head do not get registered in Tanzania because of the high cost and which are prevented from being legally imported from Kenya or other neighboring start of APEC over many agreements in this area, but it may countries. The chemical registration revenue imperative of the [testing also highlight some limits in using binding language when agency] thus appears to take precedence over a feasible solution of mutual conceiving international cooperation for trade facilitation. recognition of other (including more rigorous) testing and registration systems� (Tanzania 2005, 97). Soft law has an important role to play. Trade facilitation 9. This is an important facilitating practice for customs valuation. For provisions that focus on reform on the ground, implemen- a discussion, see De Wulf and Sokol (2005), ch. 8. tation, and monitoring of measurable objectives, as in 10. Communication by Norway to the WTO Trade Facilitation APEC, seem to deliver more effective and successful reforms. Negotiating Group on Border Agency Cooperation, TN/TF/W/48, June 9, 2005. In 1995 Norway calculated the savings associated with the two Such agreements tend to be pragmatic, flexible, and country agreements: without them, 10 new customs offices would have had to be specific and are generally well suited to the type of reforms opened on the Norwegian side of the border, and 100 new customs offi- required in the context of trade facilitation. cers would have been needed. The cost would have been about US$8 mil- lion in additional investment and US$8 million in recurring annual costs Transit issues, a very important dimension for countries for new buildings, salaries, and so on. Businesses would have incurred an with limited or no geographic access to main trade routes, estimated US$39 million in additional annual costs because of longer are rarely present in PTAs. The reason is that the PTAs with waiting times and double stops at the border. Trade Facilitation 345 11. Communication from Barbados, Fiji, Papua New Guinea, and the agreement to form the Central Corridor Transit Transport Facilitation Solomon Islands to the WTO Negotiating Group on Trade Facilitation, Agency (CCTTFA), which is modeled along the lines of the NCTA author- TN/TF/W/129, July 7, 2006. ity. The constitution establishing the Dar es Salaam Corridor Coordinat- 12. An essential piece of the architecture for the enforcement of tech- ing Committee involves stakeholders from Malawi, Tanzania, and Zambia. nical regulations and sanitary and phytosanitary (SPS) measures in inter- The Walvis Bay Corridor Group (WBCG), a nonprofit public-private national commercial exchanges is accreditation, which offers an interna- partnership, was established in 2000 and incorporated in Namibia to tionally recognized guarantee that national processes of standards coordinate and integrate public and private sector development efforts conformity assessment can be relied on. Holmes et al. (2006) offer a good along the Walvis Bay Corridor (see box 15.3). Its initial business develop- overview of the issue. In Sub-Saharan Africa there was until recently only ment efforts focused on the Trans-Kalahari Corridor, mainly to facili- one accredited expert, located in South Africa, able to provide accredita- tate institutional cooperation between Botswana, Namibia, and South tion. Three additional experts, also in South Africa, have now been Africa through the Trans-Kalahari Corridor Management Committee trained. This confirms the view that in many developing countries, mar- (TKCMC). In 2003 the three governments signed a formal trilateral corri- kets for accreditation and certification bodies may be too small. dor agreement, and the TKCMC was established through a memorandum 13. Arvis, Raballand, and Marteau (2010) show how advanced logis- of understanding. The WBCG currently serves as the secretariat for the tics services are inhibited by lack of trade facilitation: it is often impossi- TKCMC. A similar body, the Walvis Bay-Ndola-Lubumbashi Corridor ble, note the authors, to maintain multicountry inventories or to avoid Committee, has been set up for the Trans-Caprivi Corridor, and an analo- first clearance and then reexport to the gateway country. gous approach is being taken toward the Trans-Cunene Corridor. 14. Payment guarantee systems require less working capital than pay- 26. The exchange of commitments, however, is only likely when part- ment in advance (which is required when there is no guarantee), and thus ner countries have similar levels of development and patterns of compar- they help smaller agents access international trade. ative advantage that make movements of natural persons relatively bal- 15. This shortcoming has prompted EBRD and the International anced. Bin (2008) finds that provisions on mobility of business persons Finance Corporation (IFC) to create international risk-sharing funds to are present in about one-third of PTAs containing trade facilitation provi- provide small enterprises with access to trade finance. These funds help sions in Asia and the Pacific. international banks (confirming banks) cover the political and commer- 27. For instance, regional guarantee systems help establish global cial risk faced by local issuing banks when they guarantee international standards for documentary credit (EBRD 2003) and thus generate infor- trade transactions. mation that can be used for other purposes. 16. As an instance of economies of scale, air transport hubs avoid the 28. For the areas and state of progress, see APEC, Sub-Committee on inefficiency of empty cargo capacity on incoming or outgoing freight, Customs Procedures, “CAP Assessment/Evaluation Matrix, Summary by which is a problem for small, nondiversified economies. Economy, July 2009,� 2009/SOM2/SCCp/016. 17. McTiernan (2006) reports that Benin and Togo charge very high 29. Another way to create ownership for businesses is to give them fees for transit, which gives an incentive for transport from Lagos to Accra access to dispute settlement under the PTA, as NAFTA and CAFTA do for to be carried out by ship instead of overland. investment. Similar solutions could be envisaged in relation to trade facil- 18. The failure stems not merely from oversight or neglect but also itation, offering the possibility for the private sector to challenge govern- from lack of incentives for the country of transit to internalize the costs of ments that illegitimately restrict their business. more efficient transit. 30. Similarly, provisions for common institutions are absent in the 19. A counterexample is Bolivia, which has several access roads to the agreement with CEMAC but are quite developed in other agreements. sea (Schiff and Winters 2003). 31. For example, according to Article 31 of the EU-CARIFORUM 20. Arvis, Raballand, and Marteau (2010) cite the poor performance EPA, “[the parties] shall simplify requirements and formalities where pos- and unpredictability of Tanzanian railways, which have an error margin of sible, in respect to the rapid release of goods.� four to five days in predicting the arrival of a shipment, as a reason for the 32. See the proposals to the WTO by Australia, Canada, Turkey, and increase in road transit from Kenya to northern Tanzania. That traffic the United States, TN/TF/W/153, and by the United States, TN/TF/ increased by 20 percent over five years. Similarly, 75 percent of Rwandan W/144/Rev.2. trade now transits through Kenya, whereas 50 percent went through Tan- 33. The parties are Australia; Brunei Darussalam; Canada; Chile; zania only three years prior to the authors’ survey. China; Hong Kong SAR, China; Indonesia; Japan; the Republic of 21. The TIR is an international transit regime adopted by 68 parties, Korea; Malaysia; Mexico; New Zealand; Papua New Guinea; Peru; primarily in Europe, the Middle East, North Africa, and Central Asia. It the Philippines; the Russian Federation; Singapore; Taiwan, China; was initially known in French as Transport Internationaux Routiers but is Thailand; the United States; and Vietnam. now only referred to as TIR in legal texts. 34. The parties are Brunei Darussalam, Cambodia, Indonesia, Lao PDR, 22. Cooperation provisions on rules of origin have traditionally been Malaysia, Myanmar, the Philippines, Singapore, Thailand, and Vietnam. included with the provisions on rules of origin; chapters on customs cooperation deal with other elements of cooperation. 23. Aldaz-Carroll (2006) provides an example of such difficulties for References regional standards harmonization in sensitive sectors in Mercosur. 24. As noted by Arvis (2010), it may be desirable for the transit sys- Adzibgey, Yao, Charles Kunaka, and Tesfamichael Nahusenay Mitiku. tem to evolve according to the level of integration among countries. Tran- 2007. “Institutional Arrangements for Transport Corridor Manage- sit can indeed work as a series of independent national systems, but there ment in Sub-Saharan Africa.� SSATP Working Paper 86, Sub-Saharan are significant gains from cooperation, such as harmonization and infor- Africa Transport Policy Program, World Bank, Washington, DC. mation sharing, which can be implemented gradually. Aldaz-Carroll, Enrique. 2006. “Regional Approaches to Better Standards 25. The Northern Corridor Transit Agreement (NCTA) was signed in Systems.� Policy Research Working Paper 3948, World Bank, Washing- 1985 by Burundi, Kenya, Rwanda, and Uganda to promote the use of the ton, DC. Northern Corridor as an effective route for the surface transport of goods Arce, Daniel G., and Todd Sandler. 2002. Regional Public Goods: Typologies, between the partner states. The Democratic Republic of Congo joined the Provision, Financing, and Development Assistance. Stockholm: Almqvist corridor group in 1987. Trade facilitation along the Central Corridor was and Wiksell International for Expert Group on Development Issues, until recently handled through bilateral agreements, mainly between Tan- Swedish Ministry of Foreign Affairs. zania and the landlocked countries. In 2006 Burundi, the Democratic Arnold, John. 2006. “Best Practices in Management of International Trade Republic of Congo, Rwanda, Tanzania, and Uganda signed a multilateral Corridors.� Transport Paper 13, World Bank, Washington, DC http:// 346 Jean-Christophe Maur siteresources.worldbank.org/INTTRANSPORT/Resources/336291- Moïsé, Evdokia. 2005. “Trade Facilitation Reforms in the Service of Devel- 1227561426235/5611053-1229359963828/itc-1-11-07.pdf. opment: Country Case Studies.� OECD Trade Policy Working Paper Arvis, Jean-François. 2005. “Transit and the Special Case of Landlocked 12, Organisation for Economic Co-operation and Development, Paris. Countries.� In Customs Modernization Handbook, ed. Luc De Wulf and Mozambique. 2004. “Integrated Framework.� Vol. 1 of “Mozambique: Diag- Jose B. Sokol, ch. 11. Washington, DC: World Bank. nostic Trade Integration Study.� World Trade Organization, Geneva. ———. 2010. “Transit Regimes.� In Border Management Modernization, Müller-Jentsch, Daniel. 2003. Transport Policies for the Euro-Mediter- eds. Gerard McLinden, Enrique Fanta, David Widdowoson and Tom ranean Free-Trade Area: An Agenda for Multimodal Transport Reform Doyle, ch. 7. Washington, DC: World Bank. in the Southern Mediterranean. World Bank Technical Paper 527. Arvis, Jean-Francois, Gael Raballand, and Jean-Francois Marteau. 2010. Washington, DC: World Bank. The Cost of Being Landlocked: Logistics Costs and Supply Chain Reliabil- OECD (Organisation for Economic Co-operation and Development). ity. Directions in Development: Trade. Washington, DC: World Bank. 2006. “OECD/DAC Project on Trade Facilitation: Phase 1: A Review of Bin, Peng. 2008. “Trade Facilitation Provisions in Regional Trade Agree- Technical Assistance.� OECD Papers 6 (2). ments in Asia and the Pacific.� Staff Working Paper 01/08, UNESCAP Rufin, Carlos. 2004. “Regional Public Goods and Infrastructure.� In Trade and Investment Division, United Nations Economic and Social Regional Public Goods: From Theory to Practice, ed. Antoni Estevade- Commission for Asia and the Pacific, Bangkok. ordal, Brian Frantz, and Tam Robert Nguyen. Washington, DC: Inter- Brooks, Douglas H. 2008. “Regional Cooperation, Infrastructure and American Development Bank. Trade Costs in Asia.� ADB Institute Working Paper 123. Asian Devel- Sandler, Todd. 2006. “Regional Public Goods and International Organiza- opment Bank Institute, Tokyo. tions.� Review of International Organizations 1 (1, March): 5–25. Cadot, Olivier, Celine Carrère, Jaime de Melo, and Bolormaa Sauvé, Pierre, and Americo B. Zampetti. 2000. “Subsidiarity Perspectives Tumurchudur. 2006. “Product-Specific Rules of Origin in EU and US on the New Trade Agenda.� Journal of International Economic Law Preferential Trading Arrangements: An Assessment.� World Trade 3 (1): 83–114. Review 5 (2): 199–224. Schiff, Maurice, and L. Alan Winters. 1998. “Regional Integration as Devlin, Robert, and Antoni Estevadeordal. 2004. “Trade and Cooperation: Diplomacy.� World Bank Economic Review 12: 271–95. A Regional Public Goods Approach.� In Regional Public Goods, ———. 2002. “Regional Cooperation and the Role of International ed. Antoni Estevadeordal, Brian Frantz, and Tam Robert Nguyen. Organizations in Regional Integration.� Policy Research Working Washington, DC: Inter American Development Bank and Asian Paper 2872, World Bank, Washington, DC. Development Bank. ———. 2003. Regional Integration and Development. Washington, DC: De Wulf, Luc, and Jose B. Sokol. 2005. Customs Modernization Handbook. World Bank. Washington, DC: World Bank. Stevens, Christopher, Mareike Meyn, Jane Kennan, Sanoussi Bilal, EBRD (European Bank for Reconstruction and Development). 2003. Corinna Braun-Munzinger, Franziska Jerosch, Davina Makhan, and “Regional Trade Facilitation Programme: Fostering Transition through Francesco Rampa. 2008. “The New EPAs: Comparative Analysis of Documentary Credit.� Evaluation Department, EBRD, London. Their Content and the Challenges for 2008.� Final Report, Overseas Fasan, Olu. 2004. “Comparing EU Free Trade Agreements: Trade Facilita- Development Institute and European Center for Development Policy tion.� ECDPM InBrief 6F, European Centre for Development Policy Management, London. Management, Maastricht, Netherlands. Tanzania. 2005. “Integrated Framework.� Vol. 1 of “Tanzania: Diagnostic Haddad, Mona. 2008. “Trade Issues in East Asia: Overcoming Trade Barri- Trade Integration Study.� World Trade Organization, Geneva. ers from Standards and Technical Regulations.� Policy Research Toledano, Joëlle, and José Ansón. 2008. Postal Economics in Developing Report, East Asia and Pacific Region, World Bank, Washington, DC. Countries: Posts, Infrastructure of the XXIst Century? Bern: Universal http://www-wds.worldbank.org/external/default/WDSContentServer/ Postal Union. WDSP/IB/2008/02/28/000333037_20080228050849/Rendered/PDF/ Uganda. 2006. “Integrated Framework.� Vol. 1 of “Uganda: Diagnostic 427200WP0P10991port0ASEAN0Standards.pdf. Trade Integration Study.� World Trade Organization, Geneva. Holmes, Peter, Leonardo Iacovone, Rungroge Kamondetdacha, and Lara Wille, P., and J. Redden. 2007. “A Comparative Analysis of Trade Facilita- Newson. 2006. “Capacity Building to Meet International Standards as tion in Selected Regional and Bilateral Trade Agreements and Initia- Public Goods.� UNIDO Working Papers, United Nations Industrial tives.� In ESCAP, Trade Facilitation beyond the Multilateral Trade Nego- Development Organization, Vienna. tiations: Regional Practices, Customs Valuation and Other Emerging Horn, Henrik, Petros C. Mavroidis, and André Sapir. 2010. “Beyond the Issues—A Study by the Asia-Pacific Research and Training Network on WTO? An Anatomy of EU and US Preferential Trade Agreements.� Trade, 27–76. New York: United Nations. World Economy 33 (11): 1565–88. World Bank. 2005. “Beyond Trade Policy Barriers: Lowering Trade Costs McTiernan, Andrew. 2006. “Customs and Business in Africa: A Better Way Together.� In Global Economic Prospects 2005: Trade, Regionalism, and Forward Together.� Report prepared for Business Action for Improv- Development, ch. 4. Washington, DC: World Bank. ing Customs Administration in Africa (BAFICAA). http://webarchive Yang, Yongzheng, and Sanjeev Gupta. 2005. “Regional Trade Agreements .nationalarchives.gov.uk/20100918113753/http://www.sitpro.org.uk// in Africa: Past Performance and the Way Forward.� IMF Working policy/baficaa/baficaareport.pdf. Paper WP/05/36, International Monetary Fund, Washington, DC. 16 COMPETITION POLICY Kamala Dawar and Peter Holmes There is general agreement that competition among firms (PTAs) and the costs and benefits involved. It then surveys enables consumers to enjoy freedom of choice, low prices, representative arrangements between countries in the and good value for money, while at the same time promoting global North and the global South and at differing stages innovation and higher standards. On the national level, the of development. Finally, it analyzes the strengths and need for regulation to prevent anticompetitive practices is, shortcomings of the several regional competition policy accordingly, widely accepted. models and of specific agreements, explores questions of On the plane of international trade, the competition third-party discrimination and trade diversion, and looks policy issues are more complex. Abuse of market power at the practical implementation of the agreements. can span markets and national boundaries, and many The discussion leads to the conclusion that regional countries lack a competition policy framework that would competition provisions can create an incentive for imple- facilitate cooperation with other countries. menting national competition policy regimes, with a view The inclusion of competition provisions in trade agree- toward locking in such policies, increasing foreign direct ments is potentially beneficial—particularly for developing investment (FDI), and, in the case of North-South agree- countries, which suffer disproportionately from cross- ments, promoting technical assistance and learning by border anticompetitive practices. Competition law and doing. All these interactions have the potential to generate policy inherently contribute to better balance between the beneficial regional public goods. rights of producers and protection for consumers and other members of society. A well-administered competi- Competition Policy and Development: tion law will have positive spillover effects on the economy A Survey of the Literature at large, not just the particular firms or groups that bring complaints. It is argued here that, in principle, PTAs can address market The extent to which regional competition provisions in failures that national competition laws cannot and that trade agreements can promote regional public goods and they can offset, to some degree, the absence of an interna- deal with market failures depends on the nature of the pro- tional regime. This conclusion is not self-evident. Even if it visions and on their implementation and enforcement. A can be shown that optimal competition provisions in PTAs small group of countries has begun to develop cooperative are beneficial, it does not follow that what has actually been practices and appears to be active in initiating such agree- negotiated is ideal. Nevertheless, the contention in this ments. These are primarily developed countries with estab- chapter is that existing competition provisions do have a lished national competition law, existing agencies, and a potential for positive effects. strong competition culture. In other regional competition The desirability of competition is (mostly) taken for regimes, such provisions are the beginnings of state-to-state granted in advanced industrial economies. Some consen- practices that are likely to develop over time. The develop- sus exists that competition is good for economic develop- ment and effective implementation of national law and pol- ment and that the natural selection process of the market icy regarding competition and consumer protection are cannot be entirely relied on to ensure that firms can enter essential complements to regional competition policy. and exit as freely as possible. This chapter first discusses the economic case for includ- In developing countries, there are doubts about the abil- ing competition provisions in preferential trade agreements ity of markets to function so as to deliver the gains from 347 348 Kamala Dawar and Peter Holmes competition. There is suspicion of strong competition, on affected by foreign firms’ anticompetitive behavior. Evenett, the grounds that it will merely ensure the “survival of the fat- Levenstein, and Suslow (2001) show that even in globalized test� and lead to dominance by large firms. Some developing markets, cartels do not necessarily collapse rapidly of their countries, in their early period of industrialization, tried to own accord. Many cartels are based in developed countries, limit the severity of competition, especially from imports, to where their activities are legal under the laws of the protect their own enterprises. Only the most dogmatic mar- exporting countries, as long as the effects are confined to ket fundamentalist could deny that there are models in foreign markets.1 Competition authorities in developed which monopoly capitalism emerges and circumstances in countries are largely forbidden to provide confidential which it might prove beneficial. Examples can always be information on cartel activity conducted abroad, without selected in which intensified competition went wrong, and specific legal authority.2 U.S. and European Union (EU) economists need to be modest in urging the gains for devel- authorities have traditionally considered themselves capa- opment to be had from strong competition. ble of addressing foreign anticompetitive practices that In fact, however, recent research does support the thesis affect their own interests, through appeal to the effects doc- that competition is good for development, both stimulat- trine.3 In practice, however, the difficulty and cost of ing new business and benefiting consumers. Dutz and obtaining evidence on the behavior of firms located in for- Hayri (1999), after surveying the existing empirical evi- eign jurisdictions can make it difficult to bring cases. dence and conducting a major cross-country study, found As Phlips (1998) has argued, without clear evidence of a strong correlation between long-run growth and effective motivation and collusion, the burden of proof in cases enforcement of antitrust and competition policy. A recent based on overseas evidence can be very hard to sustain. An study by the United Nations Conference on Trade and example is the 1988 case in which the European Court of Development (UNCTAD) reviewed the literature and Justice rejected, on appeal, the European Commission’s commissioned case studies around the world. The results attempt to fine a cartel of foreign wood pulp producers for (Brusick et al. 2004) broadly support the view that compe- price-fixing (Vedder 1990). In the 1986 Matsushita v. Zenith tition is good for development, and the authors argue that case, the U.S. Supreme Court found cooperative conduct competition policy is very much complementary to other legal under antitrust law, although the actions did violate instruments for encouraging enterprise development. antidumping rules (Belderbos and Holmes 1995). Skeptics argue that although competition is desirable, In fact, the absence of international rules on mandating the difficulties, expense, and skilled-staff requirements cooperation on competition policy enforcement increases involved in making competition policy work effectively the temptation to resort to antidumping rules. An inter- render it unlikely that developing countries will see much national agreement on competition policy would not benefit from it. At best, the effort will be an expensive mean, however, that antidumping rules would be super- waste; at worst, it will present a further opportunity for seded. These rules are designed to cover far more types of regulatory capture, as incumbents or other potential losers behavior than does competition policy (see Bourgeois use competition policy to frustrate rather than foster com- and Messerlin 1999; Sykes 1999) and have overtly protec- petition. We cannot rule out this result a priori. There is, tionist motives. however, considerable evidence that if governments are The case for some form of international cooperation on willing to let their competition agencies act, these bodies competition policy is very strong unless one is skeptical of can be effective. Extensive case studies that are reported in antitrust policy as such.4 The EU’s experience with interna- CUTS (2003) and Brusick et al. (2004) corroborate the tional restrictive business practices has contributed to the findings of Dutz and Hayri’s (1999) cross-sectional statisti- belief that reduction of trade barriers alone is inadequate. cal analysis. The case studies suggest that competition pol- It became clear that oligopolistic firms were able to divide icy can be made to work in developing countries to the EU markets after border barriers were lifted and that anti- benefit of both development and the consumer, at costs competitive price discrimination could occur. Given the that, although not trivial, are modest in relation to the con- propensity of member states to protect their own national sumer savings from successful interventions. Dutz and champions, the EU’s founders early on deemed a suprana- Vagliasindi (2002) also find that effective enforcement of tional competition policy necessary. competition policy in transition economies is associated Anderson and Jenny (2005, 67) observe, with more rapid entry of new firms. Competition policy In the 1990s, extensive evidence surfaced that international offers more than just static gains from lower prices. cartels are alive and flourishing in the “globalising� eco- The international dimension of competition policy is nomic environment. Investigations conducted by the US clearly important to developing countries that are adversely Department of Justice, the European Commission, the Competition Policy 349 Canadian Competition Bureau and authorities in other opposed attempts to use trade agreements to press devel- jurisdictions revealed the existence of major cartels in (to oping countries to adopt competition laws. Hoekman and cite but a few of many examples) the following industries: Holmes (1999) contend that a major flaw in the EU pro- graphite electrodes (an essential input to steel mini-mill posals for a World Trade Organization (WTO) multilat- production); bromine (a flame retardant and fumigant); eral competition policy agreement was that the scheme citric acid (a major industrial food additive); lysine (an would have imposed burdensome administrative require- agricultural feed additive); seamless steel pipes (an input to ments that were not worth the benefits, since the propos- oil production); and vitamins. als on the table did not include any obligations by devel- Levenstein and Suslow (2001) argue that reductions of oped countries to provide assistance in cases against 20–30 percent in the prices of developing-country imports export cartels that affect developing countries. Neverthe- of products known to be affected by cartels would yield less, Hoekman and Saggi (2004) argue that developing benefits greater than those from a 50 percent cut in agricul- countries may be able to profit from bilateral deals in tural tariffs. Utton (2008) finds that a large share of the which they agree to the adoption of a competition regime competition problems of most developing countries in exchange for market access in the partner country. It emanates from the international sector, from such sources should be evident that the most attractive deal is one in as import distribution monopolies and cartels; the influ- which a developed country agrees to make real coopera- ence of dominant firms based in other countries, including tion available to a developing-country partner. neighbors; overseas export cartels; and regional market All in all, there is little doubt that global competition sharing. These market failures can be difficult to deal with problems exist, and it is generally accepted that interna- unilaterally, depending on where the unlawful conduct tional cooperation is desirable. What is less clear is what takes place and where the evidence is located. If the con- shape such cooperation should take. What works and what duct takes place abroad, prosecution becomes problematic. does not is essentially an empirical matter. It is therefore Export cartels are notoriously hard to chase because they necessary to identify and examine the form and history of deliberately collude outside the jurisdiction of the import- competition provisions in PTAs. ing country, and the same goes for regional market sharing. Where, however, an abuse of dominant position extends Economics of Competition Provisions beyond a national market, there is a clear administrative case for regional cooperation, as in the EU. Competition law and policy designed to regulate and curb The anticompetitive practices of multinational corpora- anticompetitve practices are now common at the national tions are disadvantageous to developing countries and have level. Anticompetitive practices include abuse of market particularly detrimental consequences in a context of eco- dominance, collusion between firms, mergers and acquisi- nomic scarcity. Levenstein and Suslow (2001) examined 16 tions that secure a dominant market position, artificially products that were cartelized during the 1990s and for restricted output that leads to artificially high prices, which reasonably reliable trade data were available. They predatory pricing, and price fixing. estimated that the total value of such “cartel-affected� goods Competition policy issues are more complex under imported to developing countries was US$81.1 billion, a globalization because abuse of market power can occur sum equivalent to 6.7 percent of all imports to these coun- unevenly across several markets and beyond the jurisdic- tries and 1.2 percent of their combined gross domestic tion of a national authority. Restrictive business practices product (GDP).5 may be carried out by domestic producers on foreign mar- This evidence suggests that developing countries should kets, or by foreign producers on domestic markets. The actively work for international cooperation on competi- effectiveness of the prohibition of these anticompetitive tion policy. Interestingly, they did just that before the practices will depend on the engagement of the various Uruguay Round of the General Agreement on Tariffs and competition agencies, on whether competition law embod- Trade (GATT). Officials in India, to take one instance, have ies the extraterritoriality principle, and on the degree to frequently expressed their support for UNCTAD’s princi- which agencies cooperate in addressing behavior that only ples and rules for the control of restrictive business prac- one jurisdiction may view as harmful. tices (UNCTAD 2000).6 It is therefore surprising that so The lack of a comprehensive and coherent approach much controversy surrounds the link between trade policy toward cross-border competition issues has led to propos- and competition policy. als for a binding multilateral competition regime. Such a Many authors have, however, argued that free trade is framework is currently rejected within the WTO. Mean- itself the best competition policy and have, in particular, while, competition provisions are increasingly being 350 Kamala Dawar and Peter Holmes included in regional trading agreements and bilateral com- for promoting competition in regional trade agreements, petition arrangements, as a halfway house or stepping- some relevant issues can be raised. stone toward agreement on the international level. This Institutional and behavioral shortcomings—lack of the development has not been unanimously welcomed, partly requisite competition culture or of the political will to pro- because any form of regionalism creates trade diversion mote domestic implementation—contribute to poor and preferentialism, and partly because of the costs of implementation of regional competition provisions. A negotiating and implementing regional competition provi- well-designed regional competition agreement needs to sions, at a time when experience with the economic and take account of these local realities and to foster the welfare effects of such regimes is as yet inconclusive. national structural and behavioral environment necessary Some analysts favor the narrower but stronger option to draw benefits from the regional competition provisions. of negotiating a mutual legal assistance treaty (MLAT) In regional groupings where individual members are at that specifies how competition authorities may assist one very different stages of economic development, where another in securing and sharing evidence that is not read- some lack a competition law or functioning enforcement ily obtainable. Under the MLAT between Canada and the agency, and where approaches to sovereignty pooling dif- United States, for example, Canada has requested that fer, establishment of a strong regional enforcement mecha- U.S. authorities obtain documentary evidence and testi- nism can be beneficial. Examples of groups in this category mony from U.S. corporate offices, using compulsory pro- are the Andean Pact, the Common Market for Eastern and cedures. Canada has also assisted in the execution of Southern Africa (COMESA), and the Southern Cone Com- search warrants at the premises of a firm in Canada that mon Market (Mercosur, Mercado Común del Sur). was allegedly party to felony violations of U.S. antitrust The case of Mercosur is instructive because its intergov- laws. The documentary and other evidence Canada pro- ernmental ministerial approach to regional competition vided to U.S. authorities contributed to the initiation of law has not been implemented effectively. This kind of fail- grand jury investigations in the United States. ure is usually attributed to the unwillingness of some Clearly, improvement of interagency relationships and members to enact national laws or to set up domestic cooperation mechanisms will facilitate coordination of regimes that can give effect to external regional obligations. competition investigations and prosecutions. But MLATs De Araújo (2001) notes that the Mercosur competition are not a complete solution; they cannot be applied in policy agreement, the Fortaleza Protocol, requires member jurisdictions that lack competition agencies, and they are states to have a national competition law but creates no less effective where national agencies are unequal in collective agency; Paraguay and Uruguay chose to ignore expertise, resources, and enforcement mechanisms. More the protocol. The lack of effective regional competition comprehensive regional competition regimes can poten- remedies may have undermined some of the benefits tially overcome some of these challenges and discrepancies, expected from the free trade schedules for this common and to them we next turn. market. Some Mercosur members have developed compe- tition law and policy, but unless all members of an agree- ment have an effective domestic law, there can be no legal Benefits of Various Types of Agreement basis for action by a member against practices organized in In principle, the economic rationale for including competi- another member state. tion law and provisions in PTAs is to prevent liberalization North-North competition agreements are more con- from being undermined by anticompetitive business prac- ducive to cooperation and coordination of activities than tices within the region, to the disadvantage of consumers other types of PTA and may generate more benefits, but and firms. Competition policy provisions in PTAs there- there are gains to be had from North-South regional com- fore have two aims: to ensure that the partner’s enforce- petition regimes, as well. Developing countries in these ment (or nonenforcement) of competition policy does not arrangements will tend to benefit from such competition undermine the market access preferences granted in the provisions as cooperation in enforcement activities and agreements, and to guarantee that cross-border competi- technical assistance. The benefits have been notable in the tion policy issues are dealt with adequately through regula- case of the Brazil–U.S. competition cooperation agreement, tory cooperation. Beyond this rule of thumb, there is little which, among other things, provides for U.S. technical evidence to support definitive conclusions about the eco- assistance that has helped improve Brazil’s expertise in the nomic benefits of the different types of competition- field of competition law and policy. This has enabled Brazil related provisions found in PTAs. Although not enough is to communicate and cooperate with the United States in yet known about the relative merits of different measures confronting anticompetitive practices. In cases like this, Competition Policy 351 consultations between the competition agencies provide an Positive Spillovers opportunity for one agency to offer its support, advice, and National competition law and policy constitute public experience to its counterpart. Notification of enforcement goods. Notably, competition provisions respond to market actions enables authorities to compare information about failures such as cartel creation or abuse of dominant mar- particular cases, and the provisions for technical assistance ket position. The provisions are designed to ensure that can be very helpful in building up capacity and expertise in the benefits of liberalization are not undermined by pri- the field. A notable example was the extensive cooperation vate restrictive business practices and to promote more the South African competition authorities received from the efficient, fairer markets. This is in the interests of both EU regarding the international merger of SmithKline businesses and consumers. Competition law and policy Beecham PLC and Glaxo Wellcome PLC. The South African inherently help balance the rights of producers and the Competition Tribunal explicitly noted that its decision protection accorded to consumers and other members of against the merger proposal was largely based on the EU’s society. stand, and both South Africa’s Competition Commission Strictly speaking, competition law is not a pure public and the EU found that the merger would significantly affect good, in the sense that the use of competition agency competition in similar markets. The merger was eventually resources to pursue a case brought by one set of interests approved subject to the merging parties’ outlicensing some occurs at the expense of those whose cases cannot be products in specific areas in order to reduce their post- heard, and certain types of firm can be denied standing on merger market share (CUTS 2003). the basis of nationality. There is no doubt, however, that a North-South PTAs will yield greater development bene- well-administered competition law will have positive fits and will have better implementation records when (a) spillover effects on the economy at large, and not just on the more developed party offers appropriate technical and those firms or groups whose cases are adjudicated. capacity-building assistance to the less developed one, and At a regional level, competition provisions can produce (b) the less developed regional partner is able to benefit regional public goods by regulating cross-border trade, as from the assistance. For those members with nascent or well as mergers and acquisitions. There is a growing need nonexistent competition regimes, technical assistance for regional cooperation to address cross-border anticom- should aim to impart the required expertise and experience petitive practices. Unilateral competition measures under- over the long term, so as to promote the generational taken by national governments will not yield the same behavioral changes necessary for a competition culture. magnitude of regional public goods. Cooperation and coordination will require an agreement between the par- Constructing Effective Enforcement Mechanisms ties and may lead to complementary regional institutional arrangements or mechanisms. The institutions that moni- PTAs that have a goal of deepening integration between the tor and enforce regional competition rules and regimes members through a customs union or common market will be shared among countries, as will the ensuing bene- may find it economically advantageous to design effective fits. Countries, by working together, will induce beneficial regional competition enforcement mechanisms. This will cross-border spillover, through, for example, information be in addition to efforts to promote competition on a provision or cooperation in enforcing competition law in national level in PTA members that are in the early stages the region. In addition, regional agencies are able to realize of implementing competition. When, in highly centralized economies of scale: even if a competition agency with arrangements, a regional authority is established to assist twice the economic or geographic reach as another costs implementation, it must be given (if it is not to be a paper more than the narrower one, the cost increase is less than tiger) strong investigative powers, adequate resources and proportionate. expertise, and the ability to issue cease-and-desist orders PTA competition provisions, where they have been nego- and collect fines. Transition economies in deep integration tiated, generally insist on core principles that include PTAs facing a strong legacy of statist economies may find it nondiscrimination, due process, and transparency. Although economically useful to include, as well, effective regulations these commitments are made regionally, there is a positive on state aid and antidumping policy. Even in those coun- spillover, and their effect is multilateral (Kulaksizoglu tries with experience of competition, implementation of 2004). For example, Turkey’s competition policy was estab- regional competition law will be more successful if the lished as a result of a bilateral agreement with the EU, but a provisions explicitly promote partners’ existing domestic U.S. or Japanese firm operating in Turkey will benefit from policy priorities, such as support for small businesses or it as much as will a Turkish or European firm. disadvantaged communities. 352 Kamala Dawar and Peter Holmes The extent to which a regional competition regime can in e-commerce, but also with respect to other deceptive deal with market failures depends on the comprehensive- practices, scams, and spam. These domestic agencies have ness of the provisions and the will to enforce them. little or no basis for acting against domestic entities that Regional competition provisions that address cartels col- are causing market injury to consumers outside the coun- lectively will yield regional public goods. Those that do not try (although domestic consumers may have effective explicitly prohibit such restrictive business practices will recourse against the same practice). Similarly, the agencies not produce any such benefits. Where the agreement deals may have no clear authority or capacity to take action with cartels, the publication and notification of cartel against entities that are located, or conducting business enforcement actions in one country will generally stimu- from, outside the domestic territory and that are targeting late enforcement efforts in other countries—a form of or entering into transactions with domestic consumers. At competition advocacy. This effect is particularly germane the national level, the ability to enforce injunctions or cease- where a formal framework exists that establishes a relation- and-desist orders to protect consumers across national bor- ship between competition authorities. Cross-jurisdictional ders is very limited, leading to lack of consumer confidence and multijurisdictional information exchange also pro- in cross-border transactions. motes the investigation and successful prosecution of cross-border restrictive business arrangements such as Implementation Costs international cartels. The costs of implementing the competition provisions of PTAs will depend on their nature and objectives and on Consumer Protection the existing domestic competition framework and level of Lack of adequate, comprehensive consumer protection is a competition culture. The challenge for negotiators and detriment to the achievement of healthy, competitive mar- policy makers is to craft the competition provisions so kets, as well as of healthy consumers. If explicit consumer that the accruing benefits are seen to exceed the imple- protection provisions are included in the regional compe- mentation costs. Regional competition laws and provi- tition regime, there is potential for realizing greater sions for cooperation between competition enforcement regional public goods. Such provisions can help preserve agencies can increase the success and efficiency of the par- the dynamic potential of consumers, while ensuring that ties’ efforts to reduce the negative impact of restrictive consumer protection measures do not become unneces- business practices. sary barriers to trade and that they are consistent with Those decentralized agreements which require the exis- international trade obligations. A regional arrangement to tence of local competition law and the authority to apply protect consumer welfare will prevent cross-border firms the law nationally, such as the North American Free Trade from locating in a jurisdiction with relatively lax consumer Agreement (NAFTA) and the U.S.–Chile agreement, will policy, which would make cross-border consumer com- not be as economically demanding as a regime that estab- plaints and redress difficult to enforce. A regional regime is lishes a fully centralized law with a supporting regional also better able to cope with information asymmetries in authority, as in the EU or COMESA. The provisions in cus- such areas as registries of licensed businesses, e-commerce toms union agreements are, in general, more specific and regulations, and so on. demand higher commitments from the parties because Regional consumer policy can address the negative their goal is regional integration. spillover effects of cross-border anticompetitive business In North-South competition agreements in which practices and can also deliver economies of scale. In order cooperation is limited to the exchange of specified infor- to deal fully with regional market failures, however, com- mation and nonmandatory notification, the costs involved plementary regional consumer policy that focuses directly are those associated with human resources (technical assis- on the cross-border demand side is needed. It must be tance, capacity building, and so on), communication, and geared toward collecting information and evidence travel. These economic burdens will be offset if the parties about practices that may be particularly injurious to are able to exchange information effectively (thus con- consumers, in a situation in which competitors can avoid tributing, for example, to the successful conclusion of an harm because they are able to pass on the costs of restric- investigation) and avoid duplication and conflicting tions to the ultimate consumers. decisions. Domestic consumer agencies increasingly acknowl- For developing countries, the preconditions for suc- edge their inability to identify legislative and enforcement cessfully implementing even the most minimal coopera- gaps in cross-border consumer protection—particularly tion provisions come at a price. For example, a qualified Competition Policy 353 staff supplied with adequate resources is needed; it has only if the agreement produces substantial and beneficial been estimated that if a country were to report every sin- multijurisdictional cooperation between the parties. It gle investigation that might have an impact in a counter- has been reported, for example, that the notification pro- part’s jurisdiction, at least five staff members would cedure specified in the competition cooperation agreement be required (Rosenberg and de Araújo 2005). Even if between Argentina and Brazil is burdensome for both developed countries entering into North-South agree- countries and that its application is not systematic (Botta ments have well-established domestic competition frame- 2009). In a context of scarce human and material resources, works, implementation of competition provisions still such a provision cannot be routinely carried out because its entails costs—both the costs related to human capacity immediate costs exceed its perceived benefits. and administration, and the political costs of alienating In general, the costs should be proportionate to the potential business support or releasing confidential or benefits that can reasonably be expected. The provisions agency information. negotiated should be as simple as possible, focusing pri- The provisions of the Revised Treaty of Chaguaramas, marily on information exchange, technical assistance, which established the Caribbean Community (CARICOM), and capacity building. Subsequently, the commitments indicate the potential economies of scale offered by regional can be expanded to include, for instance, provisions on cooperation, by allowing for resource pooling among mandatory notification and comity. Where developing neighboring countries when national capacity is not countries with less well established domestic competi- adequate for implementing and enforcing the regional tion frameworks are parties to regional competition pro- framework. These provisions have been imported into visions, notification provisions should be mandatory other arrangements negotiated by CARICOM, such as the only for the most important cases for both jurisdictions. competition chapter in the economic partnership agree- More general commitments should be implemented only ment (EPA) between the EU and the Caribbean Forum of after the necessary expertise and cooperation mecha- African, Caribbean, and Pacific States (CARIFORUM) nisms have been developed. (Dawar and Evenett 2008). Many members of regional agreements are small Many North-South regional competition regimes spec- economies with insufficient resources to fund national ify that the developed parties will provide technical assis- competition agencies. In some of the developing countries tance. The cooperation provisions tend to be used primarily that have created a national competition law, the law is not as capacity-building tools, whereby the more mature agency always well understood or adequately enforced. Given the helps develop the expertise of the newly established one, competing short-term pressures on scarce resources, rather than as a way of effectively coordinating enforcement national governments do not always look favorably on activities. Capacity-building activities involve monitoring, regional competition provisions that would yield benefits communication, travel, and staff costs, which are borne only in the longer term. It is for this reason that advocacy only by the northern parties to the agreements. and promotion of a competition culture are so important In North-North and South-South PTA competition in the early stages of implementation, when the economic chapters, the parties are generally at similar levels of costs are presumed to outweigh the gains. institutional development. Such arrangements will not Cost-benefit analyses are invaluable in persuading gov- necessarily lead to the transfer of capacity-building activ- ernment officials of the long-term benefits to be had from ities, with their associated costs. If, however, a central competitive markets. Evidence is growing that the benefits authority is mandated, it may undertake internal capacity of enforcing competition provisions against cartels go building. In 2005, the European Commission’s Directorate beyond increased economic efficiency and consumer wel- General for Competition managed a program to train fare. In 2005 the EU adopted five decisions against cartels, national judges in European Commission competition law. and the fines imposed totaled 683 million euros. In 2006 That initiative, which costs 800,000 euros annually, is per- the European Commission issued seven final decisions in ceived to be an important element in the promotion of a which 41 undertakings were fined a total of 1.85 billion common competition culture in the EU. euros (European Commission 2007). This activity has resulted in substantial savings for EU consumers, since overcharges stemming from cartels are estimated to be Cost-Benefit Analyses typically about 20–30 percent of prices (Connor 2004). In agreements between parties at similar levels of institu- Heimler and Anderson (2007) note that the EU antitrust tional development, the costs of negotiating and imple- authorities’ successes with anticartel enforcement suffice menting competition provisions will be seen as justified alone to justify the investment in the relevant institutions. 354 Kamala Dawar and Peter Holmes Evenett (2004) estimates the annual deterrent effect of restricting competition or by restraining abuse of domi- anticartel laws in the EU prior to enlargement at 96 percent nant market positions across borders. Once internal tariff of enforcement outlays, in just one sector (see table 16.1). barriers are removed, firms should not be allowed to dis- To date, very few cost-benefit analyses of the impact of tort regional or member markets through cross-border competition law enforcement in cases other than anticartel anticompetitive practices. It is possible for competition enforcement (such as abuse of dominance) are available. laws to be operated in a distortionary way—for example, The economic and human resources necessary to imple- by ruling out perfectly normal vertical arrangements com- ment even a minimal decentralized competition regime are monly employed by importers or by exempting certain significant for developed and developing countries alike. practices used by local firms. Some critics of the trade and Nevertheless, the emerging evidence on the economic and competition proposals at the WTO argue, however, that the welfare costs associated with cross-border anticompetitive converse can apply; competition provisions of PTAs may practices shows that those costs are undoubtedly higher be specifically intended to enhance market access for for- than the costs of competition enforcement. Short-term eign firms at the expense of local interests. Some competi- political costs should be weighed against the understand- tion officials, too, express concern that their offices may ing that the long-term and sustainable benefits of a become antidumping agencies if they are given inappropri- strongly enforced regional competition regime will almost ate rules on predatory pricing. These fears seem to be less always outweigh its costs. problematic in practice than had been anticipated. India, for example, as far as we can tell, did not raise any formal objection to the inclusion of competition issues in the Implications for Open Regionalism EU–India negotiations. Regional arrangements, anyway, There are reasons for believing that the inclusion of com- are less subject to capture and distortion than purely petition and consumer provisions in PTAs will benefit national ones. Although one may have sympathy for the rather than hinder open regionalism—regional arrange- argument that competition rules in PTAs should not focus ments that do not discriminate against outside countries. solely on market access, that focus is not unreasonable in a This is particularly the case if the provisions are designed trade-related regime. to harmonize national laws, rather than maintain separate Again, although the empirical evidence is limited, we and differentiated national laws. A properly written com- believe the inclusion of competition provisions at the petition law is inherently neutral and nondiscriminatory, regional level could offer an opportunity to promote and so competition policy should always increase the free open regionalism by addressing the negative impact of flow of trade and investment. If a regional member cross-border trade distortions. Effectively enforced enhances its competition law as a result of a PTA provision, regional competition provisions may be able to lock in this change will have effects beyond the other regional reforms that are politically difficult to sustain because of members. Regional competition provisions monitor busi- strong domestic lobbying by interests that do not benefit ness behavior and evaluate the economic role of large for- immediately from competition law. Regional agreements eign companies on a regional basis, without contradicting can pioneer or test-run provisions and so facilitate their the principle of open regionalism. negotiation at a multilateral level at a later date. Finally, A regional competition law may strengthen regional eco- regional competition regimes can offer a demonstration nomic integration by prohibiting or controlling agreements effect of the positive gains to be had from effective Table 16.1. Cartel Overcharges and Deterrent Effect, Vitamin Industry, 1990s Overcharges on vitamin Deterrent effect of Annual average deterrent Country imports, 1990–99 anticartel laws as a share of enforcement or group (millions of U.S. dollars) (millions of U.S. dollars) outlays (percent) EC-10 660.19 1,220.78 96 Brazil 183.37 72.09 65 Mexico 151.98 44.59 46 Peru 18.91 6.98 7 Source: Evenett 2004. Note: EC-10 refers to the 10 members of the European Community (now the EU) before the 1986 enlargement: Belgium, Denmark, France, Ireland, Italy, Germany, Greece, Luxembourg, the Netherlands, and the United Kingdom. Competition Policy 355 national cooperation to underpin and improve the mar- Heimler and Anderson (2007) note this potential in their ket liberalization process. discussion of the various jurisdictions that were involved in assessing the anticompetitive practices of the Microsoft Corporation. The breakup in one jurisdiction of a large Effects on Third Parties international corporation as a result of a finding of abuse As noted above, in practice, competition laws are unlikely of dominant position might be seen as a negative develop- to discriminate against third parties. Cooperation provi- ment in another jurisdiction where behavioral remedies sions and agreements could be viewed as excluding third are viewed as adequate for handling cases of obstructive or parties, but in fact, they are unlikely to have any significant distorting competition in a market. Yet, if a firm is broken trade-diverting effects. Indeed, Brusick et al. (2004) have up by the judgment of one jurisdiction, that could, in argued that a PTA which provides for members to adopt practice, lead to its dismantling elsewhere. national competition laws and apply them in a nondis- Another potentially negative effect on third parties is criminatory manner vis-à-vis national, intra-PTA, and that if there is no law in the third-party jurisdiction pro- third-party firms will, other things being equal, have trade- hibiting cartels, the third party may unknowingly serve as a creating effects. safe haven for international cartels which collude to restrict In the case of regional competition provisions, the the market in order to protect or increase their profit mar- impact on third parties will depend to a large extent on the gins. Firms and consumers located in a jurisdiction with lax nature of the competition regime and is contingent on competition provisions will, nevertheless, benefit from the the character and existence of preexisting national compe- existence of a region with strong competition law enforce- tition laws and enforcement institutions. In many cases, as ment that is able to successfully prosecute the cartel. noted in the section on spillovers, regional competition Some PTAs have addressed the discriminatory effects of laws can have a positive impact on third parties, as meas- regional competition regimes. The Canada–Costa Rica PTA ures implemented to protect competition and consumers provides that measures taken to proscribe anticompetitive in one market will also benefit consumers elsewhere. activities are to be applied on a nondiscriminatory basis. Regional competition regimes increase the efficiency and The free trade agreement between Colombia, Mexico, and quality of markets, while lowering prices for goods and the República Bolivariana de Venezuela requires state- services. Not only is the competition regime generally owned monopolies to act on the basis of commercial con- advantageous for markets and consumers; it can also pro- siderations in operations in their own territories and not to vide information, demonstration effects, and cooperation use their monopoly positions to engage in anticompetitive to third-party agencies. practices in a nonmonopolized market in such a way as to Nevertheless, a PTA designed to confront anticompeti- affect enterprises in other member states. tive practices only insofar as they may affect trade between Another area with potential for trade distortions is the PTA members could in theory have a de jure discrimina- use of competition measures in place of antidumping meas- tory effect on non-PTA firms. At the national level, the ures in intraregional trade, in cases where the parties employ remedies and institutional provisions included in the different criteria and conditions and where antidumping regional agreement will be made available only to member measures would still apply to third parties. In fact, how- states, which may induce trade diversion against competi- ever, PTAs rarely abolish antidumping provisions. tive third-party producers. Agreements that create com- mon laws and policies with direct effect could theoretically Regional Competition Policy in Practice treat third parties differently from those agreements which only commit members to ensure national treatment in From a trade perspective, promoting competitive markets competition rules. It is difficult to see, however, how dis- helps ensure access to those markets by foreign firms. The crimination could apply in practice in these agreements, lack of a comprehensive multilateral competition agree- since third-country firms that are established in one party ment has drawn attention to regional provisions as a are usually treated like any other firm in competition law, potential tool for controlling cross-border restrictive busi- except sometimes with respect to mergers and takeovers. ness practices. Because of the principles of nondiscrimina- Different approaches to the assessment of liability and, tion and national treatment, trade agreements can include in particular, the imposition of different remedies can competition objectives even without the negotiation of a cause negative spillovers to third parties in the sense that competition chapter, but they will then not have a coher- measures adopted in one jurisdiction can affect commer- ent, independent regime for directly tackling harmful, cial decisions and the welfare of consumers in another. restrictive business practices (see box 16.1). 356 Kamala Dawar and Peter Holmes Box 16.1. Competition Policy and International Cooperation The World Trade Organization (WTO) system contains binding rules that are designed to promote a competitive environment. General national treatment and most favored nation (MFN) obligations enjoin nondiscrimination in competition policy. The provisions for trade in goods in the General Agreement on Tariffs and Trade (GATT) address the activities of state enterprises and subsidies and prohibit state-initiated voluntary export restraints. The General Agreement on Trade in Services (GATS) competition rules go further toward regulating the operation of monopoly services providers, in that they incorporate provisions for resolving disputes where competition is restricted. The GATS Telecommunications Reference Paper contains competition provisions for the telecommunications market, committing the signatories to ensure that established suppliers with dominant market power do not prohibit potential competitors from entering the market. Finally, there are provisions for members to address anticompetitive licensing practices in the Trade-Related Aspects of Intellectual Property Rights (TRIPS) Agreement. These procompetition obligations do not, however, represent a comprehensive agreement by the parties to prevent cross-border restrictive business practices and to apply the same criteria across various sectors of the economy. If governments are to tackle cross-border restrictions on competition effectively, they need to cooperate on law enforcement. Without policies to establish and develop cooperation and coordination among competition agencies, the challenges of addressing cross-border anticompetitive practices persist. Although bodies such as the Organisation for Economic Co-operation and Development (OECD) and the International Competition Network (ICN) are able to promote competition on a voluntary, intergovernmental level, their membership is largely made up of developed countries that are already engaged in these activities domestically, and they cannot offer any significant legislative push beyond this mandate. It can be argued that without effective competition laws to buttress markets, partial and discriminatory liberalization merely serves to protect emergent dominant firms or cartels in the region from external competition. Regional Competition Models centralizing the regional competition law requires creation of a complementary regional institutional mechanism to A comparative survey of competition provisions at the conduct investigations, enforce actions, and assess and levy regional level indicates great diversity. The overview in this penalties. In addition, the uniformity of court rulings needs section focuses on describing the legal obligations set out to be guaranteed, through a superior regional court, a in the provisions of each agreement (see table 16.2); it does process of binding preliminary opinions, or both. not attempt to assess the successfulness of implementation. The Treaty on the Functioning of the European Union Within each regime, associations are classified as North- (TFEU) is the leading example of a centralized regime North, South-South, and North-South. among northern or developed-country partners. The com- Centralized regimes. The most comprehensive regional petition provisions cover, among other things, agreements competition regime is the fully centralized system with sup- or concerted practices between enterprises (Article 101) porting regional institutions. A supranational law addresses and abuses of dominance by enterprises (Article 102). Arti- anticompetitive practices that affect trade between the cle 107 prohibits state aid that distorts competition. Since members or that distort competition within the region and 1989, the EU has had associated rules concerning concen- establishes a distinct regional jurisdiction. Regional and trations that may affect trade between members. Parties do domestic laws may overlap, but only the regional laws can maintain separate and distinct national competition laws, adequately address anticompetitive practices that affect trade as well as national competition authorities that may differ between treaty members or anticompetitive practices that substantially from one another, but from the creation of the take place in the regional territory beyond the jurisdictional association, regional competition law has promoted a soft boundaries of any one member of the PTA. Without harmonization of member states’ competition laws. The regional law, political entities may use the effects doctrine, European Commission’s Directorate General for Competi- as the EU and the United States have done, to respond to tion has primary competence for applying EU competition anticompetitive practices beyond their borders, but deci- laws. The European Court of Justice has reaffirmed, sions against firms located elsewhere may be unenforceable. through case law, the direct effect of European law and its Competition laws in the fully centralized model are superiority to national law.7 directly applicable within the territory of a member and The Andean Pact; the West African Economic and are superior to any national law or judgment that is incon- Monetary Union (WAEMU; in French, UEMOA, Union sistent with the regional law. Regional competition laws Économique et Monétaire Ouest-Africaine); and COMESA may also have direct effect in members’ jurisdictions, are centralized South-South competition regimes. The giving firms or citizens the right to invoke the regional law Andean Pact system is based on supranational rules in the domestic courts of the member countries. Fully enforced by community bodies. Decision 608 empowers Competition Policy 357 Table 16.2. Models of Regional Competition Regimes Model North-North South-South North-South Key characteristics Centralized EU COMESA Regional authority WAEMU/UEMOA Regional law Andean Pact Regional enforcement Partially ANZCERTA CARICOM Regional authority centralized Regional law Domestic enforcement Partially Mercosur U.S.–Brazil No regional authority decentralized EU–Jordan Regional law Domestic enforcement Decentralized SACU NAFTA No regional authority Canada–Chile No regional law Canada–Costa Rica Domestic law subject to harmonization criteria Source: Authors’ compilation. Note: ANZCERTA, Australia–New Zealand Closer Economic Relations Trade Agreement; CARICOM, Caribbean Community; COMESA, Common Market for Eastern and Southern Africa; EU, European Union; Mercosur, Southern Cone Common Market (Mercado Común del Sur); NAFTA, North American Free Trade Agreement; SACU, Southern African Customs Union; WAEMU/UEMOA, West African Economic and Monetary Union/Union Économique et Monétaire Ouest-Africaine. the General Secretariat to tackle cross-border anticompet- Partially centralized regimes. The second competition itive practices more effectively by imposing sanctions. model establishes regional competition law but supports WAEMU competition law applies to practices that have an it with an only partially centralized agency. As with the intraregional effect; countries that do not have national fully centralized regime, the independent regional law has competition laws may apply regional competition law direct applicability and takes precedence over national within their own boundaries. laws and judgments that are inconsistent with it. Although COMESA’s competition regulations and rules are the central agency has a mandate to receive complaints derived from Article 55 of its association treaty. As in the and initiate independent investigations, it must work with EU, agreements or concerted practices between enter- the member states’ competition agencies and national prises that restrict or are designed to restrict competition courts to process case actions leading to enforcement in the COMESA common market are generally prohib- and remedies. ited.8 Members may not grant, in the form of subsidies, In deep integration North-North arrangements such as state aid that restricts or threatens to distort competition the Australia–New Zealand Closer Economic Relations between member states (Article 52). Article 49 mandates Trade Agreement (ANZCERTA), the main objective is to the elimination of quantitative and other restrictions expand free trade by eliminating barriers to trade and pro- between members. The regulation concerning abuse of moting fair competition. In addition to coordinating dominant position requires the COMESA Competition bilateral trade in goods and services, ANZCERTA pro- Commission to assess vertical restraints on a rule-of- motes integration in such areas as quarantine, customs, reason basis, and horizontal agreements are illegal. The transport, standards, and business law. Antidumping regulations provide for a premerger notification system measures have been removed, and the parties’ competition under which the commission scrutinizes larger mergers, authorities and courts have concurrent or overlapping above certain turnover thresholds. The cooperation provi- jurisdiction. This enables either competition authority to sions for the COMESA Commission and for member control the misuse of market power in the trans-Tasman states specify the application and enforcement of the com- market without a need for independent supranational petition regulations and rules. Where there is concurrent institutions. Complaints relating to the abuse of dominant jurisdiction of the commission and national courts, con- market position can be filed and heard in either jurisdic- sistency in the application of competition law must be tion, and valid and enforceable subpoenas and remedial ensured. These regional regulations are directly, fully, and orders can be issued in the partner country. To underpin this uniformly effective in all member states. arrangement, the parties have signed a separate bilateral 358 Kamala Dawar and Peter Holmes enforcement agreement that provides for extensive investi- dures for cooperation or for recourse to dispute settlement. gatory assistance, the exchange of confidential informa- NAFTA’s Chapter 15, covering competition policy, monop- tion, and coordinated enforcement. olies, and state enterprises, requires members to adopt or The rules governing competition policy within maintain measures proscribing anticompetitive business CARICOM, a South-South agreement, are contained in conduct and to take appropriate action, but it does not set Chapter 8 of the Revised Treaty of Chaguaramas. That out any more specific competition rules. The provisions chapter establishes a Community Competition Commis- formalize existing consultations and cooperation between sion (CCC) with jurisdiction over all cases of cross-border the parties on the effectiveness of their national competi- anticompetitive conduct. Article 30(b) obligates members tion laws, as well as cooperation on the enforcement of those to enact competition policy legislation and establish com- laws via mutual legal assistance, notification, consultation, petition enforcement bodies. Chapter 8 requires members and exchange of information. The parties’ rights to apply to cooperate in the determination of competition legisla- antidumping or countervailing measures are preserved tion; to take the necessary legislative measures to ensure (Article 1902). consistency and compliance with the rules of competition; Another North-South arrangement, the Canada– and to set penalties for anticompetitive business conduct. Costa Rica Free Trade Agreement, includes provisions Provisions are made for cooperation between national that specify the substantive requirements of a satisfactory authorities in member states and within the CCC so as to domestic law, along with matters of due process and achieve compliance with the rules of competition. Under transparency. Although a cooperation mechanism is not Article 173(e–h), it is the responsibility of the CCC to created, there is some potential for competition authori- cooperate with national authorities, provide support, and ties to cooperate informally. Timelines are set for the facilitate exchange of information and expertise. The CCC establishment of national laws to address certain anti- is responsible for taking effective measures to ensure that competitive practices. nationals of other member states have access to competent The treaty of the Southern African Customs Union enforcement authorities, including the courts, on an equi- (SACU), a South-South agreement, commits members to table, transparent, and nondiscriminatory basis. establish competition policies and to cooperate in the Partially decentralized regimes. Further down the scale of enforcement of competition laws and regulations (Article centralization are regimes that have a regional law but no 40). Remedies relating to unfair trading practices are also independent regional body with powers of investigation provided for. Article 41 states that the SACU council shall and enforcement. Thus, the application of the law is left develop, within the context of the larger customs union, entirely to the members. National competition authorities policies and instruments to address unfair trade practices have the jurisdiction to bring cases, and they are also the between member states, on the advice of the regional recipients of complaints of any violation of the regional commission. competition law. An example is Mercosur, whose competition protocol Scope provides for a regional competition framework without any central agency. This partly decentralized arrangement In addition to representing differing degrees of centraliza- has had significant implications for the enforcement of the tion, PTA competition provisions have substantive and provisions. An intergovernmental committee assists coop- procedural requirements that vary depending on how eration and the allocation of investigations and cases comprehensive the competition regime is. A full taxonomy among the members. To ensure some harmonization of provisions, as outlined by Solano and Sennekamp among the parties, Mercosur’s competition provisions set (2006), could consist of measures and provisions on the out common principles to establish the minimum require- following issues: ments for its members’ domestic laws and procedures. Decentralized regimes. In the least centralized regional 1. Adoption, maintenance, and application of competition competition regime, members do not create a regional law; law instead, they agree to cooperation principles and criteria 2. Establishment of bodies for cooperation and coordina- for national laws addressing anticompetitive practices that tion and for enforcing competition law are detrimental to the functioning of the PTA. 3. Anticompetitive acts, and measures to be taken against NAFTA, a North-South arrangement that includes them Canada, Mexico, and the United States, does not rely on any 4. Nondiscrimination, due process, and transparency in institutions for enforcement, nor are there detailed proce- the application of competition law Competition Policy 359 5. Prohibition of the use of antidumping measures against entry into a market. Such provisions can help protect signatories’ commerce both consumers and firms, but they do not directly 6. Permitted forms of recourse to trade remedies (e.g., address demand-side market imperfections stemming antidumping measures, countervailing duties, and safe- from lack of consumer information or inability to switch guards) suppliers. Although this supply-side type of regulation 7. Application of dispute settlement procedures in compe- traditionally falls within the scope of consumer law, tition matters there are PTAs that mainstream their consumer law pro- 8. Flexibility and progressivity, or “special and differential� visions within or alongside the competition regimes to treatment. address the legislative and enforcement gaps in cross- border trade relating to consumer protection. Enhanced For the purposes of this section, it is useful to identify notification, information sharing, and investigative assis- two sets of competition provisions in PTAs: those that tance among member states can work to protect foreign envisage harmonization of the competition rules of the consumers from domestic anticompetitive business contracting parties, and those that provide for coopera- practices and to shield domestic consumers from parallel tion on competition-related issues (Holmes et al. 2005). foreign practices. The main families of provisions can be divided according Australian regional competition agreements are notable to representative parties—the EU, the United States, in their inclusion of consumer protection provisions. In and Canada. the Australia–U.S. agreement, Article 14.6 of the chapter The EU has tended to negotiate PTAs that employ lan- on competition-related issues is dedicated to cross-border guage similar to Articles 101, 102, and 106 of the TFEU, consumer protection.9 The parties, under this article, are to which implicitly promotes harmonization of dynamics further strengthen cooperation and coordination among even where no approximation is mandated (as in the their respective agencies, including the U.S. Federal Trade EU–Jordan PTA). The PTA provisions generally prohibit Commission (FTC) and the Australian Competition and anticompetitive agreements and abuse of dominant posi- Consumer Commission (ACCC), in areas of mutual con- tion that affect trade, but they do not provide commen- cern. Such areas include (a) assistance with enforcement surately robust provisions to ensure coordination and and investigations and (b) consultation and coordination cooperation among the parties. The Euro-Mediterranean on enforcement actions against violations of consumer agreements are of this type, except for the one with Algeria, protection law that have a significant cross-border dimen- which has stronger provisions. Other, more elaborate EU sion. Unlike the Australia–Papua New Guinea and Aus- agreements, such as that with Chile, include provisions tralia–Republic of Korea PTAs, Australia’s agreement with that allow trade measures such as safeguards but prohibit the United States includes separate agency-to-agency the use of anticompetitive state assistance, require nondis- agreements in the fields of competition and consumer crimination on the part of state monopolies, and mandate protection. Another difference is that the notification notification of state aid. provisions are stronger, in that notifications are to take The U.S. and Canadian PTAs with competition provi- place without the necessity of requests by the other coun- sions prohibit anticompetitive behavior. The agreements try. The Australia–U.S. PTA also contains a reciprocal can be divided into PTAs that establish commitments to agreement to provide relevant evidence in cases where create or enforce competition laws and agencies and PTAs national consumer protection laws have been violated that focus on cooperation and coordination between the (Article II.C). parties, on notifications, and on the behavior of state enter- The Australia–Korea PTA contains provisions on the prises and state monopolies. There are notable exceptions application of the parties’ competition and consumer to these broad models; for example, the Canada–Costa protection laws; the notification provisions are weaker Rica PTA is more comprehensive and more procedurally than in the PTA with the United States. The agreement demanding than other Canadian PTAs. with Papua New Guinea has an objective of promoting cooperation and coordination in the application of the countries’ competition and consumer protection laws. In Consumer Policy these cases, the parties have, at the national level, joint Most of the competition regimes in the PTAs surveyed competition and consumer protection agencies. focus on the supply-side behavior of firms and aim to An example of a South-South PTA that addresses the identify and remove barriers (such as cartels, monopo- consumer welfare aspect of regional competition frame- lies, and other restrictive business practices) to a firm’s works is the COMESA treaty. The COMESA Competition 360 Kamala Dawar and Peter Holmes Commission has powers and duties to enforce the con- Where dispute settlement mechanisms are made avail- sumer protection provisions of the competition regulations able for the competition provisions, the complaints tend and to provide support to member states in promoting and to be limited to those between states. A private party protecting consumer welfare (Articles 6 and 7). wishing to bring a complaint must first persuade a gov- ernment to submit a claim on its behalf. Governments tend to bring claims only after a cost-benefit analysis has Hard versus Soft Law been conducted, in view of the political and economic Little evidence is available for making strong recommen- resources required. Clearly, a country complainant needs dations regarding the design of appropriate dispute settle- to believe that it has some chance of winning the dispute. ment mechanisms covering competition provisions. It can Furthermore, it must have determined that winning will be noted, however, that when PTAs contain few binding not set a precedent which will not be in its longer-term competition provisions, there is less reason for a binding self-interest. dispute settlement system. There is not much to be gained In the competition provisions found in regimes such by arbitrating “best endeavor� principles that merely as COMESA, the EU, and NAFTA, this area of private encourage the application of effective domestic competi- action is seen as an effective setting for addressing, for tion laws or cooperation principles. example, exclusionary practices and abuses of dominance A halfway house toward incorporating a dispute settle- in supply chains. It is there that private complainants can ment mechanism is a PTA that allows parties to subject more easily identify the contractual practice that is disputes arising from the application of the competition affecting their commerce and bring that practice before a provisions to specific consultation procedures, short of court or authority for legal assessment and action. A sys- dispute settlement procedures. This is the case in the tem of private rights is arguably more effective at catch- Canada–Costa Rica PTA, discussed above. ing minor actions that may go undetected by competition Under a partially centralized competition regime such authorities. Private rights of action can promote the com- as CARICOM’s, an intergovernmental committee, rather petition rules and principles that aim to create a level than a specially created independent authority, implements playing field and harness the beneficial economic effects the agreement. It is not clear whether such bodies can exert of liberalization. enough influence to ensure that a member state complies Nevertheless, private rights of action are not in them- with certain treaty obligations relating to competition law. selves sufficient to effectively address cross-border competi- In the absence of a central authority, treaty objectives can tion issues relating to implementation and enforcement. easily be undermined by differences in domestic laws and Individual consumers or firms cannot obtain the necessary judgments. information to assess anticompetitive practices or do not There are further challenges to overcome where regional have the expertise to determine their full effect. This is par- competition provisions include binding commitments to ticularly the case for such restrictive business practices as set up domestic competition laws or to provide for domes- cross-border hard-core cartels and vertical restraints. Collu- tic procedures such as positive comity, which implies that sive activity is generally conducted with great secrecy. Infor- each country will consider the other’s national interest mation relating to cartel activities tends to be obtained when enforcing its own competition laws.10 Not only may through investigations, from whistle-blowers who seek for- individual members be inexperienced in implementing mal amnesty in exchange for incriminating information, or competition law, but there may also be political reasons, through other actions that require resources and powers related to preservation of sovereignty, for deciding to not available to individual actors. Without this information, exclude these obligations from the general dispute settle- it is rare for a cartel to face charges and for remedies to be ment mechanism. The parties may want to avoid having identified and imposed. In supply chains where anticom- rulings or decisions of national competition authorities and petitive behavior leads to high prices, costs can be passed to courts overturned by supranational dispute settlement pro- the end user without being identified along the way. Conse- ceedings. It may be that the negotiators were only able to quently, the likelihood that a private actor will be able to agree to language that ultimately was too vague to be sub- bring complaints about such practices before a national ject to formal legalistic dispute settlement proceedings. court is not great. Although private rights of action are There are also unwanted consequences attached to enforc- important as rights and serve as an effective monitoring and ing another party’s competition obligations, as it may lead enforcement mechanism, there are cases where they are nei- to tit-for-tat retaliation whereby a party seeks to ensure that ther sufficient nor appropriate and therefore require sup- the complaining party enforces all its obligations, as well. port through active governmental intervention. Competition Policy 361 Implementation Issues or utilize it properly, or to engage in case coordination. Where national laws and agencies already existed prior to It is difficult to assess the implementation of competition the regional agreement, there is more likelihood that the provisions. Some regional regimes commit the parties to provisions will be implemented and that benefits will be establish national competition laws and policies that, in realized. fact, were already in place prior to the agreement and Poor implementation performance based on lack of therefore cannot be attributed to the PTA—although the structural preconditions on the domestic front can to some PTA may have a reinforcing effect. If regional competition extent be addressed through the negotiation of more appro- provisions are excluded from dispute settlement provi- priate competition provisions in the PTA. These regional sions, there will not be any disputes or case law to indicate laws can prioritize the development of competition regimes nonimplementation. Finally, it is difficult to attribute evi- at the national level and, in so doing, address behavioral dence of interagency cooperation solely to the provisions issues by offering the legislative impetus and policy lock-in of a PTA, rather than to interagency contacts through, for necessary for sustained reform. Regional policy can be used example, the International Competition Network. as an exogenous force to overcome domestic inertia or Where assessment is possible, most existing research vested interests that are obstacles to implementation. A suggests that the level of implementation of regional com- regional law can compensate for an absence of national petition provisions tends to be low, particularly in develop- laws, and this has been an important feature of COMESA- ing countries. This has led to questioning of the value of type centralized arrangements. incorporating such provisions into PTAs, in view of the In a fully centralized competition regime such as burdens of negotiating them and building institutions. But COMESA, the competition laws are underpinned by insti- the research has also generated analyses that seek to iden- tutions and dispute settlement mechanisms that have been tify and solve implementation challenges in order to har- empowered to investigate, prosecute, and remedy anti- ness the regional public goods that such provisions can competitive practices. Where the agreement is between potentially provide. southern parties, some without any national competition To some extent, domestic implementation is hampered regimes, the acknowledged first challenge is to harmonize because, in most regions, competition policy is a relatively national competition laws and regional competition. This new area of regulation. Time is required to build the neces- dynamic is seriously constrained by the differing capaci- sary expertise and competition culture to establish the law ties of individual national competition authorities, which and enforce it effectively on the domestic level, before an limit the usefulness of regional law and reduce political agency is able to take advantage of the benefits of intera- support for the competition authority among members. gency cooperation. It is apparent that in some cases com- The structural and behavioral factors reinforce each other petition culture is lacking at the national level, and conse- because implementing such a fully centralized model of quently there is not sufficient political will to provide the regional competition policy requires ongoing advocacy necessary resources, and not enough institutional author- and capacity building, in addition to legislative develop- ity to push competition reform measures through. ment to ensure that the regional law is appropriate for In South-South arrangements, implementation tends to national conditions and needs. The focus on both struc- be particularly poor in countries in which national compe- tural and behavioral variables involved in implementing tition laws and authorities were underdeveloped prior to the EU competition regime evolved over the course of signature of the agreement. In the absence of the requisite 50 years, with several modernizing phases to increase effi- laws, institutions, and expertise, the country cannot absorb ciency, reduce the costs of implementation, and foster or take advantage of the benefits offered by the provisions. buy-in by member states. Indeed, if there is no national competition regime, it is Reliance on bottom-up implementation initiatives will clear that the fundamental economic benefits of prohibit- be less successful where a competition culture and but- ing restrictive business practices within the national econ- tressing regional laws and institutions are lacking. Merco- omy have not in fact been realized. Without national insti- sur is a South-South PTA that differs from COMESA in tutional structures, implementation of regional laws to having only a loose intergovernmental framework, with lit- address cross-border anticompetitive practices will be seen tle supranational power to promote competition at the to create more costs than benefits. Even where regional national level. In the absence of national competition regimes establish measures such as cooperation, notifica- regimes, there is no regional authority with the mandate to tion, consultations, and so on, they are of no use without promote and advocate implementation. The competition the human resources and expertise to request information regime under Mercosur is often described as moribund. 362 Kamala Dawar and Peter Holmes In North-South agreements, the implementation for example, is regularly ranked by the Global Competition record is clearly better, although the motivation to include Review as having one of the world’s three best competition competition law in PTAs is usually driven by the more authorities. In the EU treaty, which establishes an inde- developed party. The effectiveness of implementation is pendent law and competition authority with the resources partly a function of the ability of the northern party to and independence to implement the competition commit- push its interest in ensuring a competitive playing field for ments, there have been notable successes in tackling cartels its firms in the less developed parties, through national and other restrictive cross-border business practices. The laws and authorities. Although this transfer of competi- binding nature of the provisions, with supporting monitor- tion law does not bode well for implementation, particu- ing and enforcement agencies, is clearly much more likely to larly if the provisions negotiated are based on a model meet with success than when agreements have loose, non- exported from a developed region, there are also positive binding clauses that are not subject to dispute settlement. factors. To counterbalance the challenges that arise from Furthermore, the regime has had several decades to create transplanting a competition regime, the more developed the structural and behavioral qualities needed for success in party can offer technical assistance, capacity building, and investigating, prosecuting, and remedying regional anti- information, increasing the chances that the less devel- competitive practices. oped regional partner will achieve a positive implementa- By contrast, the Australia–New Zealand deep integra- tion record. tion PTA does not include a regional competition agency If the southern parties to a North-South agreement do and is not subject to a binding formal dispute settlement not possess the characteristics necessary to absorb the mechanism. Both countries’ courts do, however, have juris- benefits from exchange of information, the full potential diction throughout the region, and the PTA has a good of the benefits to be gained from the northern party’s implementation record. Here, as with the EU, the countries experiences and know-how cannot be realized. If, how- have the added advantage of a close geographic, historical, ever, the agreement prioritizes the provision of technical and cultural relationship. Although their national competi- assistance and capacity building, the parties will be more tion laws were initially based on different models, they likely to cooperate effectively in the long run, to the benefit were at relatively similar stages of institutional develop- of both the northern and the southern parties. It is then ment, which promoted harmonization. In situations such that positive results emerge from the avoidance of duplica- as this, and where competition agencies already exist, the tion of research and other activities, and of conflicts in main focus of the provisions is on enhancing cooperation judgments and legal interpretations. When the advantages and coherence among the parties on the basis of comity of regional competition provisions are tangible and signifi- rules—positive, negative, or both. A well-designed regime cant, as manifested in lower prices, better quality, and more of cooperation mechanisms can facilitate the implementa- efficient regional markets, the economic and political costs tion of measures to tackle hard-core cartels, just as legisla- of implementation are more likely to be perceived as justi- tive shortcomings inhibit cooperation (Alvarez and Wilse- fied in light of the economic and consumer welfare gains. Samson 2007). The Brazil–U.S., NAFTA, Canada–Costa Rica, EU–Jordan, It is evident and significant that the implementation of and EU–South Africa competition regimes have had some competition provisions at a regional level depends on ade- success in advocating a competition culture and in pro- quate enforcement of competition polices at the national moting cooperation between competition authorities level (Alvarez and Wilse-Samson 2007; Alvarez and Horna through learning by doing.11 In these decentralized North- 2008). Enforcement will be improved if, in addition to South regional competition regimes, the commitments are focusing on the necessary structural changes, behavioral confined to national implementation and cooperation but- changes are effected. Where competition is a new phenom- tressed by technical assistance. Although there is no explicit enon, the inclusion of competition provisions in regional reference to special and differential treatment for the agreements will merely signal the importance of regulating southern party, flexibility in implementation can be built the restrictive practices of businesses, rather than provide into the competition provisions. For example, the substantive means of regulating anticompetitive practices. EU–Jordan treaty states that Jordan is to have five years to That is, regional competition provisions will be a first step implement the provisions regulating unfair competition, in the long process of competition advocacy aimed at cre- and further flexibility is offered with respect to the removal ating a competition culture. of competition-distorting state aid (Article 53.3). To increase the advocacy of competition, PTAs should In North-North PTAs with fully centralized competi- emphasize cooperation on competition provisions. Some tion regimes, the implementation record is better. The EU, areas for cooperation are case investigations, legal treaties, Competition Policy 363 exchange of staff, exchange of experience, and peer reviews, culture to demand them and to allow the competition in addition to compliance with regional competition rules. authority to implement its decisions without undue inter- Strengthening of implementation capacity in developing ference from lobbying groups. PTA members must be accompanied by a reinforced com- mitment on the part of developed countries to effectively Conclusions address the main competition policy concerns of their trading partners. This is particularly important where Despite limited evidence about the impact of the regional regional competition provisions were negotiated as a small competition regimes, it is clear that lack of regional com- part of a much broader trade agreement and were pushed petition can undermine the benefits from liberalization by more developed parties with an interest in opening up of regional markets. Implementation of effective regional developing-country markets to foreign trade and business competition law and policy can help address cross-border activity. restrictions or regional market failures and generate posi- Implementation will be more successful where the pro- tive spillovers, such as more efficient markets that offer visions reflect the diversity of social policy objectives within better-quality goods for lower prices, further encouraging the regional framework. A PTA can explicitly promote the investment. Additional benefits may be derived by main- existing domestic policy priorities of the parties, such as streaming consumer policy into regional competition support for small and medium-size enterprises or for disad- laws, given that the option of cooperation in dealing with vantaged groups. The domestic policy space for fostering cross-border consumer protection issues complements marginalized communities, consumer welfare, or small both economic and social development objectives. businesses through exclusions from competition regula- Although the number and variety of North-South and tions, as well as for promoting market efficiency and market South-South PTAs with some element of competition law access, will make the agreement more appropriate to local is increasing, implementation records have been poor. needs and increase the chances of successful implementa- This is particularly the case for PTA members that did not tion. If the regional competition provisions are viewed in possess a national competition law prior to accepting the light of the existing trade policies operating in the par- regional commitments and for PTAs whose members are ties to the PTA, they are more likely to be implemented. at very different stages of competition regime develop- Levels of implementation will also relate to the ability ment. A regional competition regime in which effective of the parties to enforce the competition provisions national laws are lacking offers no legal basis for a member effectively. Because most of the provisions are excluded to take action against anticompetitive practices organized from the dispute settlement mechanism covering the in another member state on the basis of the effects on its general trade provisions, only consultation mechanisms own territory. Even where regional regimes establish are available for discussing issues related to implementa- measures such as cooperation, notification, and consulta- tion. These, however, are unlikely to provide the same tions to generate regional benefits, these are of no use incentives to implement as would binding provisions nationally without the human resources and expertise to covered by binding dispute settlement mechanisms that absorb or respond to the information. provide legal certainty and remedies. Implementation is more successful in North-South Although competition provisions are generally not agreements, partly because of the ability of the northern subject to dispute settlement processes and procedures, party to push its interests in ensuring that the less devel- enforcement mechanisms can nevertheless be incorpo- oped parties provide a competitive playing field for its rated within the provisions themselves. For example, the firms through national law and authority. The more devel- inclusion of private rights of action will enable individu- oped party will be able to offer technical assistance, capac- als harmed by anticompetitive practices to seek redress ity building, and information that can increase the less and damages in local courts. developed regional partner’s potential to achieve a positive Finally, implementation will be facilitated if regional implementation record. authorities are endowed with strong investigative powers, By crafting appropriate regional competition provisions, adequate resources and expertise, and the ability to issue PTAs can serve as vehicles for addressing both the structural cease-and-desist orders and collect fines. The greater the and the behavioral challenges obstructing successful imple- independent powers, budgets, and resources of the compe- mentation. In countries with little or no experience with tition agencies, the greater the chances of satisfactory competition policy, regional laws can act, even if temporar- implementation will be. Such powers will only be bestowed ily, as an alternative to the expense of establishing and on competition agencies where there exists a competition implementing domestic competition laws. For members 364 Kamala Dawar and Peter Holmes with nascent or nonexistent competition regimes, technical 3. According to the effects doctrine, domestic competition laws are applicable to foreign firms, but also to domestic firms located outside assistance should be designed to impart the requisite the state’s territory, when their behavior or transactions produce an expertise and experience over the long term. This will also “effect� within the domestic territory. The “nationality� of firms is irrel- assist in promoting the necessary behavioral reforms. evant for the purposes of antitrust enforcement, and the effects doctrine Although PTA competition provisions can offer the leg- covers all firms irrespective of their nationality (Institute of Competition Law database, http://www.concurrences.com/article.php3?id_article= islative impetus and the policy lock-in necessary for sus- 12374&lang=en). tained reform, a more appropriate first objective may be to 4. Some authors believe all competition and antitrust policy is cap- focus on establishing at the national or subregional level a tured by “losers�—that is, inefficient firms that lose market share when faced with competition from stronger firms (see McChesney and Shughart culture that values competition. In PTAs involving members 1995)—but there is no convincing evidence for this view. with no or little experience with competition, the provisions 5. A billion is a thousand million. could initially be restricted to information exchange, techni- 6. For a description of Indian interest in this area, see Shroff (2005). 7. For example, Internationale Handelsgesellschaft mbH v. Einfuhr- cal assistance, and capacity building. More general commit- undVorrattsstelle für Getreide und Futtermittel, ECJ case 11/70, 1970 ECR ments could be implemented after the necessary expertise 1125. and cooperation mechanisms have been developed. Provi- 8. Exceptions to this rule must meet four requirements. The two pos- sions could also be included that oblige members, over a itive requirements are that the agreement, decision, or concerted practice contribute to the improvement of the production or distribution of goods specified period of time, to adopt competition laws that can or the promotion of technical or economic progress and that it allow con- address the full range of private and state-created anticom- sumers a fair share of the resulting benefit. The two negative requirements petitive practices and outcomes. In such a scenario, soft law are that the agreement not impose restrictions unnecessary to the attain- ment of the positive objectives stated and that it not afford the firms con- may offer advantages if governments are uncertain of the cerned the possibility of eliminating competition in a substantial part of underlying technical issues or the judicial consequences of the market in question. the provisions. If competition regimes are at an early stage of 9. The text is available at Office of the U.S. Trade Representative, “Free Trade Agreements Australia,� http://www.ustr.gov/trade-agreements/free- implementation, nonjudicial mechanisms such as voluntary trade-agreements/australian-fta. peer review, consultations on implementation issues, and 10. For example, the European Commission’s report on the applica- informal diplomatic methods may be more appropriate. tion of the EU–U.S. agreement states, These can subsequently be complemented by a nonbinding In all cases of mutual interest it has become the norm to establish mechanism for review of competition policy and by non- contacts at the outset in order to exchange views and, when appro- binding consultations, promotion of voluntary implementa- priate, to coordinate enforcement activities. The two sides, where appropriate, seek to coordinate their respective approaches on the tion of competition policy obligations, and, ultimately, con- definition of relevant markets, on possible remedies in order to vergence or harmonization. ensure that they do not conflict, as well as on points of foreign law If highly centralized regimes are to function adequately, relevant to the interpretation of an agreement or to the effective- ness of a remedy. Cooperation under this heading has involved the the regional monitoring and enforcement body must be synchronization of investigations and searches. This is designed to endowed with strong investigative powers, adequate make fact-finding action more effective and helps prevent compa- resources and expertise, and the ability to issue cease-and- nies suspected of cartel activity from destroying evidence located in the territory of the agency investigating the same conduct after its desist orders and collect fines. When the advantages of counterpart on the other side of the Atlantic has acted. (European regional competition provisions are tangible and signifi- Commission 1999, 313) cant, the resource requirements and political costs of 11. Brazil and the United States have a competition cooperation implementation will be perceived as being justified by the agreement (rather than an free trade agreement). economic and consumer welfare gains alone. Identification of these gains will require further empirical research and cost-benefit analyses, along with the necessary advocacy Selected Bibliography and legislative revisions at the national and international Alvarez, Ana María, and Pierre Horna. 2008. “Implementing Competition levels to build an economic and social environment that is Law and Policy in Latin America: The Role of Technical Assistance.� Chicago-Kent Law Review 83 (1): 91–128. http://www.cklawreview.com/ better able both to demand competitive markets and to wp-content/uploads/vol83no1/Alvarez.pdf. supply them. Alvarez, Ana María, and Laurence Wilse-Samson, eds. 2007. “Implement- ing Competition-Related Provisions in Regional Trade Agreements: Is It Possible to Obtain Development Gains?� UNCTAD/DITC/CLP/ Notes 2006/4, United Nations Conference on Trade and Development (UNCTAD), New York and Geneva. http://www.unctad.org/en/docs/ 1. Some provisions of U.S. law may allow litigation in the United ditcclp20064_en.pdf. States. These were unsuccessfully invoked in the Empagran case concern- Anderson, Robert, and Frédéric Jenny. 2005. “Competition Policy, Eco- ing alleged price fixing for vitamins; see Popofsky (2005). nomic Development and the Role of a Possible Multilateral Frame- 2. This prohibition can be reversed in specific cases, for example, work on Competition Policy: Insights from the WTO Working Group when authorized by the U.S. International Antitrust Enforcement Assis- on Trade and Competition Policy.� In Competition Policy in East Asia, tance Act of 1994. ed. Erlinda Medalla. Abingdon, U.K.: Routledge. Competition Policy 365 Baldwin, Richard, Simon Evenett, and Patrick Low. 2009. “Beyond Tariffs: and Consumer Policy for Latin American Countries,� United Multilateralising Deeper PTA Commitments.� In Multilateralising Nations Conference on Trade and Development (UNCTAD), Lima, Regionalism: Challenges for the Global Trading System, ed. Richard March 22–24. http://www .evenett.com/publicpolicy/presentations Baldwin and Patrick Low. Cambridge, U.K.: Cambridge University .html. Press. Evenett, Simon J., Margaret C. Levenstein, and Valerie Y. Suslow. 2001. Belderbos, Rene, and Peter Holmes. 1995. “The Economics of Matsushita “International Cartel Enforcement: Lessons from the 1990s.� World Revisited.� Antitrust Bulletin 20 (Winter): 825–57. Economy 24 (9, September): 1221–45. Also issued as Policy Research Bhagwati, Jagdish N., and Arvind Panagariya. 1996. The Economics of Pref- Working Paper 2680, World Bank, Washington, DC, 2002. erential Trade Agreements. Washington, DC: American Enterprise Heimler, Alberto, and Robert D. Anderson. 2007. “What Has Competition Institute Press. Done for Europe? An Inter-Disciplinary Answer.� Aussenwirtschaft 4. Botta, Marco. 2009. “The Cooperation between the Competition Authori- SSRN: http://ssrn.com/ abstract=1081563. ties of the Developing Countries: Why Does It Not Work? Case Study Hoekman, Bernard. 1999. “Free Trade and Deep Integration: Antidumping on Argentina and Brazil.� Competition Law Review 5 (2): 153–78. and Antitrust in Regional Agreements.� Policy Research Working Paper Bourgeois, Jacques H. J., and Patrick A. Messerlin. 1999. “The European 1950, World Bank, Washington, DC. http://www-wds.worldbank.org/ Community’s Experience.� In Brookings Trade Forum: 1998 Trade, external/default/WDSContentServer/WDSP/IB/1998/07/01/000009265 Competitiveness, Global Economics, Antidumping, ed. Robert Z. _3980928162543/Rendered/PDF/multi0page.pdf. Lawrence. Washington, DC: Brookings Institution Press. Hoekman, Bernard, and Peter Holmes. 1999. “Competition Policy, Devel- Bourgeois, Jacques, Kamala Dawar, and Simon J. Evenett. 2007. “A Com- oping Countries and the WTO.� World Economy 22 (August): 875–93. parative Analysis of Selected Provisions in Free Trade Agreements.� Hoekman, Bernard, and Kamal Saggi. 2004. “Trading Market Access for Prepared for the Directorate General for Trade, European Commis- Competition Policy Enforcement.� Policy Research Working Paper sion, Brussels. Series 3188, World Bank, Washington, DC. http://www-wds.worldbank Brusick, Philippe, Ana María Alvarez, and Lucian Cernat, eds. 2005. Com- .org/external/default/WDSContentServer/WDSP/IB/2004/02/10/00001 petition Provisions in Regional Trade Agreements: How to Assure Devel- 2009_20040210145554/Rendered/PDF/325550wps3188.pdf. opment Gains. Geneva: United Nations Conference on Trade and Holmes, Peter, Anna Sydorak, Anestis Papadopoulos, and Bahri Özgür Development (UNCTAD). Kayali. 2005. “Trade and Competition in RTAs: A Missed Opportunity?� Brusick, Philippe, Lucian Cernat, Ana María Alvarez, and Peter Holmes, In Competition Provisions in Regional Trade Agreements: How to Assure eds. 2004. “Competition, Competitiveness and Development: Lessons Development Gains, ed. Philippe Brusick, Ana María Alvarez, and Lucian from Developing Countries.� United Nations Conference on Trade Cernat. Geneva: United Nations Conference on Trade and Development and Development (UNCTAD), Geneva. http://www.unctad.org/en/ (UNCTAD). www.unctad.org/en/docs/ditcclp20051_en.pdf. docs/ditcclp20051_en.pdf. Hughes, Timothy T. 2004. “Designing Competition Institutions for Differ- Connor, John M. 2004. “Price-Fixing Overcharges: Legal and Economic ent Stages of Institution Building.� Presented at the Asia-Pacific Eco- Evidence.� American Antitrust Institute Working Paper 04-05, Ameri- nomic Cooperation (APEC) Fourth Training Program on Competi- can Antitrust Institute, Washington, DC. SSRN: http://papers.ssrn tion Policy, Ho Chi Minh City, August 3–6. http://www.jftc.go.jp/ .com/sol3/papers.cfm? abstract_id=1103516##. eacpf/05/APECTrainingProgramAugust2004/usa.hughes.pdf CUTS (Consumer Unity and Trust Society). 2003. Pulling Up Our Socks. Kulaksizoglu, Tamer. 2004. “Measuring the Effectiveness of Competition: Report based on the 7-Up Project. Jaipur, India: Consumer Unity and Evidence from the Turkish Cement Industry.� Munich Personal Trust Society. RePEc Archive (MPRA) Paper 357, Munich University Library, Dawar, Kamala, and Simon J. Evenett. 2008. “The CARIFORUM-EC EPA: Munich, Germany. http://mpra.ub.uni-muenchen.de/357/1/MPRA An Analysis of Its Competition and Government Procurement Provi- _paper_357.pdf. sions.� Working Paper, German Agency for Technical Cooperation Levenstein, Margaret, and Valerie Suslow. 2001. “Private International (GTZ), Eschborn, Germany. http://www.gtz.de/en/dokumente/en-epa- Cartels and Their Effect on Developing Countries.� Background paper cariforum-and-beyond-procurement-and-competition-2008.pdf. for World Development Report 2001, World Bank, Washington, DC. de Araújo, José Tavares, Jr. 2001. “Toward a Common Competition Policy http://www-unix.oit.umass.edu/~maggiel /WDR2001.pdf. in Mercosur.� Presented at the conference on “Competition Policy in McChesney, Fred S., and William F. Shughart II, eds. 1995. The Causes and Infrastructure Services: Second Generation Issues in the Reform of Consequences of Antitrust: The Public-Choice Perspective. Chicago, IL: Public Services,� Inter-American Development Bank, Washington, DC, University of Chicago Press. April 23–24. Organization of American States (OAS), Washington, DC. Phlips, Louis. 1998. “On the Detection of Collusion and Predation.� In http://www.sedi.oas.org/DTTC/TRADE/PUB/STAFF_ARTICLE/tav01 Applied Industrial Economics, ed. Louis Phlips, 269–83. Cambridge, _merc_e.asp. U.K: Cambridge University Press. Dutz, Mark, and Aydin Hayri. 1999. “Does More Intense Competition Popofsky, Mark S. 2005. “Extraterritoriality and the Empagran Case.� Lead to Higher Growth?� Policy Research Working Paper 2320, World ABAnet, American Bar Association, Washington, DC. http://www Bank, Washington, DC. http://www-wds.worldbank.org/external/ .abanet.org/antitrust/at-committees/at-ic/pdf/spring/05/popofsky.pdf. default/WDSContentServer/IW3P/IB/2000/05/25/000094946_000504 Qaqaya, Hassan, and George Lipimile, eds. 2008. “The Effects of Anti- 05325137/Rendered/PDF/multi_page.pdf. Competitive Business Practices on Developing Countries and Their Dutz, Mark A., and Maria Vagliasindi. 2002. “Competition Policy Imple- Development Prospects.� United Nations Conference on Trade and mentation in Transition Economies: An Empirical Assessment.� Development (UNCTAD), New York and Geneva. http://www.unctad OECD CCNM/GF/COMP/WD(2002)13, Organisation for Economic .org/en/docs/ditcclp20082_en.pdf. Co-operation and Development (OECD), Paris. Rosenberg, Barbara, and Mariana Tavares de Araújo. 2005. In Competition European Commission. 1999. XXVIII Report on Competition Policy 1998. Provisions in Regional Trade Agreements: How to Assure Development Luxembourg: European Communities. http://ec.europa.eu/competition/ Gains, ed. Philippe Brusick, Ana María Alvarez, and Lucian Cernat, publications/annual_report/1998/en.pdf. 189–212. Geneva: United Nations Conference on Trade and Develop- ———. 2007. Report on Competition Policy 2006. Luxembourg: Office for ment (UNCTAD). http://www.unctad.org/en/docs/ditcclp20051_en Official Publications of the European Communities. http://ec.europa .pdf. .eu/competition/publications/annual_report/2006/en.pdf. Shroff, Natashaa. 2005. “Bilateral Antitrust Cooperation Agreements.� Evenett, Simon J. 2004. “Cost-Benefit Analyses of Implementing Research Report, Competition Commission of India, New Delhi. Competition Law and Policy.� Presented at a seminar on “Strength- http://www.cci.gov.in/images/media/ResearchReports/BilateralAntitrust ening Institutional and Capacity Building in the Area of Competition Natashashroff09022007_20080411101524.pdf. 366 Kamala Dawar and Peter Holmes Solano, Oliver, and Andreas Sennekamp. 2006. “Competition Provisions UNCTAD (United Nations Conference on Trade and Development). in Regional Trade Agreements.� OECD Trade Policy Working Paper 2000. “The Set of Multilaterally Agreed Equitable Principles and Rules 31, Organisation for Economic Co-operation and Development for the Control of Restrictive Business Practices.� United Nations, New (OECD), Paris. doi:10.1787/344843480185. http://www.oecd.org/ York. http://www.unctad.org/en/docs/tdrbpconf10r2.en.pdf. officialdocuments/displaydocumentpdf/?cote=com/daf/td%282005% Utton, Michael A. 2008. International Competition Policy. Cheltenham, 293/final&doclanguage=en. U.K.: Edward Elgar. Sykes, A. O. 1999. “Antidumping and Antitrust: What Problems Does Each Vedder, Christoph. 1990. “A Survey of Principal Decisions of the Euro- Address?� In Brookings Trade Forum: 1998 Trade, Competitiveness, pean Court of Justice Pertaining to International Law: I. Wood Pulp.� Global Economics, Antidumping, ed. Robert Z. Lawrence. Washington, European Journal of European Law 1 (1–2): 365–77. http://207.57 DC: Brookings Institution Press. .19.226/journal/Vol1/No1/sr2.html. 17 GOVERNMENT PROCUREMENT Kamala Dawar and Simon J. Evenett Theoretical and empirical knowledge on issues related to Economic and Developmental Dimensions government procurement in preferential trade agreements of Government Procurement (PTAs) is sparse. It consists mainly of taxonomies of legal At the outset, it is critical to understand the development provisions on government procurement and economic context in which discussions of government procurement models of across-the-board (i.e., nondiscriminatory or policy in trade agreements take place. Proposals for public nonpreferential) reductions in discrimination against for- procurement provisions should be informed by circum- eign bidders for state contracts. To the best of our knowl- stances in developing countries. As with many “trade edge, no ex post empirical assessments have been conducted and . . . � matters, it would be unwise to design or assess on the impact of PTA provisions dealing with government proposals for trade disciplines related to public procure- procurement on trade flows or on the effectiveness of ment without a clear understanding of existing state national procurement institutions, nor are there evalua- purchasing practices and their potential developmental tions of the extent to which these PTA provisions have been significance. In particular, it is important to examine the implemented or of whether national procurement authori- potential scope of government procurement, the relevant ties have changed their practices as a result of PTA provi- factors affecting national procurement regimes, the objec- sions. The analysis of government procurement in PTAs is a tives of these regimes, and the underlying principles of nascent discipline, and readers are cautioned accordingly. good procurement policy before drawing inferences about The purpose of this chapter is to describe what is known trade negotiations and obligations bearing on public concerning public procurement provisions in PTAs and procurement. what factors ought to be taken into account by policy mak- ers and analysts as they evaluate policy options in this area. Procurement Spending as a Share of Total Although an effort is made to draw appropriate policy Government Spending implications, the findings are largely tentative and will need to be revised in the light of new evidence and changes A distinction must be made between all government in thinking as to what constitutes effective public procure- spending and what is typically referred to as state spending ment policy. on goods and services. (Spending on wages, salaries, and The section that follows examines the developmental pensions is part of the former, but not the latter.) The sig- aspects of government procurement and associated reforms nificance of this difference for development is that wage in the context of trade agreements. The second section sur- rates are much lower in developing countries than else- veys the major types of PTA provisions on government pro- where, and so the cost-effective way of supplying a given curement found in selected agreements, with particular level of public service is to use more labor-intensive meth- attention to dispute settlement provisions. The third section ods. Consequently, the share of spending on capital and then assesses the government procurement provisions intermediate goods will, other things being equal, be lower found in PTAs in a multilateral context, focusing on the in developing countries, and this—setting aside the level of important matter of discrimination. Some lessons for pol- national income—accounts for the relatively small size of icy making are set forth in the concluding section. these countries’ public procurement expenditures.1 367 368 Kamala Dawar and Simon J. Evenett However small public procurement expenditures are in is placed on transparency and good governance. A num- relation to the size of the economy, the manner in which ber of jurisdictions recognize that the management of such monies are spent has an important developmental procurement systems is a distinct, highly valued profes- effect. Many public goods and services have a direct or sion, and that, as in many cases involving talented person- indirect effect on economic performance and living nel, staff retention and motivation are important standards—in particular, the living standards of the poor, challenges.2 who tend not to be able to afford private alternatives. Effec- tive national procurement policies can help improve the exe- Objectives of National Procurement Regimes cution of state infrastructure projects, yielding export and growth benefits. A government that is pursuing recognized A common feature of public procurement policies that development goals should, consequently, strive to limit almost always colors debates about reform in both devel- waste and corruption in its public procurement regime. oping and industrial countries is the multiplicity of objec- tives assigned to these policies. A review of national public procurement legislation and implementing regulations Factors Affecting National Procurement Regimes shows that the following objectives are commonly targets in Developing Countries of public procurement policy: Procurement regimes do not necessarily affect all levels of government in the same way. Total state spending on goods • Value for money, typically taken to mean minimizing and services is distributed across various levels of govern- procurement costs ment within a nation. Constitutional arrangements—in • Macroeconomic management particular, federalist structures—affect which levels of gov- • National security ernment do the spending, how much they spend, and for • Redistribution to the poor what purposes. It is unwarranted to assume that just • Industrial and regional development because one level—say, the central government—engages • Promotion of small and medium-size enterprises in public procurement reform (perhaps unilaterally, per- (SMEs) haps in the context of a trade agreement), other levels will • Support for state-owned enterprises and their employees follow suit. Constitutional niceties matter when evaluating • Pursuit of governance-related targets. the likely significance and impact of public procurement reform. In practice, pursuit of any but the first target amounts to Government procurement in developing countries could designing procurement systems that sacrifice the value-for- also conflict with other—at times, externally imposed— money goal, in whole or in part, for some other objective. constraints on the composition and implementation of Advocates for giving preference to a particular regime typ- public spending. Externally funded budget support pro- ically appeal to some apparently inviolable principle such grams, tied aid from donors, and debt relief initiatives all as transparency or defense of small business, but the risk of affect the level of government spending on goods and serv- wasting scarce state resources is high if governments are ices and the extent to which the associated contracts are swayed by abstract principles. open to national and international competition. A distinct The alternative for governments is to use other state implication of these schemes is that their existence may instruments, when available, to attain a particular target.3 limit the scope of national or regional public procurement It is, to be sure, possible to use a national procurement reform initiatives, unless particular care is taken to recon- regime to support local industrial development, and a cile the latter initiatives with bilateral or other interna- government policy of imposing high costs on foreign tional obligations. enterprises attempting to establish and do business in the In addition, many developing-country governments expe- country will indeed stimulate incumbent firms. But if rience institutional and administrative capacity constraints, those existing firms are particularly confident, they may including a paucity of trained procurement staff. As the simply raise the prices charged the state buyer. In such a discussion below illustrates, there are many methods of case, not only is the policy misguided, but it might exac- public procurement, and the state officials responsible erbate the exercise of market power and the distortion of need to be well versed in their design and execution. Per- resources within the economy. Much is at stake in the formance evaluation of contractees is required, and con- design of public procurement policy—not just the tract management, reporting, and accountability are capacity to do some good, but also the danger of doing important tasks, especially in an era in which a premium further harm! Government Procurement 369 When pursuing an objective other than value for national procurement regimes can be substantial. For an money, policy makers and analysts need to ask under what average developing country that spends about 15 percent circumstances government demand would directly and of its national income on goods and services, a 10 percent least expensively meet the targeted objective. To this line of saving on procurement contracts is equivalent to 1.5 per- argument, some will respond that the “real� world is cent of gross domestic product (GDP)—an amount that imperfect and often requires second- or third-best solu- may exceed the total amount of aid received by many tions. Experience indicates that this objection would be developing countries. stronger if it were made after alternative government policy Equality of opportunity. Entrenching equality of oppor- instruments had been evaluated and rejected as potential tunity to compete for state contracts involves eschewing means to a stated end. provisions that limit, bar, or discourage firms from bid- ding, on the basis of location, sourcing decisions, and employment practices.5 Bans on foreign bidders, as seen in Four Broad Principles of Good Procurement Policy certain “buy-national� legislation passed during the Despite, or because of, the variety of government objectives 2008–09 global economic downturn, involve violations of for public procurement policy, most jurisdictions, interna- equality of opportunity (see box 17.1). The matter here is tional accords, and pronouncements of international not simply a case of domestic versus foreign firms but also organizations on public procurement tend to refer to a core of discrimination between foreign firms, as discussed in set of “principles� for the implementation of national pol- the next section. icy in this area. Governments desire to retain the freedom The adverse welfare effect from discriminating against to use procurement policy to pursue policy objectives that foreign bidders is, however, not straightforward to estab- may be different from their neighbors’ objectives. In this lish theoretically. Baldwin (1970) and Baldwin and context, the core principles may be seen either as a limited Richardson (1972) show that when the quantity of a good approach to liberalization or as an agreement on higher- that the government seeks to buy is smaller than the total level disciplines that guarantee good policy making and quantity supplied by domestic firms, prohibiting foreign governance. For governments negotiating a common firms from bidding on state contracts merely reshuffles approach to procurement reform—perhaps through a PTA—these similarities in principles may make it easier to reach consensus despite differences in overall objectives for public procurement policies. Box 17.1. Persistence of Discrimination: Procurement Over time, the following four principles appear to have Practices and the Global Economic Crisis gained considerable common support: (a) efficiency (value for money); (b) equality of opportunity to compete for The 2008–09 global economic downturn has created doubts about the effectiveness of the rules and disciplines governing state contracts (nondiscrimination); (c) transparency government procurement contained in trade agreements. (control of corruption; accountability); and (d) encour- The widespread use of fiscal stimulus packages has added a further layer of factors and potential complexity. Some agement of investments and partnerships (public-private governments have included “buy-national� provisions in partnerships).4 The principles may be codified in national fiscal stimulus packages to coerce state agencies into buying constitutions, national and subnational laws, implement- “domestic products.� Defining what exactly a domestic product is often proves elusive, and so the laws underpinned ing regulations, and binding and nonbinding international by such notions can be confusing. accords. Associated with the principles are particular steps This being said, buy-national provisions have the that the government can take to attain them. potential to affect the international outsourcing decisions of firms and the operation of their supply chains. A Efficiency. It is widely accepted that the value-for-money developing country may find that both its intermediate and objective is best achieved by encouraging (through the final goods producers lose sales abroad when a trading design of an easy-to-understand, easy-to-participate-in, partner implements restrictive buy-national policies. These policies introduce a form of cross-border discrimination and fair procurement regime) the maximum number of against foreign commercial interests in an area of corporate bidders for a state contract. Simulation evidence strongly strategy making (international supply chains) that has benefited significantly from open borders over the past two suggests that the expected cost to the government of a con- decades. Moreover, for developing countries in which tract falls as the number of bidders increases, and especially participation in international supply chains is viewed as a as the number rises toward five or six (McAfee and way of encouraging the upgrading of exporters, the consequences of being barred from certain commercial McMillan 1989; Deltas and Evenett 1997). opportunities through buy-national provisions may not be For many developing countries, the inefficiency and confined to lost sales. opportunity cost of suboptimal levels of competition in 370 Kamala Dawar and Simon J. Evenett purchases from foreign producers from state buyers to made as nondiscriminatory as possible. For instance, if a local private sector buyers, without any impact on local particular skill is absolutely needed, the qualification prices and local production levels. In other words, the exis- requirement should be based on that skill—on nothing tence of discrimination and its subsequent removal may else, and certainly not on how that skill was acquired. In have no effect on resource allocation. many instances, however, governments are adamant in The same analyses showed that only when the total asserting that only the graduates of specified national insti- amount demanded by a government exceeds the total tutions have the skill in question. quantity supplied by domestic firms does banning foreign Transparency. The importance of transparency in gov- bidders increase domestic output and prices and limit ernment procurement is generally well accepted. It is often imports. If the good in question is one that is supplied in argued (Anderson et al. 2009) that transparency helps small quantities in a developing country—perhaps because improve governance and limit corruption and discrimina- the legal and governance environment is less than ideal for tion; the latter consideration points to a potential comple- business—such a ban on foreign procurement can indeed mentarity across principles. However, not every aspect of lead to expansion of domestic output.6 For this reason, the procurement process can be made mechanical and nationalistic procurement policies are regarded in some transparent (the evaluation of intangible attributes of bid- quarters as part of the industrial policy toolkit. ders is an example), and the pursuit of more transparent In brief, the economics of discrimination in public pro- procurement policies will not completely eliminate oppor- curement is different from that of tariffs, precisely because tunities to engage in discrimination. Furthermore, achiev- the former applies only to a subset of buyers. ing transparency is costly. Although transparency can More recent analyses focus on cases in which discrimi- encourage more firms to bid for state contracts, thereby nation did limit market access (see, e.g., Evenett and intensifying competition and lowering procurement costs, Hoekman 2005). The increase in prices paid by state buyers it also entails costs, such as delays in awarding procedures. following a ban on foreign procurement tends to encour- The optimal degree of transparency is therefore unlikely to age the entry of domestic firms willing to supply the gov- be infinite, and reasonable people can disagree over that ernment, and so the longer-term effects of procurement degree. Still, the general principle that the procurement discrimination depend on the magnitude of local barriers process should be known, understandable, and inexpensive to entry. With no such barriers (whether administrative or to monitor remains key. in the shape of anticompetitive practices by incumbent The relationship between transparency and market firms), the procurement discrimination could, in the long access can be ambiguous (Evenett and Hoekman 2005). term, lead to an expansion of domestic output. In a com- Making procurement regulations easier to understand and petitive market, moreover, prices would fall in the longer more accessible will encourage foreign bidders for state term to the lowest level of average costs of the most effi- contracts but will also attract domestic ones. Whether the cient local firm, which may or may not be equal to those of share of state contracts awarded to foreign firms goes up or the most efficient foreign rival. If it turns out that in the down will depend on the relative responsiveness of both longer run the most efficient local firm has costs equal to types of firm to improvements in transparency. It is quite or less than those of its most competitive foreign rival, the possible that a foreign trading partner could argue for the government will end up paying prices at or below world inclusion of transparency-related provisions in a PTA and prices; implying that under these circumstances there is no subsequently discover that the implementation of those adverse price impact from discrimination in the long run. provisions actually benefits the domestic contractors of the The policy implication of this argument is as follows: the partner country. longer-term impact of procurement discrimination on Improved transparency is one of the few areas in which resource allocation and state budgets is contingent on there is some empirical evidence of the impact of procure- national competition law and its enforcement and on poli- ment reform.7 Information on contracts entered into by cies toward the entry of new business. member states of the European Union (EU) between 1995 In some cases, such as the provision of health care and and 2002 and on the number and “nationality� of firms other professional services, the principle of equality of bidding for those contracts shows that during the period, opportunity is tempered by the realization that it should the average number of bidders increased by 30 percent, the only apply to qualified or sufficiently expert or experienced number of foreign subsidiary bidders rose to 30 percent of bidders. Without challenging the contention that expertise the total, and the dispersion of prices paid for comparable is needed to fulfill certain government contracts, the ques- products by state buyers fell by 30 percent. Interestingly, it tion arises as to whether the qualifications to bid can be was found that during the same period, 78 percent of all Government Procurement 371 state contracts examined went to small and medium-size question arises as to why provisions in a trade agreement enterprises, suggesting that transparency reform has present a more credible, more effective, or more feasible not eliminated the capacity of SMEs to compete for these option. Much depends on the legislative and constitutional contracts. history of the developing country in question—reneging Encouraging investment and public-private partnerships. on national legislative and constitutional commitments Since the early 2000s, the principle of encouraging public- may involve less risk for some governments than breaking private partnerships in government procurement has their pledges to a powerful trading partner. The key point gained momentum. In recognition of both tight budget is that alternative reform vehicles exist for public procure- constraints and the growth of private sector capital ment regimes, and the case needs to be made that, given its markets, governments have sought to fund investment (or history and other relevant circumstances, a developing capital) projects with contributions from the private sector. country’s interests are best served by signing public pro- Although the contractual circumstances are hardly uni- curement provisions. Put this way, it may be the case that form, a private sector partner typically puts up the capital no generalizations about the desirability of public procure- for a state project in return for the right to operate the ment provisions in trade agreements are possible. And just related state facilities and charge users of those facilities. because not every developing nation will benefit from such Many such partnerships are effectively off the govern- provisions does not imply that no developing country will. ment’s balance sheet, precisely because the private sector As is shown in figure 17.1, two-thirds of the PTAs noti- advanced all the financing, but the associated transactions fied to the World Trade Organization (WTO) since 2000 are still part of government procurement and ought to be include provisions related to government procurement, treated as such. and about 28 percent of extant PTAs treat government pro- curement in a comprehensive way. (See the annex to this chapter for a list of PTAs with government procurement Implications for Negotiation of Trade Obligations provisions.) But PTAs are not the only instruments that Concerning Public Procurement regulate government procurement on an international The motives for negotiating and agreeing on public pro- scale (see box 17.2) Nonbinding guidelines, such as the curement provisions in PTAs are not limited to market access. Provisions of trade agreements fall broadly into three categories, delineated by their specific objectives: Figure 17.1. PTAs Containing Government entrenching rules, limiting cross-border discrimination Procurement Provisions, 2009 (thereby opening markets), and promoting state-state cooperation and the orderly settlement of disputes. Gov- ernment procurement, like other behind-the-border issues, falls into a fourth category, as the provisions on this subject also deal with the establishment, funding, operations, and 28% review of the public institutions associated with the 37% national procurement regime. A government might strate- gically accept binding rules on its national procurement regime because, in the government’s assessment, these rules are the most effective way of reforming national prac- tices. Consequently, it is misleading to think in terms of the 35% gains of these provisions solely in terms of what additional sales can be made in a trading partner. Market access is not the only possible benefit, and PTAs can contribute toward institutional improvements that have significant develop- ment payoffs. PTAs with comprehensive government The possibility that trade obligations can be used to procurement provisions improve a national procurement regime immediately PTAs containing a few government procurement provisions raises the question of whether there are other, potentially more effective, vehicles available to governments for attain- PTAs without government procurement provisions ing the stated ends. In principle, changes in national legis- lation or in a nation’s constitution are alternatives, and the Source: Anderson et al. 2009. 372 Kamala Dawar and Simon J. Evenett Box 17.2. Three International Government Procurement Instruments The three basic types of international instruments on government procurement, and the principal examples or actors, are as follows: Model procurement codes, guidelines, and statements of principles or best practices • United Nations Commission on International Trade Law (UNCITRAL) Model Law on Procurement of Goods, Construction and Services • United Nations Convention against Corruption • Asia-Pacific Economic Cooperation (APEC) Nonbinding Principles Procurement guidelines imposed by central financial institutions • World Bank • Regional development banks Binding agreements or directives • World Trade Organization (WTO) Government Procurement Agreement (plurilateral) • Preferential trade agreements (PTAs): European Union (EU) directives; Common Market for Eastern and Southern Africa (COMESA); West African Economic and Monetary Union/Union Économique et Monétaire Ouest-Africaine (WAEMU/ UEMOA) United Nations Commission on International Trade Law curement processes internationally, reducing transparency (UNCITRAL) Model Law and the Asia-Pacific Economic and competition. Cooperation (APEC) Nonbinding Principles, as well as A government procurement provision may be explicitly guidelines of lending agencies, play important roles in set- discriminatory, but this does not imply that the implemen- ting out international norms for reforming domestic pro- tation of the provision is necessarily harmful to the com- curement frameworks.8 mercial interests of third parties. Indeed, many provisions The APEC principles, as stated in the 1994 Bogor Decla- in PTAs require changes in national procurement regimes ration, include a transparency standard, value for money, that, as a legal matter, need only be shared with signatories. open and effective competition, accountability and due If, however, operating dual administrative systems is very process, fair dealing, and nondiscrimination. These princi- costly, a signatory may decide that it is cheaper to share all ples are similar to the objectives, whether binding or the PTA-induced improvements with all of its trading part- nonbinding, enunciated in other international instru- ners. In that case, the agreement may allow de jure discrim- ments, including the WTO’s Government Procurement ination with respect to a particular provision, but, de facto, Agreement (GPA). All international frameworks stress the no discrimination occurs. This observation does not imply importance of transparency and an optimal use of that there is no discrimination in the public procurement resources and acknowledge the need for nondiscrimina- provisions in PTAs but, rather, that it is possible for a dis- tion and competition in procurement markets, within a criminatory provision to generate most favored nation rules-based procurement system. The nonbinding princi- (MFN) benefits. Put simply, criticism of PTAs on the basis ples cannot, however, provide any significant degree of leg- of the effects of discriminatory tariff reforms need not islative push or legislative certainty. carry over to public procurement provisions. Again, PTAs with government procurement provisions have straightforward generalizations may not be possible. similar general objectives, but PTA negotiations, particu- Just because a PTA contains potentially discriminatory larly among similar countries, can allow for the provisions provisions does not imply that its implementation will to be better tailored to parties’ needs. Parties to a PTA may cause harm to nonsignatories; a PTA may trigger MFN share cultural interpretations of principles such as fair- improvements in public procurement institutions. It ness, accountability, and integrity, and the consequent would also be wrong to infer that because a national pro- ability of a PTA to promote the harmonization of procure- curement regime could be improved, PTA provisions are ment rules may enable bidders to better predict methods the best vehicle for doing so. Open-minded, case-by-case of tendering, selection, and adjudication, thereby increas- assessments of the merits of such provisions are probably ing the efficiency and competitiveness of the system. There the best counsel for policy makers and those that advise is, however, a risk of creating a patchwork of different pro- them. Government Procurement 373 Finally, reverting to the earlier discussion, individual ensure that both parties’ suppliers are given equal access to governments’ preferences as to the objectives of procure- each others’ government procurement markets. ment vary. Procurement provisions negotiated in a PTA EU and U.S. PTAs with industrial economies such as should reflect both agreement on policy-neutral ways to Chile and Australia, although regarded as relatively com- achieve better regulation (such as transparency) and some prehensive, are less ambitious regarding market access, con- degree of acceptable exceptionality that can be accommo- taining instead general principles of nondiscrimination, dated through specific exceptions or exemptions, as dis- reciprocity and transparency.9 The economic partnership cussed in the next section. agreement (EPA) between the EU and the Caribbean Forum of African, Caribbean, and Pacific (ACP) States (CARIFORUM) is unique in including only transparency An Overview of Government Procurement as a general principle, without any binding commitments Provisions in PTAs regarding market access. Surveys of government procurement provisions in PTAs The U.S.–Jordan PTA contains a single commitment on worldwide indicate that these regimes exhibit a wide vari- government procurement, to the effect that the parties ety and may overlap (Bourgeois, Dawar, and Evenett 2007; support Jordan’s accession to the WTO Government Pro- Dawar and Evenett 2008). The most comprehensive curement Agreement. This clause could promote the mar- regimes—for example, the PTA between the EU and Chile, ket access interests of both parties and of all existing the North American Free Trade Agreement (NAFTA), and members of the GPA. So far, PTAs entered into by the EU the Dominican Republic–Central America Free Trade have not included clauses committing a party to accede to Agreement (CAFTA–DR) PTA—contain detailed provi- the GPA. Instead, their negotiated texts tend to set recipro- sions on government procurement and related issues, such cal and gradual liberalization of procurement markets as a as dispute settlement. At the other end of the spectrum, goal without specifying the scope or coverage of the agree- some PTAs omit procurement altogether; examples are the ment. This is the case with the EU–Morocco association East African Community (EAC) and the Association of agreement, which states that the council set up by the Southeast Asian Nations (ASEAN). Other agreements set agreement must implement the mutual opening of pro- out minimal provisions covering only transparency, cooper- curement markets.10 ation, or the gradual liberalization of procurement markets. To make the picture even more varied, not all industrial The variety of regimes largely stems from the fact that countries have chosen to include government procurement government spending is the preserve of sovereign decision provisions in their agreements, as illustrated by the making, providing a readily available (but not necessarily Canada–Costa Rica and New Zealand–China PTAs. South- effective) tool for favoring particular domestic policies, South PTA provisions also vary widely. In ASEAN, the sectors, or communities. Consequently, the willingness of EAC, and the Southern Cone Common Market (Mercosur, governments to use PTAs to reform, or to codify the reform Mercado Común del Sur), for instance, no government of, public purchasing practices depends on national cir- procurement provisions have been negotiated, whereas in cumstances and international opportunities. Policy makers the CAFTA–DR agreement, the general principles accord have many options available to them as they consider their national treatment and nondiscrimination to all parties to country’s strategy toward government procurement provi- the PTA. In the Common Market for Eastern and Southern sions in PTAs. Where governments have been proactive, Africa (COMESA), the guiding principle is to promote significant provisions have been developed, as discussed regional procurement integration through cooperation next. and information exchange rather than through binding procurement laws. Examples of Government Procurement Provisions in PTAs Defining the Scope of Government The most comprehensive government procurement agree- Procurement Provisions in PTAs ment to date is the Australia and New Zealand Govern- ment Procurement Agreement (ANZGPA). The general Most PTAs that include government procurement provi- principle behind this agreement is to form a single govern- sions tend to follow the WTO GPA positive-list approach. ment procurement market “to maximise opportunities for This means that the scope of the provisions is defined competitive [Australia and New Zealand] suppliers and during the negotiations, and the obligations apply only to reduce costs of doing business for both government and procurements by the entities listed in the annexes to the industry.� The procurement provisions are designed to text. (In the GPA, the relevant provisions are Annexes 1–3 374 Kamala Dawar and Simon J. Evenett to Appendix I.) The positive-list system allows for greater above the thresholds negotiated by the parties. To the national flexibility and a more incremental approach to extent that other PTA provisions limit or condition dis- procurement reform for the included entities. For example, crimination, the thresholds agreed in a PTA have an NAFTA only regulates federal or central government enter- important effect on the extent to which domestic firms will prises and certain parastatals.11 State and provincial gov- face additional competition from foreign rivals after the ernment entities are excluded, although the governments PTA comes into effect. “encourage� voluntary and reciprocal participation by Thresholds may differ according to the type of procure- their respective subnational units. The PTA between the ment and the level of government making the purchase. In EU and Chile includes a positive list for the European fed- effect, these thresholds allow for a partial opening up of the eral entities covered for each EU member state. Although covered sectors, offering governments flexibility to pro- most member states also follow a positive-list approach at mote other policy objectives through certain excluded the subnational level, some EU members, such as Finland, sectors of the procurement market. Using thresholds to have chosen to employ a negative-list approach, which promote an incremental approach to reform extends shel- means that, except as explicitly specified, all public or pub- ter from these obligations to sectors and entities operating licly controlled entities or undertakings that do not have an below the threshold. The provisions of the EPA between industrial or commercial character are subject to the govern- the EU and CARIFORUM, for example, apply only to cen- ment procurement provisions. Chile, by contrast, follows a tral government contracts in excess of one of the highest strict positive list for both central and municipal levels. thresholds yet negotiated, US$200,000. The negative-list approach is used most notably in the Notwithstanding the need to tailor procurement provi- deep integration PTA between Australia and New Zealand: sions to the domestic circumstances of a country, a cost- all government entities are subject to the procurement benefit analysis of high thresholds and preferential policies obligations except those that are explicitly listed as exempt. is essential. Such analyses, however, are scarce. As was This broad approach complements the objectives of the noted above, using discriminatory procurement as a devel- Australia–New Zealand Closer Economic Relations Trade opment tool detracts from the beneficial effects of applying Agreement (ANZCERTA), of creating a single market value-for-money criteria to the expenditure of public between the two parties and achieving the maximum ben- funds. These benefits are usually sought in the most inte- efits from bilateral trade. The choice of a negative-list grated PTAs, such as the ANZGPA, which, consequently, do approach to negotiations on the entities covered by not set any thresholds. government procurement provisions is typically taken to Countries that are parties to PTAs and are also members signal greater liberalizing ambition. The total amount of of the WTO GPA cannot set higher thresholds in bilateral commercial opportunities created depends, however, on a and regional agreements involving other GPA members wider range of factors. than those stated in the GPA. That agreement’s thresholds Following the WTO GPA approach, and mirroring its vary according to level of government (central govern- text, the EU–Chile obligations apply to “any law, regula- ment, subcentral government entities, and government- tion, procedure or practice regarding any procurement, by owned enterprises) and are expressed as special drawing the entities of the Parties, of goods and services including rights (SDRs), an accounting unit used by the Interna- works, subject to the conditions specified in the relevant tional Monetary Fund (IMF).13 In order to capture further Annexes.�12 As with the WTO GPA and most PTAs, the benefits from liberalizing procurement markets, the more EU–Chile obligations do not apply to some kinds of con- recent U.S. agreements (except those with Bahrain and tracts, including international agreements; contracts per- Morocco) have negotiated thresholds lower than those taining to land acquisition, broadcasting, arbitration, agreed to in the GPA. employment, and financial services; and certain research and development (R&D) contracts. U.S. PTAs, the Incorporating Development or Domestic Policy Objectives ANZGPA, and the WTO GPA include both goods and serv- ices and, most notably, big-ticket items such as construc- To afford further flexibility in implementing government tion services. procurement provisions, parties to PTAs with government procurement regimes may negotiate specific exceptions in the scope or coverage of the agreement (see box 17.3). Thresholds Most of the more comprehensive agreements explicitly The coverage of government procurement provisions in a prohibit government entities from imposing measures PTA applies only if the value of the procurement is at or aimed using requirements regarding domestic content, Government Procurement 375 Box 17.3. Examples of Flexible Provisions in Government Procurement PTAs Scope: Level of entities covered by the provisions • Central or federal level only: U.S. preferential trade agreements (PTAs); set to commitments in the World Trade Organization (WTO) Government Procurement Agreement (GPA) • Subcentral entities: European Union (EU) PTAs; set to WTO GPA commitments • All entities: Australia and New Zealand Government Procurement Agreement (ANZGPA); New Zealand–Singapore PTA Thresholds: Minimum value at which procurement is covered by provisions • WTO GPA thresholds or lower: GPA parties • Transitionally higher than WTO GPA thresholds: PTAs between the United States and Bahrain, the United States and the Dominican Republic–Central America Free Trade Agreement (CAFTA–DR), and the United States and Oman • Higher than WTO GPA thresholds: economic partnership agreement (EPA) between the EU and the Caribbean Forum of African, Caribbean, and Pacific (ACP) States (CARIFORUM) • No thresholds: ANZGPA; New Zealand–Singapore PTA; Chile–Costa Rica PTA Exclusions • Defense procurement • Transport: U.S. PTAs • Financial services: Chile PTAs (e.g., U.S.–Chile; Mexico–Chile) • Health, education, and welfare: New Zealand (e.g., Trans-Pacific Strategic Economic Partnership) Set-asides • Preferential opportunities for indigenous persons: Australian agreements (e.g., Singapore–Australia PTA, Article 15) • Small and minority businesses: U.S. general note to annexes Offsets • Prohibited: U.S.–Singapore, Japan–Switzerland, EU–Chile, and other PTAs • Not prohibited: EU–CARIFORUM EPA, European Free Trade Association (EFTA), Singapore–Australia PTA, EU–Jordan PTA, and others licensing of technology, investment, and the like to encour- negotiated rules allowing minimum local-content rules for age local development or improve balance of payments awardees of state contracts. Such provisions are not con- accounts. These measures are known as offsets and are fined to South-South arrangements. Promotion of the defined in Article XVI of the WTO GPA. There are, how- development of local industry and local employment is ever, other measures that give the parties flexibility to pro- included in the Australia–Singapore FTA, which explicitly mote domestic policy objectives. For example, under allows the Australian government to promote employment NAFTA’s joint programs for small business, a committee is for significant indigenous communities. established to report on the efforts being made to promote In addition to provisions allowing for set-asides and government procurement opportunities for members’ other exceptions, cooperation and technical assistance can small enterprises. NAFTA also initially allowed Mexico’s also be specified in the agreement. For instance, the national oil and electric companies to set aside one-half of EU–Chile government procurement provisions state that their procurement each year for domestic suppliers. (This the parties will seek to provide technical assistance provision was entirely phased out by 2003.) Unlike the on issues connected with public procurement, paying United States and Canada, Mexico is not a member of the special attention to the municipal level. The North-South WTO GPA, and its procurement is therefore subject only to EU–CARIFORUM agreement includes provisions on the obligations negotiated in PTAs. Another NAFTA excep- cooperation and technical assistance, with commitments tion allowed Mexico to impose local-content requirements of financial resources. There are also asymmetrical require- for turnkey construction projects. For capital-intensive ments that the EU show “due restraint� in resolving dis- projects, Mexico negotiated set-asides for as much as 25 putes in favor of CARIFORUM parties. This is ambiguous, percent for local inputs, and for up to 40 percent Mexican but it could result in the EU’s resolving disputes by giving content for labor-intensive projects. the CARIFORUM partners the benefit of the doubt, or In the free trade agreement (FTA) between Mexico demanding less retaliation. The agreement also allows a and Nicaragua, in addition to set-asides, both parties significant implementation period to give CARIFORUM 376 Kamala Dawar and Simon J. Evenett countries ample time to prepare for implementation and the accountability of the officials and the procuring agen- provides for development support. cies. The existence of the mechanism improves the system’s In addition to exceptions in individual schedules, pro- reputation, reducing barriers to entry caused by a poor curement obligations in U.S. PTAs (like those of the WTO perception of the integrity or due process of procurement GPA) are subject to national security and general excep- markets. An important point is that complaints do not tions, to exclude sensitive sectors from the overall rules. concern the performance of the contract, once it has been These exceptions vary in their scope and details. The U.S. initiated. (That type of performance complaint is covered PTAs with Chile and Singapore include “essential security� in the contract itself.) provisions specifically applicable to government procure- Domestic review mechanisms can be located within a ment obligations, but this type of national security excep- contracting agency, a dedicated independent entity, or the tion is not present in the U.S.–Morocco and U.S.–Australia general court system. Each option has advantages and dis- agreements. In general, the government procurement advantages relating to issues such as perceived independ- exceptions in the EU’s PTAs broadly follow the template ence, expertise, efficiency, and authority. A common model contained in Article XX of the General Agreement on Tar- used in the more comprehensive PTAs is to house the bid iffs and Trade (GATT); Article 161 in the EU–Chile PTA is challenge mechanism in a designated body or agency. A an example. U.S. PTAs are different in that they include deep integration model such as the ANZGPA typically additional specific measures, such as those necessary to embeds the monitoring procedures in an annex to the protect human, animal, or plant life or health (e.g., agreement. The annex then commits the parties to identify U.S.–Chile PTA, Article 9.16).14 a designated body as the responsible authority and point of contact for complaints. Monitoring is triggered by the examination of alleged breaches of the agreement by the Settling Disputes over Government Procurement in PTAs other party. The designated body investigates the com- An effective dispute resolution mechanism is critical to the plaint, and if the matter cannot be resolved, it has the effectiveness of government procurement disciplines in power to refer the case to the ministerial level for further PTAs. It also has broader application because government investigation, if necessary. This is a very streamlined model procurement dispute settlement provisions include a num- that is able to balance the lack of formal procedures with ber of features that might usefully be employed in other speed and efficiency and is mainly suitable for more highly behind-the-border policies (see Porges, ch. 22 in this integrated markets. volume). The NAFTA bid challenge system is much more pro- Within a comprehensive PTA government procure- cedurally extensive. It obligates the parties to adopt and ment regime, there can be two or possibly three levels of maintain bid challenge procedures that allow suppliers dispute settlement. The first level consists of the proce- to submit challenges concerning any aspect of the pro- dures and institutions governing disputes between the curement process. It seeks to ensure that the contracting procuring entity and a disappointed bidder (the disputes entities accord fair and timely consideration to any being those relating to the procurement procedure complaint and sets out minimum time limits for the sub- itself). The second level is the system governing disputes mission of complaints. Independent reviewing authori- between the parties to the agreement—that is, state-state ties must be identified, and the “entities normally shall disputes regarding compliance or the implementation of follow the recommendations of the reviewing authority� the government procurement provisions within the for bringing the actions into conformity with their agreement. At the third level, there may be a clause gov- obligations. erning disputes between state parties that are simultane- A similar template obligating the parties to provide pro- ously members of other relevant international agreements cedures that are transparent, timely, impartial, and effective with overlapping jurisdictions that also include dispute can be found in the CAFTA–DR, EU–Chile, U.S.–Chile, settlement mechanisms. and EU–CARIFORUM agreements. Most of these agree- Complaints about the procurement process. The availabil- ments do not specify the measures available for remedy of ity of dispute settlement of complaints matters because it breaches or the amount of compensation. The EU–Chile means there is a self-policing and self-enforcing mecha- agreement, for example, states that the challenge proce- nism for the procedural provisions set out in an agree- dures shall provide for correction of breach of the provi- ment.15 A dispute settlement mechanism provides an sions or for compensation for damages, which is limited to essential forum for airing complaints and obtaining relief, the costs of tender preparation and protest. The govern- and it offers the parties due process rights while enhancing ment procurement provisions in the EU–Mexico FTA Government Procurement 377 concerning the gradual and reciprocal liberalization of sig- tions are binding. Thus, in comparison with integrated sys- natory procurement markets are less ambitious, and the tems such as those of the EU, EFTA, or COMESA, which challenge procedures are correspondingly less well devel- have permanent international courts to settle disputes oped. The agreement simply states that the PTA’s joint between member states, individuals, and the organization’s council should decide on the construction of “clear� chal- institutions, NAFTA is less institutionalized and relies lenge procedures. mostly on ad hoc arbitration and diplomacy. As a model for some of the more comprehensive PTA The state-state dispute settlement provisions in the provisions, Article XX of the WTO GPA, on challenge pro- U.S.–Chile and CAFTA–DR agreements have a similar cedures, includes similar language concerning first- framework, typically reflecting the multimember nature of instance consultation procedures, transparency and good the PTAs. Both sequence dispute settlement procedures in governance of the review proceedings, and time limits. The the same manner, although the CAFTA–DR PTA has an article further states that the review body should be a court additional provision allowing for multiple complainants or impartial entity subject to judicial review or similar pro- and third-party participation in dispute proceedings and cedures. The challenge provisions include the option of setting out different procedures. Under this agreement, a taking rapid interim measures to correct breaches of the party that considers it has a substantial trade interest in a agreement and limitation of compensation for damages dispute between other parties to the agreement may partic- suffered to the costs of tender preparation or protest. ipate after sending a written notice explaining its interest in Those PTAs that have soft cooperation provisions cov- the matter to the other parties within seven days of delivery ering government procurement or best-endeavor clauses of the initial request for consultations (Article 20.4, on do not include bid challenge mechanisms because these are consultations). That party automatically becomes one of to be developed within domestic legislation. Where review the consulting parties and may request a meeting of the mechanisms are included in such agreements, they neces- agreement’s commission if the matter is not resolved sarily apply only to procurement that lies within the scope within a specified time. and coverage negotiated by the parties. Overlapping jurisdictions and dispute settlement mecha- Existing approaches in state-state disputes. Where nism “shopping.� Article XXII of the WTO GPA, on consul- neither party to a PTA is a member of the WTO GPA, dis- tations and dispute settlement, provides that the WTO Dis- putes about nonimplementation of the government pro- pute Settlement Understanding (DSU) applies if a GPA curement provisions are governed only by the PTA itself. party considers that an objective of the agreement or a ben- This is the case, for instance, with ANZCERTA, the deep efit accruing to the party from the agreement is being nulli- integration PTA between Australia and New Zealand. No fied or impaired because another member has failed to specific procedure is established to govern disputes related carry out its obligations. The Dispute Settlement Body to noncompliance of government procurement provi- (DSB) has the sole authority to establish panels of experts to sions; instead, the agreement states that “the close and consider the case and to accept or reject the panels’ findings long-standing political relationship between Australia and or the results of an appeal. It monitors the implementation New Zealand means that any issues of grievance or con- of the rulings and recommendations and can authorize cern are addressed through discussion between the two retaliation when a country does not comply with a ruling. Governments.� The DSB’s recommendations are to be implemented, and NAFTA’s dispute resolution provisions are applicable to the relevant party must state its intention of doing so at a all disputes regarding the interpretation of application of DSB meeting held within 30 days of the report’s adoption. If NAFTA and are “intended to resolve disputes by agree- it does not, the complaining side may ask the DSB for per- ment, if at all possible.� Because the process encourages the mission to impose limited trade sanctions against the non- use of arbitration for settlement of disputes between par- compliant party; these sanctions are ideally restricted to the ties, it begins with government-to-government consulta- same sector as the nullification or impairment. tions. When these general disputes are not resolved Because of its legalistic and binding framework, the through consultation within a specified period of time, DSB is one of the strongest trade arbitration forums. Its either party can request that the dispute be referred ini- strength makes it attractive to weaker states that are unable tially to the “good offices� of NAFTA’s Free Trade Commis- to exert diplomatic pressure on a noncompliant party. But sion. If the dispute is not resolved within a fixed time a lack of coherence in international trade law can result period, the matter can be referred to a panel for ad hoc when governments are faced with multiple and overlap- arbitration. Each party selects two panel members, and the ping obligations and jurisdictions to choose from when chair is chosen by consensus. The panel’s recommenda- trying to settle a dispute. 378 Kamala Dawar and Simon J. Evenett To offer more predictability and transparency, some grounds that the United States refused to form a panel to PTAs include rules that dictate the choice and sequencing settle the dispute under NAFTA. The WTO Dispute Settle- of dispute settlement systems. In some PTAs, a request for ment Body declined jurisdiction because the dispute did the establishment of a WTO panel excludes the jurisdiction not fall within the scope of the WTO agreements. of the regional forum, or the use of one mechanism Clearly, parties tend to choose the recourse most likely excludes the use of another (to prevent “forum hopping�). to generate a favorable outcome for themselves. The vari- A PTA cannot take away the jurisdiction of the WTO GPA: ables considered will include the scope and legal status of if a party to both a PTA and the GPA appeals to the WTO the measure in dispute; the applicable law; the procedures, Dispute Settlement Body, the panel will not refuse it structure, and time frame of each mechanism; the reme- because of PTA rules. (The panel will, however, refuse the dies available; and the inherent characteristics of each dispute if it does not concern the application of the GPA or dispute settlement process. That is, consideration of the another WTO agreement.) political circumstances of the dispute can also influence its The U.S.–Chile and CAFTA–DR PTAs have clauses clar- resolution, as is seen in the dispute between Mexico and ifying the forum to be used in a dispute. A complaining the United States over soft drink sweeteners (for back- party may select the forum, which may be either the PTA ground, see Davey and Sapir 2009). concerned, another PTA to which the parties are party, or the WTO DSM, as appropriate. Once the party requests a Costs of Implementing Government panel under the jurisdiction of one agreement, the forum Procurement Provisions selected is to be used to the exclusion of others. The arrangement in the EU–Chile agreement is more Information on the costs of implementing government elaborate: it provides that unless the parties agree other- procurement provisions in PTAs is scarce.16 Still, enough is wise, when a party seeks redress for a violation of an known about the content of potential PTA provisions that obligation under the forum exclusion clause (which is some tentative observations can be made about the nature equivalent in substance to an obligation under the WTO), and extent of these costs. it shall have recourse to the relevant rules and procedures Setting up an effective government procurement regime of the WTO Agreement, which apply notwithstanding the requires specialized institutional frameworks and expert- provisions of the PTA. Once dispute settlement procedures ise. Although the economic and welfare benefits to be have been initiated in a selected forum, that forum is to be gained from fair and transparent procurement markets are used to the exclusion of any other likely to be sizable as a share of the total amount spent on NAFTA’s members are the United States and Canada, goods and services by the government, the procedures and which are parties to the WTO GPA, and Mexico, which is challenge mechanisms which enable bidders to feel confi- not. In the case of a dispute relating to procurement dent that procurement processes are fair and transparent between the United States and Canada, those countries will require resources. The implementation of a PTA, can choose whether to use the PTA or the WTO forum. however, does not necessarily entail new costs, as the gov- Mexico’s procurement is exclusively governed by NAFTA, ernment may have borne some or all of these costs before- and so it must adhere to the procedures specified in that hand. What share of additional implementation costs is agreement. PTA-specific is an open, factual question, the answer to The NAFTA dispute settlement procedures are oriented which is likely to differ according to the country and the toward diplomatic or negotiated solutions. They provide preexisting procurement regime. that disputes regarding any matter that arises under both Implementation costs are likely to vary across govern- NAFTA and the General Agreement on Tariffs and Trade ment procurement provisions. For example, transparency (GATT), or under any agreement negotiated under these provisions that require signatories to publish all relevant arrangements, or under any successor agreement, may procurement regulations in a foreign language will require be settled in either forum at the discretion of the complain- translators with legal expertise, and retaining the special- ing party. A party is to notify any third party of its inten- ized legal talent to adjudicate complaints on procurement tion to bring a dispute, with a view to agreement on a matters and present appeals before tribunals is a distinct single forum. If the parties cannot agree, the dispute is nor- resource challenge. The nature and timing of implementa- mally to be settled under the NAFTA procedure. This has tion costs will thus differ among classes of government led in the past to some danger of legal fragmentation, as procurement provisions and capacity-building needs. can be seen in the U.S.–Mexico soft drinks dispute, when The costs associated with setting up an effective Mexico tried to bring a NAFTA dispute to the WTO on the national procurement system are usually borne by the Government Procurement 379 jurisdiction concerned. The review of PTAs in this section established. This committee is to meet as necessary, but not indicates that no regional institutions had to be created to less than once a year. It may establish working parties to implement any of the agreements. Even in the PTAs with carry out specific functions. Jordan’s decision to accede to the most comprehensive government procurement com- the GPA, however, was made independently and before the mitments, the institutional implications are all at the agreement with the United States. national level. This contrasts with regional competition In sum, much of any burden of implementation of PTA regimes that have established a regional competition law obligations is borne at the national (and, potentially, and a competition authority to monitor the implementa- subnational) levels. But there is another significant oppor- tion of the provisions (see Dawar and Holmes, ch. 16 in tunity cost that some governments may perceive from sign- this volume). Although setting up an effective and trans- ing government procurement provisions: constraints on parent government procurement system is costly, there are the use of state spending as an “industrial policy� tool for no further costs for establishing and running the regional promoting targeted industries.17 This is not the place for a procurement institutions provided for in these PTAs. This full exposition on industrial policy; suffice it to say that is so even in the case of the deepest government procure- assertions about the implications of government procure- ment integration framework, the ANZGPA between ment reforms for such policies have been made in both Australia and New Zealand. The implementation of the developing and industrial countries. An important ques- agreement relies on existing national bodies, and no supra- tion is whether public procurement policies optimally tar- national institutions need to be established. The agreement get whatever market failure is holding back the industries contains the monitoring requirement that each jurisdic- in question. tion must have at least one designated body to be the point of contact for complaints, with the authority, responsibil- Technical Assistance ity, and expertise to handle and investigate complaints across government and public sector agencies covered by Technical assistance requirements related to government the agreement. If the complaint is multijurisdictional, all procurement in PTAs may be quite substantial. For example, the relevant designated bodies and, where necessary, minis- NAFTA requires the parties to provide, on a cost-recovery ters are included in the procedure. basis, information concerning training and orientation pro- Similarly, the CAFTA–DR procurement regime includes grams relevant to their government procurement systems commitments to establish or designate at least one impar- and to grant nondiscriminatory access to any programs they tial administrative or judicial authority, independent of conduct. Such activities include training of government procuring entities, to receive and review challenges by sup- personnel directly involved in government procurement pliers relating to the obligations of the party. Again, this bid procedures; training of suppliers interested in pursuing gov- challenge mechanism is to be established at the national ernment procurement opportunities; an explanation and level and does not entail any institutional buildup, assum- description of specific elements of the party’s government ing that there is already a domestic institution for review- procurement system, such as its bid challenge mechanism; ing domestic complaints. The same is true of those PTAs and information about government procurement market with provisions that are restricted to ensuring transparency opportunities. NAFTA parties are also required to establish in government procurement, such as the EU–CARIFORUM at least one contact point to provide information on the agreement. Each party must identify or designate at least training and orientation programs. These NAFTA provi- one impartial administrative or judicial authority that is sions are much more extensive than the analogous obliga- independent of its procuring entities to receive and review tions in other PTAs. For instance, the EU–Chile PTA states challenges by suppliers arising in the context of covered only that the parties will seek to provide technical assistance procurement. on issues connected with public procurement, with special In the case of the U.S.–Jordan agreement, the institu- attention to the municipal level. tional implications are broader than a commitment to For PTA developing-country members that are also establish national procurement institutions to implement signatories of the WTO GPA, the technical assistance pro- the obligations of the PTA. Article 9 of the agreement states visions are more substantive than at the regional level. that, pursuant to Jordan’s application for accession to the Article V of the GPA, on special and differential treatment WTO GPA in 2000, the parties shall enter into negotiations for developing countries, states that each developed- with regard to Jordan’s accession to the WTO GPA. The country party shall, on request, provide all the technical GPA requires that a committee on government procure- assistance which it may deem appropriate to developing- ment, composed of representatives from each the party, be country parties to help resolve their problems in the field 380 Kamala Dawar and Simon J. Evenett of government procurement and that this assistance must ment, as in other situations of less-than-perfect competi- be provided in a nondiscriminatory way. Assistance is to be tion (see Baldwin, ch. 3 in this volume), deviations from directed toward the solution of particular technical prob- free trade could theoretically be welfare improving. For lems relating to the award of a specific contract and is to instance, imposing price preferences on foreign suppliers include translations into an official WTO language of qual- (i.e., raising their bids by a certain percentage) will induce ification documentation and of tenders made by suppliers them to lower their bids for state contracts. If enough of of developing-country parties. Developed-country mem- those foreign suppliers enjoy a cost advantage over their bers are required to establish, individually or jointly, infor- domestic rivals, the probability that a foreign supplier mation centers to respond to reasonable requests from could still win the contract may be high enough that the developing-country parties for a variety of specified pro- expected cost of the government procurement will fall. In curement information. these circumstances, promoting a level playing field may not necessarily be welfare maximizing. Yet, the lack of capacity to collect and analyze the necessary information Government Procurement Provisions to discriminate optimally between suppliers calls into in PTAs in a Multilateral World question whether a state purchaser could in practice tap In assessing the trade-negotiating options facing a govern- the gains from exploiting its buying clout. Moreover, as ment in the area of government procurement practices, it is Deltas and Evenett (1997) show in a series of simulations commonplace to compare nondiscriminatory PTA options of procurement bidding situations, the gains from discrim- with multilateral reform conducted on an MFN basis. Such ination may be small (when expressed as a percentage of assumptions, however, should not always guide policy the purchase price), and even tiny mistakes in setting the advice in the government procurement arena, precisely optimal degree of discrimination could result in increased, because the GPA is a plurilateral accord that extends “con- not reduced, procurement costs. All in all, a rule of thumb cessions� only to other signatories to that accord. As far as of open competition is recommended to policy makers. market access is concerned, there is no nondiscriminatory This implies opening up state procurement contracts to the multilateral benchmark to rely on, and as long as the maximum number of appropriately qualified bidders. national institutions associated with implementing public Second, traditional concerns related to PTAs about the procurement are exempted from the WTO’s national treat- risk and cost of trade diversion could also apply to the pref- ment provisions, there is no multilateral benchmark there, erential liberalization of government procurement in these either. agreements. Whether trade is diverted from efficient to The absence of benchmarks does not imply that pursuit less-efficient foreign suppliers will depend on the treat- of nondiscriminatory reform through a trade agreement is ment of foreign bids before the PTA was concluded. If bids impossible. At present, such reform can take place unilater- from foreigners were allowed and price preferences were ally or, somewhat paradoxically, in the context of a prefer- applied on an MFN basis before the PTA was enacted, the ential PTA. In the future, however, should WTO members traditional concerns about trade creation and trade diver- extend traditional nondiscriminatory disciplines on a mul- sion arise.18 By contrast, if no bids from foreign firms were tilateral scale, another means of liberalizing public pur- allowed before the PTA was signed, the additional competi- chases on an MFN basis will become available. With these tion for state contracts is likely to push down the price paid remarks, it should be clear that any analogies between tariff by the public buyer, and there was no trade to divert in the reform strategies and public procurement reform strategies first place. Overall, this suggests that if a PTA generates, for in the context of trade agreements have to be drawn very the first time, significant foreign competition for state con- carefully. Blanket application of the logic of the former to tracts, there are likely to be economic benefits in the form the latter is likely to be misleading. of lower procurement costs. For procurement regimes that already benefit from considerable foreign competition, the benefits of preferential reform through PTAs may be Theoretical versus Actual Costs of Discrimination ambiguous and even adverse. There are theoretical reasons why provisions on government A third consideration is that there are various forms of procurement in PTAs could be discriminatory. First, as in discrimination in government procurement systems against any situation in which a buyer has market power—as some foreign bidders. Simulations have been conducted of the state buyers with large budgets might—discrimination increases in procurement costs created by price preferences, against certain suppliers can, in principle, reduce the costs by measures that raise the costs of foreign bidders, and by to the purchaser. In the context of government procure- outright limitations on the number of foreign bidders. Government Procurement 381 Although these simulations do not claim to have covered here could relate to more contracts, lower price prefer- every single case or procurement setting, price preferences ences, lower thresholds, and so on) The use of such a provi- were found to inflict the least harm, and cost-raising meas- sion would ratchet up the degree of competition over time ures were the next least harmful. Given these findings, it is in government procurement markets while simultaneously paradoxical that discrimination though price preferences is ensuring that all beneficiaries of this clause fight for con- most often singled out for banning in PTAs. A perhaps tracts on the same terms (limiting discrimination between preferable alternative might be to eliminate or limit the PTA signatories.) Third-party MFN provisions do not, of other forms of discrimination, even if it comes at the course, limit discrimination against exporters from coun- expense of higher price preferences, at least initially.19 tries that have not signed a PTA, so it would be wrong to conclude that they eliminate all forms of discrimination across trading partners. Transparency and Third-Party MFN Because of the potential confusion regarding third- The implementation of PTA provisions that improve the party obligations in future PTAs, it is becoming more transparency of government procurement procedures (on the common to clarify these issues of overlapping PTA plausible assumption that the entry of both domestic and membership and jurisdictions within the agreement itself. foreign firms is encouraged by such improvements) may Government procurement provisions in the more compre- lead to either reductions or increases in the share of state hensive PTAs aim to open public procurement to foreign contracts awarded to foreign firms (Evenett and Hoekman competition on a preferential basis. In order to protect 2005). There is some evidence that small and medium-size these preferences, some procurement provisions require enterprises are more sensitive to the transparency of third-party MFN guarantees so as to limit the extent to national procurement regimes, in which case PTA-induced which preferential procurement is undermined by subse- improvements in transparency may increase the propor- quent PTAs.22 tion of state contracts awarded to a class of firms that gov- Third-party MFN clauses that extend preferential access ernments, for other reasons, typically wish to favor. automatically reduce the geographic discrimination The example of implementation of improvements in implied by proliferating PTAs. They also allow a govern- transparency on an MFN basis carries over to other PTA- ment to free ride on the negotiating clout of their PTA induced improvements in public procurement regimes partners in future agreements: countries with relatively that are implemented on a nondiscriminatory basis. One small procurement outlays can choose to negotiate for the might ask why a government would voluntarily extend inclusion of these provisions if it seems that the other party benefits to non-PTA signatories.20 The answer is that the may subsequently negotiate an agreement with a larger implementation costs of operating two or more procure- country. Albania, the former Yugoslav Republic of Mace- ment regimes for different trading partners may exceed the donia, and Turkey have in the past included third-party cost of operating a single reformed regime. Indeed, consis- MFN provisions and have followed this strategy. tent with the literature on multilateralizing regionalism It should be noted that the arguments reviewed above (e.g., Baldwin and Low 2009), it is possible that a PTA can are based on economic principles and evidence from simu- induce in a signatory an institutional innovation or lations of procurement auctions. We do not know of a sin- improvement that is willingly shared with all trading part- gle econometric analysis that seeks to estimate the effect of ners. Again, concerns about the inherently discriminatory implementing the government procurement provisions of nature of PTAs must be tempered. The fact that certain a PTA.23 Policy makers and readers should bear this in PTA provisions on public procurement are written in a dis- mind when assessing the above arguments. criminatory manner does not mean that they are so imple- mented or that the net effect of all the PTAs provisions is to Relationship to Open Regionalism limit the commercial opportunities of nonsignatories. There may, however, be more subtle intertemporal rela- Some experts, recognizing the pervasive nature of PTAs, tionships between PTA provisions on government pro- have argued that their discriminatory impact might be lim- curement and nondiscrimination. The use of third-party ited if terms could be defined ex ante under which MFN clauses is a case in point. Under such a clause, should nonsignatories could enter a trade bloc. Clarity of terms of A, a party to a PTA with B, subsequently sign another PTA entry and the desire to limit the loss of commercial oppor- with a third party, C, and in so doing offer C better access tunities from being outside a bloc are said to encourage to A’s government procurement market than B, then B is entry and, ultimately, to expand the amount of trade con- entitled to the same access as C.21 (The better market access ducted under freer, if preferential, terms. In principle, the 382 Kamala Dawar and Simon J. Evenett key ingredients for such open regionalism to work are pres- options should not be approached dogmatically. Rather, a ent in the context of PTA provisions on government pro- case-by-case evaluation is appropriate, taking due account curement, as long as these provisions are implemented on a of the state of national purchasing practices before any nondiscriminatory basis. reform was launched. For example, PTA provisions that Yet, it could be that the preferential trade-related gov- call for improvements in transparency should be treated ernment procurement accord most likely to induce entry is differently from those relating to market access. Moreover, the WTO’s plurilateral GPA, which now has 40 members. pointing to WTO obligations as a nondiscriminatory Existing GPA members are unlikely to offer more market benchmark is not an accurate reflection of the existing access to a new member (so, the terms of entry are clear), state of that organization’s plurilateral accord. Even more and the scope of public procurement covered in the GPA is confusingly, the tension between discrimination and liber- greater than under any existing PTA. This is not to say that alization in PTAs may not be as relevant for public pro- open regionalism could not happen anywhere else but only curement as it is for tariffs. that it would be strange to see such developments happen- Policy analysts would do best to understand, first, ing outside the WTO when the WTO accord provides the national procurement regimes and, second, the potential greatest incentives to join. If that logic is to be taken seri- sources and magnitudes of benefits from reform. Then, ously, the best hope for open regionalism is probably open consideration should be given to which reform vehicle plurilateralism. Then, the provision of the U.S.–Jordan (unilateral, bilateral, regional, or multilateral) offers the PTA that supports Jordan’s accession to the WTO GPA greatest promise over the time frame contemplated. The could be seen to promote open regionalism and open ability to motivate and sustain a constituency in favor of plurilateralism most directly. procurement reform is an important consideration and is likely to vary across reform options, time spans, and juris- dictions. To date, there is insufficient evidence to confi- Implications for Trade Negotiating Strategy dently recommend one reform vehicle over another. and for Evaluation of Policy Options For analysts, much remains to be done in analyzing In recent years, more and more PTAs have included compliance with, and the effect of, government procure- provisions on government procurement, ranging from ment provisions in PTAs. It is unsatisfactory that the evi- transparency-only clauses to the creation of a single dential base, whether in terms of legal compliance or of regional procurement market. These provisions affect an economic effects, that is needed to guide policy making is important area of state behavior—public purchasing— so thin. The hard and, some might say, tedious work of and there is, no doubt, interest in their potential devel- tracking what has happened after governments have taken opment impact. The word “potential� is used deliberately on obligations on procurement in PTAs is still to be done. because policy makers, officials, and analysts would be For example, in many PTAs the emphasis has been on try- wise to differentiate between what has been done and ing to eliminate the more transparent forms of discrimina- what could be done. Arguably, this is an area of trade tion (such as price preferences), and this may have had the policy making in which, at present, the former is far from unintended consequence of driving discrimination into approaching the latter. nontransparent forms. For many countries, government procurement outlays are a sensitive matter, at least in terms of interest groups Annex. PTAs with Government Procurement and politics. The desire for value for money in public pur- Provisions chasing has often been tempered by support for favored industries and groups. This has complicated but has not The following PTAs containing government procurement precluded the negotiation of government procurement provisions had been notified to the WTO as of December provisions in PTAs, often with the full range of exceptions 2009. and other devices used to limit the impact of such provi- Australia–Chile sions. Procedures to review the cost of those exceptions and Canada–Costa Rica to suggest alternative measures of helping favored indus- Canada–European Free Trade Association (EFTA) tries should be given greater consideration in the future. Canada–Peru More generally, many of the rules of thumb that trade Caribbean Community (CARICOM) (services) economists have developed concerning the relative efficacy Central European Free Trade Agreement (CEFTA) of different tariff reform strategies do not carry over to Chile–China public procurement reforms. Evaluations of public policy Chile–Colombia Government Procurement 383 Chile–Costa Rica Mexico–Honduras (services) Chile–El Salvador Mexico–Nicaragua Dominican Republic–Central America–United States Free Pacific Island Countries Trade Agreement (PICTA) Trade Agreement (CAFTA–DR) Panama–Costa Rica EFTA (services) Panama–El Salvador EFTA–Chile Singapore–Australia EFTA–Croatia Singapore–New Zealand EFTA–Egypt, Arab Rep. Singapore–Panama EFTA–Jordan Singapore–Peru EFTA–Korea, Rep. Thailand–Australia EFTA–Lebanon Thailand–New Zealand EFTA–Macedonia, FYR Trans-Pacific Strategic Economic Partnership EFTA–Mexico Turkey–Albania EFTA–Morocco Turkey–Bosnia and Herzegovina EFTA–Singapore Turkey–Croatia EFTA–South African Customs Union (SACU) Turkey–Egypt, Arab Rep. EFTA–Tunisia Turkey–Georgia European Union (EU)–Albania Turkey–Macedonia, FYR EU–Algeria Turkey–Morocco EU–Cameroon Turkey–Palestinian Authority EU–Caribbean Forum of African, Caribbean, and Pacific Turkey–Syrian Arab Republic (ACP) States (CARIFORUM) economic partnership Turkey–Tunisia agreement (EPA) Ukraine–Macedonia, FYR EU–Chile Ukraine–Moldova EU–Egypt, Arab Rep. United States–Australia EU–Israel United States–Bahrain EU–Jordan United States–Chile EU–Mexico United States–Morocco EU–Montenegro United States–Oman EU–Morocco United States–Peru EU–South Africa United States–Singapore Iceland–Faroe Islands Source: WTO data. Israel–Mexico Japan–Brunei Darussalam Japan–Chile Notes Japan–Indonesia 1. A feature of the literature on government procurement is the Japan–Mexico paucity of comparable cross-country estimates of the total amounts spent Japan–Philippines on goods and services. OECD (2001) is the most recent study to have assembled information for many industrial and developing countries, and Japan–Singapore it used 1998 United Nations data. The study shows that in 1998 the level of Japan–Switzerland government spending on goods and services worldwide was, on average, Japan–Thailand 14.5 percent of gross domestic product (GDP); in industrial countries it Japan–Vietnam was 17.1 percent. 2. Although the functions listed in this paragraph may be less inten- Jordan–Singapore sively executed in developing countries, the experiences of certain industrial Jordan–United States countries would suggest that it is unwise to assume that resource con- Korea, Rep.–Chile straints are confined to poorer countries. 3. This argument is developed at greater length and with specific ref- Korea, Rep.–Singapore erence to the imprecise notion “policy space� in Dawar and Evenett Mexico–Chile (2008). Mexico–Costa Rica 4. Strictly speaking there is a wrinkle in the value-for-money princi- ple. To the extent that a government buyer has market (monopsony) Mexico–El Salvador power, the pursuit of this objective may result in prices at which market Mexico–Guatemala outcomes are inefficient—that is, the prices do not equal the marginal Mexico–Honduras (goods) costs of production or the societal costs of producing the last unit of the 384 Kamala Dawar and Simon J. Evenett good or service in question. Given the size of many government pur- 17. Government procurement policies may be used to favor outright chasers relative to total demand in the market for a given good or service, a certain group in society or, indeed, specific individuals. Leaving aside this logical possibility has received remarkably little consideration. whether such favoritism is legal, it has long been suspected that resistance 5. Prevalent forms of discrimination against foreign firms in national to reform of public purchasing practices in some jurisdictions has been procurement regimes include price preferences (which raise foreign bids influenced by the desire to preserve such practices. It is unclear why there by a certain percentage before they are compared with the bids of domes- would be less resistance to reform induced by a trade agreement than to tic firms), outright bans on foreign bidders, local content-related restric- unilateral reform. tions, and standards adopted in the procurement process that raise the 18. Formally, the case of price preferences is different from that of costs of foreign firms. tariffs in that the PTA leads to a loss of government revenues (the lost tar- 6. Observant readers will have noted that any such procurement iff revenues on the favored trade). ban does not remedy the ultimate cause of the small size of the private 19. The replacement of other forms of discrimination by a transpar- sector. ent price-based instrument of discrimination is, of course, not new, as 7. There is some evidence in the corruption literature that improve- attempts at “tariffication� of nontariff measures can attest. The same logic ments in the transparency of government procurement processes result in applies in the procurement case. a shift of government spending toward more homogeneous goods and 20. Indeed, it is frequently contended by European Commission offi- away from goods that are so heterogeneous that bids cannot be easily cials that many government procurement contracts in the EU can be con- compared (which increases the potential for graft to affect the allocation tested by parties from all its trading partners. of state contracts). For a summary of such evidence, see Evenett and 21. One feature of these provisions is that a “weak� negotiating partner Hoekman (2005). Bourgeois, Dawar, and Evenett (2007) summarize the can benefit from a subsequent “tougher� negotiator’s ability to extract more little empirical evidence, mostly from the European Union, on the impact market access concessions from trading partners with which both the weak of procurement reforms and PTA procurement provisions. The principal and stronger parties ultimately sign a PTA. These tactical considerations findings are summarized in this chapter. may add to the attraction of third-party MFN clauses for weaker negotiat- 8. United Nations Commission on International Trade Law, “1994— ing parties, which may number among them many developing countries. UNCITRAL Model Law on Procurement of Goods, Construction and 22. For example, according to Article 67, on further negotiations, of Services, with Guide to Enactment,� http://www.uncitral.org/uncitral/en/ the EFTA–Mexico PTA, “In the case that the EFTA States or Mexico offer, uncitral_texts/procurement_infrastructure/1994Model.html. after the entry into force of this Agreement, a GPA or NAFTA Party, 9. The type of discrimination referred to in the general principles is respectively, additional advantages with regard to the access to their sometimes referred to as conditional MFN, in that each party offers MFN respective procurement markets beyond what has been agreed under this treatment to every other party to the accord. Nonsignatories are not Chapter, they shall agree to enter into negotiations with the other Party offered MFN. with a view to extending these advantages to the other Party on a recipro- 10. This clause is also seen in the EU agreements with Mexico and cal basis.� Article 18, on review of commitments, of the South Asian Free Jordan. The Trade, Development, and Cooperation Agreement (TDCA) Trade Area (SAFTA), states, “If, after this Agreement enters into force, a between the EU and South Africa directs the Cooperation Council estab- Party enters into any agreement on government procurement with a non- lished under the agreement to review progress made in opening fair, equi- Party, it shall give positive consideration to a request by the other Party for table, and transparent procurement markets. incorporation herein of treatment no less favourable than under the 11. The parastatals referred to are government-controlled enterprises; aforesaid agreement. Any such incorporation should maintain the overall they include the U.S. Tennessee Valley Authority (TVA), Canada’s St. balance of commitments undertaken by each Party under this Agree- Lawrence Seaway Authority, and Mexico’s national oil company Petroleos ment.� Article 160, on further negotiations, of the EU–Chile agreement Mexicanos (PEMEX) and national electric company, the Comisión Fed- specifies, “If either Party should offer in the future a third party additional eral de Electricidad (CFE). advantages with regard to access to their respective procurement markets 12. The WTO GPA obligations apply to the goods and services listed beyond what has been agreed under this Title, it shall agree to enter into in Annexes 4 and 5 of the agreement, again typically following a negoti- negotiations with the other Party with a view to extending these advan- ated positive-list approach. tages to it on a reciprocal basis by means of a decision of the Association 13. The thresholds are, for central government entities, 130,000 SDRs Committee.� Article 9, on further negotiations, of the U.S.–Chile PTA for procurement of goods and services and 5 million SDRs for procure- states, “On request of either Party, the Parties shall enter into negotiations ment of construction services; for subcentral government entities, with a view to extending coverage under this Chapter on a reciprocal 200,000 SDRs for goods and services (except for the United States and basis, if a Party provides, through an international agreement entered into Canada, which apply a 355,000 SDR threshold) and 5 million SDRs for after entry into force of this Agreement, access to its procurement market construction services (except for Japan and the Republic of Korea, which for suppliers of a non-Party beyond what it provides under this Agree- apply a 15 million SDR threshold); for government-owned enterprises, ment to suppliers of the other Party.� 400,000 SDRs for goods and services (except for the United States, which 23. There are studies that seek to forecast the increase in trade if gov- sets a US$250,000 threshold for federally owned utilities) and 5 million ernment procurement policies were liberalized. These forecasts, however, SDRs for construction services (except for Japan and Korea, which apply a are not the same as an estimate of the impact of the implementation of an threshold of 15 million SDRs). actual PTA with government procurement provisions. 14. By contrast, the exceptions to Article XXIII of the WTO GPA are more streamlined in that they cover security interests and social and phil- anthropic protection within one article. Bibliography 15. A variety of terminology, depending on the country or organiza- tion, exists to describe the mechanisms for resolving a complaint by a Anderson, Robert D., and Christopher R. Yukins. 2007. “Policy & Legal would-be contractor on the conduct of the procurement process prior to Frameworks for Opening the Chinese Procurement Markets.� GWU the decision on selection: domestic review mechanisms; protests by disap- Law School Public Law Research Paper 407, George Washington Uni- pointed bidders; complaints; or, simply, appeals. The WTO GPA refers to versity Law School, Washington, DC. “bid challenges,� and the UNCITRAL Model Law, to “bid review.� ———. 2008. “International Public Procurement Developments in 2008: 16. This paucity of information should not be surprising, given that, Public Procurement in a World Economic Crisis.� GWU Law School unlike the tariff provisions of PTAs, the behind-the-border provisions of Public Law Research Paper 458, George Washington University Law such accords are a much more recent policy innovation. School, Washington, DC. Government Procurement 385 Anderson, Robert D., Anna Caroline Müller, Kodjo Osei-Lah, Josefita The WTO Agreement on Government Procurement, ed. Bernard Pardo De Leon, and Philippe Pelletier. 2009. “Government Procure- M. Hoekman and Petros C. Mavroidis. Ann Arbor, MI: University of ment Provisions in Recent Regional Trade Agreements: Characteriza- Michigan Press. tion, Analysis, and Implications vis-à-vis The WTO Agreement on Dimitri, Nicola, Gustavo Piga, and Giancarlo Spagnolo, eds. 2006. Hand- Government Procurement.� World Trade Organization, Geneva. book of Procurement. Cambridge, U.K.: Cambridge University Press. Arrowsmith, Sue, and Arwel Davies, eds. 1998. Public Procurement: Global Evenett, Simon J., and Bernard M. Hoekman. 2005. “Government Revolution. The Hague: Kluwer Law International. Procurement: Market Access, Transparency, and Multilateral Trade Arrowsmith, Sue, and Martin Trybus, eds. 2003. Public Procurement: The Rules.� European Journal of Political Economy 21 (1, March): 163–83. Continuing Revolution. The Hague: Kluwer Law International. Hunja, Robert R. 2003. “Obstacles to Public Procurement Reform in Arrowsmith, Sue, John Linarelli, and Don Wallace, Jr. 2000. Regulating Developing Countries.� In Public Procurement: The Continuing Revo- Public Procurement: National and International Perspectives. The lution, ed. Sue Arrowsmith and Martin Trybus, 13–22. The Hague: Hague: Kluwer Law International. Kluwer Law International. Baldwin, Richard, and Patrick Low, eds. 2009. Multilateralizing Regional- Jones, D. S. 2007. “Public Procurement in South East Asia: Challenge and ism: Challenges for the Global Trading System. Cambridge, U.K.: Cam- Reform.� Journal of Public Procurement 7 (1): 3–33. bridge University Press. McAfee, R. Preston, and John McMillan. 1989. “Government Procure- Baldwin, Robert E. 1970. Nontariff Distortions of International Trade. ment and International Trade.� Journal of International Economics Washington, DC: Brookings Institution. 26 (3–4): 291–308. Baldwin, Robert E., and J. David Richardson. 1972. “Government Pur- OECD (Organisation for Economic Co-operation and Development). chasing Policies, Other NTBs, and the International Monetary Crisis.� 2001. The Size of Government Procurement Markets. Paris: OECD. In Obstacles to Trade in the Pacific Area: Proceedings of the Fourth Schooner, Steven L. 2002. “Desiderata: Objectives for a System of Govern- Pacific Trade and Development Conference, ed. H. Edward English and ment Contract Law.� Public Law and Legal Theory Working Paper 37, Keith A. J. Hay, Ottawa: Carleton School of International Affairs. George Washington University Law School, Washington, DC. Bourgeois, Jacques, Kamala Dawar, and Simon J. Evenett. 2007. “A Com- Schooner, Steven L., and Christopher R. Yukins. 2003. “Model Behaviour? parative Analysis of Selected Provisions in Free Trade Agreements.� Anecdotal Evidence of Tension between Evolving Commercial Public Prepared for the Directorate General for Trade, European Commis- Procurement Practices and Trade Policy.� International Trade Law and sion, Brussels. Regulation 9. SSRN: http://papers.ssrn.com/sol3/papers.cfm?abstract Campos, J. Edgardo, and Jose Luis Syquia. 2006. Managing the Politics of _id=346000. Reform: Overhauling the Legal Infrastructure of Public Procurement in Schooner, Steven L., Daniel I. Gordon, and Jessica L. Clark. “Public Pro- the Philippines. World Bank Working Paper 70. Washington, DC: curement Systems: Unpacking Stakeholder Aspirations and Expecta- World Bank. http://www-wds.worldbank.org/external/default/WDS tions.� George Washington University Law School, Washington, DC. ContentServer/WDSP/IB/2005/10/31/000090341_20051031141413/ SSRN: http://ssrn.com/abstract=1133234. Rendered/PDF/340440Man0Politic0Reform.pdf. Thai, Khi V., and Gustavo Piga, eds. 2007. Advancing Public Procure- Davey, William J., and André Sapir. 2009. “The Soft Drinks Case: The ment: Practices, Innovation and Knowledge Sharing. Boca Raton. FL: WTO and Regional Agreements.� World Trade Review 8 (1): 5–23. PRAcademics Press. http://journals.cambridge.org/action/displayFulltext?type=1&fid= WTO (World Trade Organization). 2000. “Synopsis of ‘Systemic’ Issues 4591316&jid=WTR&volumeId=8&issueId=01&aid=4591308. Related to Regional Trade Agreements Committee on Regional Trade Dawar, Kamala, and Simon J. Evenett. 2008. “Cariforum EPA and Agreements.� WT/REG/W/37, March 2, WTO, Geneva. Beyond. The Cariforum–EC EPA: An Analysis of Its Government Yukins, Christopher R. 2006. “A Case Study in Comparative Procurement Procurement and Competition Law–Related Provisions.� Working Law: Assessing UNCITRAL’s Lessons for U.S. Procurement.� Public Paper, Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ), Contract Law Journal 35 (3, Spring). Eschborn, Germany. Yukins, Christopher R., and Steven L. Schooner. 2007. “Incrementalism: Deltas, George, and Simon J. Evenett. 1997. “Quantitative Estimates of the Eroding the Impediments to a Global Public Procurement Market.� Effects of Preference Policies.� In Law and Policy in Public Purchasing: Georgetown Journal of International Law 38 (529). 18 Intellectual Property Rights Carsten Fink Rules for the protection of intellectual property rights level is often a divisive topic. Developed-country industry (IPRs) have become a common feature of preferential trade groups typically advocate the inclusion of a comprehensive agreements (PTAs). Historically, the North American Free IPR chapter in the PTAs signed by their governments. For Trade Agreement (NAFTA), which came into force in example, the U.S. Chamber of Commerce states on its Web 1994, was the first major trade agreement to include specific site, “We strongly advocate legislative and enforcement obligations on protection of patents, trademarks, copyright, actions by our top trading partners to bar the theft of intel- and other forms of IPRs. As PTAs have proliferated over the lectual property created and owned by U.S. artists, past two decades, so has the number of bilateral or regional researchers, and companies. The U.S. government should IPR rulebooks. Many PTAs, however, contain no obligations push our trading partners to uphold intellectual property for IPR protection. rights, and countries that fail to do so should be subject On the multilateral level, 1994 also saw the signing of to sanction under existing multilateral trade rules and the Agreement on Trade-Related Aspects of Intellectual bilateral accords� (emphasis added).1 In some developing Property Rights (TRIPS), at the close of the Uruguay countries, by contrast, PTA negotiations have been accom- Round of trade negotiations. The TRIPS Agreement is now panied by street protests alleging harmful consequences if about 15 years old, and the intellectual property field has the government were to sign on to higher levels of IPR undergone many changes. Governments have accordingly protections. felt a need to update and clarify certain IPR standards in This chapter presents an introduction to the treatment their international commercial relations, leading to the of IPRs in PTAs. It begins with a brief discussion of the introduction of IPR provisions in many PTAs. context in which negotiations on IPRs take place, empha- Another reason to embed IPR rules in a PTA is that these sizing the difference between traditional trade negotiations rules typically represent an important element in the overall on tariff liberalization and negotiations on IPR standards. package of quid pro quos necessary to strike a trade deal. For Subsequent sections summarize the main IPR provisions developed countries that host substantial IPR-producing found in PTAs, with a focus on the extent to which PTA industries, stronger protection of IPRs in foreign countries obligations go beyond the WTO’s TRIPS Agreement is usually an important market access concern. In exchange (“TRIPS+� provisions). The PTAs of the two main trading for preferential access to their markets for manufactured or blocs are surveyed first. U.S. agreements contain TRIPS+ agricultural goods, developed countries therefore press obligations for most forms of IPRs; agreements concluded their trading partners to commit to strong standards of IPR by the European Union (EU) are not equally comprehen- protection. From this point of view, it is not surprising to sive but do introduce new rules in certain areas. The treat- find IPR chapters more frequently in North-North and ment of IPRs in other PTAs is then briefly described. North-South PTAs than in South-South PTAs. In addition to self-standing IPR chapters in PTAs, intel- Just as the TRIPS Agreement is probably the most lectual property obligations are also found in bilateral or controversial agreement among the multilateral trading regional investment accords. Sometimes these investment rules of the World Trade Organization (WTO), the negoti- agreements take the form of another PTA chapter, and ation of new IPR standards at the bilateral and regional sometimes they are separate treaties. Regardless of their 387 388 Carsten Fink form, these accords are part of the legal framework govern- tariff liberalization, the same cannot be said about the ing bilateral or regional commercial relations and are adoption of ever-higher protection standards for IPRs. therefore briefly reviewed. Finally, the last section summa- Patents, copyright, and related intellectual property rights rizes the main messages emerging from the discussion. seek to give incentives for inventive and creative activities. Before proceeding, it is worth pointing out what this Owners of these forms of intellectual property benefit chapter does not cover: it offers neither a thorough from temporary market exclusivity, which allows them to introduction to the economics of different forms of IPRs generate rents above competitive returns and so recoup the nor a comprehensive review of the TRIPS Agreement. initial knowledge-generating investments. Governments, The annex to the chapter contains additional resources however, need to strike a proper balance between the inter- on these topics. ests of intellectual property holders and the public at large, which experiences market exclusivity as a barrier to the free dissemination of new knowledge. In practice, this Negotiating Considerations balance is reflected in, for example, the limited duration It is instructive to compare trade negotiations that involve of and the exceptions to the market exclusivity conferred the adoption of new IPR standards with more traditional by IPRs.3 trade negotiations focused on the exchange of tariff reduc- “Optimizing� the degree of intellectual property protec- tion commitments. In the latter case, country A promises tion is a challenging task. It is difficult to predict the future to lower its tariffs faced by exporters in B, and country B productivity of firms’ research and development activities promises to lower its tariffs faced by exporters in A. and the societal value of new technologies. In the absence Although this type of proposition seems only “fair,� it is not of reliable empirical guidance, existing patent and copy- grounded in pure economic logic. Economic theory mostly right regimes are typically an outcome of history, rules of makes a case for unilateral trade liberalization: countries thumb, and the influence of vested interests. There is thus are always better off if they eliminate import protection, no guarantee that the adoption of IPR rules in trade agree- regardless of what other countries do.2 Nonetheless, econo- ments will necessarily improve economic welfare. At some mists go along with the concept of reciprocity because gov- high level of protection, further strengthening is bound to ernments may find it politically easier to make a case for be welfare reducing. In addition, although the argument own-market opening if other countries also open their for free trade applies to all PTA partners, a case can be markets. Protectionist lobbies that may expect to lose from made for differential levels of IPR protection. For example, greater foreign competition may be less powerful if they are poorer developing countries often host few IPR-producing pitted against export-oriented constituents that stand to industries, and consumers and governments face tighter gain from enhanced market access abroad. budgets than in developed countries, suggesting that a At first sight, the logic for negotiating IPR rules in trade stronger emphasis on the free dissemination of technolo- agreements seems similar. For example, a developing gies than on strong incentives for innovation may be country with weak IPR protection may not find it politi- appropriate. cally feasible to strengthen its level of protection without The second major difference between negotiations on the support of exporters that benefit from enhanced tariffs and on IPR rules pertains to the preferential charac- commercial opportunities in a large developed-country ter of trade commitments. In the case of tariffs, it is the market. Similarly, the developed-country government parties to a PTA that benefit from liberalization commit- may find it difficult to overcome protectionist agricultural ments, whereas goods imported by a PTA member from a or industrial lobbies that fear competition from lower- nonmember continue to face the higher tariff prevailing wage developing-country producers without the support prior to the conclusion of the PTA (unless the nonmember of IPR-producing industries seeking stronger protection has its own preferential agreement with the member). In of their intellectual assets abroad. More generally, policy theory, the same could be done with the adoption of new makers may perceive the “inadequate� protection of intellectual property standards: the beneficiaries of these domestically generated IPRs in foreign countries as inher- standards could be confined to nationals of one of the ently “unfair,� undermining support for market opening PTA parties. For example, country A could say that only at home. intellectual property owners from its PTA partner B may Notwithstanding these parallels, there are two impor- benefit from the PTA’s intellectual property rules, whereas tant differences between traditional negotiations on tariffs intellectual property owners from nonmember country C and negotiations on IPRs. First, whereas most trade theo- are treated according to the previously prevailing stan- ries predict economic welfare gains as a result of reciprocal dards. Implementing such a discriminatory approach Intellectual Property Rights 389 would not be straightforward, as what constitutes nation- ing. In addition, in the typical grand bargain under a ality in the case of intellectual property ownership would North-South PTA, the developing country makes a non- have to be determined.4 Still, just as PTAs incorporate preferential commitment on IPRs in exchange for prefer- elaborate rules on the origin of traded goods, it would be ential market access to northern markets for agricultural possible to specify detailed nationality criteria for intellec- or manufactured goods. The latter, however, may be tem- tual property titles. porary because the value of trade preferences diminishes if Nevertheless, IPR rules in PTAs are almost always the northern PTA partner signs additional PTAs with third nondiscriminatory in character.5 The reason is that most countries or reduces tariffs on an MFN basis in the context PTA parties are also WTO members and are therefore of a multilateral trading round. bound by the rules of the TRIPS Agreement. Article 4 of These considerations are important and explain why TRIPS applies one of the WTO’s cornerstone principles— developing countries typically adopt a “defensive� negoti- most favored nation (MFN) treatment—to the protection ating stance in PTA negotiations on IPRs. Nevertheless, of intellectual property. Specifically, this article states, there are several reasons why developing countries, in the “With regard to the protection of intellectual property, end, are willing to go along with a comprehensive PTA that any advantage, favour, privilege or immunity granted by a entails new IPR rules. Member to the nationals of any other country shall be To begin with, governments may expect certain eco- accorded immediately and unconditionally to the nation- nomic benefits from stronger IPRs. Domestic IPR owners als of all other Members. . . .� Unlike the WTO’s agreements stand to benefit, even if they are still small in number. The on trade in goods and in services, the TRIPS Agreement possibility of attracting foreign direct investment (FDI) does not provide for an exception to the MFN principle that through stronger IPRs is also often considered an impor- permits preferential treatment emanating from a PTA. In tant benefit—although the empirical evidence on the other words, if a country were to extend superior treatment importance of IPRs in FDI decisions is mixed.6 to intellectual property holders in its PTA partners, other Even if the net welfare effect of signing on to a PTA WTO members could invoke the WTO’s dispute settlement intellectual property chapter is negative, the economic mechanism to request the extension of the special benefits benefits from the full package of PTA commitments may to its own nationals. be sufficiently large to lead to a positive overall welfare The fact that PTA concessions on IPRs are typically not effect. For example, it is conceivable that for a small devel- granted preferentially has important bargaining implica- oping country, the (discounted) welfare gains from obtain- tions. Suppose that country A is negotiating separate bilat- ing preferential market access to a large developed-country eral PTAs with countries B and C. In the case of goods market such as the United States or the EU will outweigh trade, it is possible to “sell� the same liberalization conces- any negative welfare effects in the IPR domain, even if pref- sion twice. For example, country A could offer a lower erential market access is temporary.7 At the same time, it is import tariff for sugar to country B in exchange for coun- important to recognize that certain segments of society try B’s reducing its import tariff on T-shirts. Simultane- would still be worse off with a PTA in such a scenario. ously, country A could offer the same reduction of its Although the presence of positive nationwide welfare import tariff for sugar to country C in exchange for coun- effects suggests that losers from the PTA could be compen- try C’s reducing its import tariff on toys. Compare this sated, the establishment of compensation mechanisms with a commitment to protect IPRs more stringently. If (e.g., tax reforms) is entirely a matter of domestic policy country A subscribes to stronger IPR rules in its PTA with and may face its own political-economy constraints—a country B, country C automatically benefits without need- dilemma well-recognized in the literature on the distribu- ing to offer any tariff concession to country A. From a bar- tional implications of trade reform. gaining perspective, the MFN requirement in the TRIPS The above considerations suggest that governments are Agreement therefore provides incentives to negotiate IPRs well advised to study carefully the potential economic at the multilateral level, where broader bargains involving benefits and costs associated with a new trade agreement. the whole WTO membership can be struck. Although it is an obvious step, carrying out a cost-benefit The above considerations raise a basic question: why analysis of the effects of a PTA with a diverse set of liberal- do developing countries still find it beneficial to sign PTAs ization commitments and legal undertakings—including that entail a substantial strengthening of the IPR regime? the adoption of new IPR standards—is a challenging task. Since most intellectual property titles in poorer nations Typically, the government agency conducting the trade are held by foreign residents, it is not clear that higher negotiations (e.g., the ministry of trade or of foreign standards of protection will necessarily be welfare enhanc- affairs) has only partial responsibility for—and only partial 390 Carsten Fink expertise in—the trade-related topics negotiated in a PTA. assessment models.8 Notwithstanding these challenges, In particular, the adoption of new intellectual property some analysis should be possible in most countries and can standards affects multiple sectors and requires cooperation usefully inform policy making before, during, and after the from other ministries, ranging from health and education negotiation of a PTA. to agriculture and industry. Internal coordination of prior analysis and the development of negotiating positions often face bureaucratic obstacles, especially in resource- TRIPS+ Provisions in U.S. PTAs constrained environments. Having considered the context in which PTA negotiations on Methodology and data availability pose special chal- IPRs take place, it is only natural to ask, where is the beef? lenges for assessing the economic effects of new IPR stan- What is the content of the “new� IPRs rules most commonly dards. The first need is for an understanding of the likely found in PTAs? We start with a review of U.S. PTAs (generally legislative changes that will be required as a consequence of called free trade agreements, or FTAs, in U.S. official usage) the conclusion of the PTA. For example, many developing because the IPR chapters negotiated by the United States are countries already have IPR laws that exceed the require- arguably the most comprehensive and far-reaching among ments of the TRIPS Agreement, and a PTA may just lock in the bilateral and regional IPRs rulebooks found worldwide. these existing standards. To be sure, such lock-ins have eco- Having concluded only two PTAs before the mid- nomic ramifications, because they reduce the freedom of 1990s—the agreement with Israel, and NAFTA—the U.S. governments to change their minds and amend laws at government embarked on a series of bilateral and regional some future time, but the economic effects of a lock-in are negotiations in the late 1990s. These “new generation� different from those of an actual legislative change. negotiations have so far led to the conclusion of 14 agree- Once the legislative landscape has been mapped, the ments, encompassing 20 trading partners; negotiations next step is to model the potential impact of the stronger with additional trading partners are pending. Table 18.1 IPR rules. When assessing the trade effects of tariff reduc- offers an overview of the current U.S. PTA landscape.9 tions, economists can draw on large datasets that provide The negotiation of strong IPR standards is a central information on actual and bound tariff rates, as well as objective in the U.S. agreements. Indeed, the U.S. Trade current and past trade flows. Partial and general equilib- Promotion Authority Act of 2002, under which most rium models for assessing policy changes are readily agreements have been negotiated, expressly states as a available from different sources. By contrast, data on the negotiating objective the promotion of intellectual prop- economic variables relevant for assessing the effects of IPR erty rules that “reflect a standard of protection similar to policy changes—for instance, prices, research and develop- that found in United States law.�10 The negotiating man- ment (R&D) investments, or license fees—are much harder date of the executive branch of the U.S. government was to come by, and economists cannot draw on ready-made refined in 2007, following the congressional elections in Table 18.1. U.S. Free Trade Agreement (FTA) Landscape FTAs signed but Negotiations launched, but no FTAs ratified by U.S. Congress not yet ratified agreement concluded Israel, 1985 Colombia Malaysia North American Free Trade Agreement (NAFTA), 1994 Korea, Rep. Thailand (Canada, Mexico) Panama Southern African Customs Union (SACU) Jordan, 2001 United Arab Emirates Chile, 2003 Trans-Pacific Strategic Economic Partnership Singapore, 2003 (Brunei Darussalam, Chile, New Zealand, Australia, 2004 Singapore) Morocco, 2004 Free Trade Area of the Americas (FTAA) Dominican Republic–Central America Free Trade Agreement (CAFTA–DR), 2005 (Costa Rica, Dominican Republic, El Salvador, Guatemala, Honduras, Nicaragua) Bahrain, 2006 Oman, 2006 Peru, 2007 Source: Office of the U.S. Trade Representative, www.ustr.gov. Intellectual Property Rights 391 November 2006. In what follows, we first discuss the main pharmaceutical producers. The TRIPS Agreement strikes provisions of U.S. PTAs that were signed between 2001 and a balance by requiring the protection of pharmaceutical 2006.11 We then describe the change in the negotiating patents and processes but allowing WTO member states to mandate and outline how it was reflected in the free trade override the market exclusivity of patents by issuing what agreement between the United States and Peru—the only are termed compulsory licenses. These are government agreement to have been ratified by the U.S. Congress since authorizations that allow generic pharmaceutical compa- the revisions of 2007. nies to produce a particular drug and sell it in competition with the patent holder, on payment of a license fee. Over the past decade, more than 10 developing countries have Agreements Ratified between 2001 and 2006 resorted to compulsory licensing—primarily, but not IPR chapters of U.S. PTAs include provisions on all types of exclusively, for HIV/AIDS drugs—with the aim of making intellectual property instruments and specify the mecha- medicines available at affordable prices to patients in need nisms available for enforcing exclusive rights. (This discus- (see Fink 2008). sion draws on Fink and Reichenmiller 2005.) A detailed U.S. PTAs contain a number of provisions that limit the review of all these provisions is beyond the scope of this ability of governments to introduce competition from chapter. Instead, we focus on some key obligations in generic producers. First, three agreements (those with which these agreements go beyond the protection stan- Australia, Jordan, and Singapore) limit the use of compul- dards found in the TRIPS Agreement. We do not analyze sory licensing to emergency situations, antitrust remedies, the extent to which these TRIPS+ obligations required (or and cases of public, noncommercial use. The TRIPS will require) domestic legislative changes. As noted above, Agreement does not contain any such limitation, merely U.S. PTA partners may have already protected IPRs more requiring that compulsory licenses be considered on their stringently than required by TRIPS, leaving the agreement individual merits.13 Thus, under TRIPS, a government to bind only existing policies under international law. may, for example, invoke the protection of public health as Table 18.2 outlines the key TRIPS+ provisions in U.S. a reason for issuing a compulsory license, without the need PTAs ratified between 2001 and 2006. The details vary to declare a national emergency. from agreement to agreement, but there are certain com- Second, to effectively make use of compulsory licenses, mon elements, as described next. generic drug manufacturers need to be able to obtain regula- Protection of patents and pharmaceutical test data. As tory permission to enter the market. Provisions in U.S. PTAs in TRIPS, all the agreements provide for a patent term of impose an obstacle in this respect. Except for the agreement 20 years. All except the agreement between the United with Jordan, all agreements require that regulatory agencies States and Jordan require, in addition, the extension of the prevent marketing approval of a generic drug during the patent term to compensate for delays caused by regulatory patent term without the consent of the patent holder; this is approval processes, such as approval for marketing a new sometimes referred to as “patent-registration linkage.� (The drug. Some agreements call for patent term extensions TRIPS Agreement does not impose any obligation of this when delays in the granting of the patent itself occur. kind.) In other words, even if a government is able to issue a Four U.S. agreements (those with Australia, Bahrain, compulsory license, generic producers may not be able to Morocco, and Oman) extend the scope of patentability by market their drugs, rendering the compulsory licensing mandating that patents be available for new uses of known instrument ineffective. products. All agreements go beyond TRIPS in enhancing Third, obtaining marketing approval for drugs requires patent protection for plants and animals. The strongest the submission to regulatory authorities of test data on a agreement in this regard is the one with Morocco, which drug’s safety and efficacy. Such data are protected by separate explicitly mandates the provision of patent protection for legal instruments that differ from country to country. The life forms. Others do not exempt plants or animals from TRIPS Agreement only requires that test data be pro- patentability, which is a flexibility provided for under tected against “unfair commercial use.� By contrast, most TRIPS. The weakest agreement is the one with the six U.S. PTAs explicitly mandate test data exclusivity, as pro- countries of the Central America Free Trade Agreement vided for under U.S. law. Once a company has submitted and the Dominican Republic (CAFTA–DR), which only original test data, no competing manufacturer is allowed calls for “reasonable efforts� to provide for patentability of to draw on these data for a period of five years to request plants.12 marketing approval for its own drug.14 The new compi- Much controversy has surrounded the provisions lation of comparable test data by competing manufac- in U.S. FTAs that strengthen the market exclusivity of turers may take several years and may be prohibitively Table 18.2. Principal TRIPS+ Provisions in U.S. Free Trade Agreements (FTAs) Ratified between 2001 and 2006 392 Provision U.S.–Jordan U.S.–Singapore U.S.–Chile U.S.–Morocco U.S.–Australia CAFTA–DR U.S.–Bahrain U.S.–Oman Protection of patents and pharmaceutical test data Patent term Extension given for Extension given for delays caused by regulatory approval process. In addition, extension given when a delay in the granting of the patent exceeds four delays caused by years from the filing of the application (five years for U.S.–Chile) or two years after a request for examination (three years for U.S.–Chile). regulatory approval process. Second-use patents No specific provision. Obligation to provide patents for new No specific Same as U.S.–Australia. uses of known products. provision. Patenting of life No general exclusion of plants and animals from patentability.a Explicit obligation Exclusions allowed “Reasonable Explicit obligation General forms to provide patent only for moral, efforts� have to to provide patent patentability protection for health, and be undertaken to protection for exclusion for plants and safety reasons. provide for plants, but animals, but not animals. patentability of animals can be for plants. plants. excluded. Compulsory Compulsory licenses limited to national TRIPS standards apply. Same as TRIPS standards apply. licenses emergencies, as antitrust remedy, and U.S.–Singapore. for public noncommercial use. Linkage between Patent owner must Marketing approval of a generic drug is prohibited during the patent term, unless authorized by the patent owner. In addition, the patent holder must patent status be notified when be notified of the identity of the generic company requesting marketing approval. and drug marketing marketing approval is approval sought during the patent term. Test data protection TRIPS standards Data exclusivity for Data exclusivity for Data exclusivity for Data exclusivity for five years. In addition, data exclusivity applies in all PTA member for apply. In five years. In five years. five years. countries, once first obtained in another territory. In the case of the U.S.–Bahrain pharmaceutical addition, length addition, where Additional three- and U.S.–Oman agreements, additional three-year data exclusivity is triggered by products of protection drug regulators year data “new clinical information� (with equivalent provisions on cross-border application). should be the rely on foreign exclusivity same as in the marketing triggered by originator’s approvals, “new clinical country. data exclusivity information.� applies automatically at home. Parallel imports of TRIPS standards Patent holders may TRIPS standards Patent holders may limit parallel imports TRIPS standards apply. patented apply. limit parallel apply. through licensing contracts. products imports of pharmaceutical products through licensing contracts. Side letters on No. Yes. No. Yes. public health? Copyright protection Membership in Only certain Yes. No. Yes. WIPO provisions apply. conventions Term of copyright Same as TRIPS. Life of author plus 70 years. If decided on a basis other than the life of the author, the term is 70 years (95 years in the U.S.–Oman agreement) from the protection publication or creation of the work. Technological “Adequate� “Adequate� protection against acts of circumvention. Ban on circumvention devices. Civil liability in case of willful infringement. Criminal liability in case protection protection and of willful infringement for commercial purposes. Exempted are nonprofit libraries, archives, and educational institutions, as well as acts related to measures “effective� reverse engineering, troubleshooting, protection of minors, computer or network security, and lawfully authorized government activities. remedies against acts of circumvention. Ban on circumvention devices. Liability of Internet No specific Limited liability of Internet service providers on the condition that they block infringing content upon notification by the copyright holder.b service provision. providers Burden of proof in Burden of proof placed on the defending party to show that works are in the public domain, but copyright owners still have to prove infringement. case of copyright infringement Parallel importation Copyright holder TRIPS standards apply. Copyright holder TRIPS standards apply. of copyrighted has right to block has right to block works parallel imports. parallel imports. Enforcement of intellectual property rights Institutional No specific Resource constraints cannot be invoked as an excuse for not No specific Same as U.S.–Chile. Same as U.S.–Singapore. flexibility in IPR provision. complying with specific enforcement obligations. provision. enforcement Border measures Scope of border Apply to imported, exported, and transiting goods. Apply only to Same as U.S.–Chile. measures not imported goods specifically (similar to defined. TRIPS). Civil and Obligation to fine infringers of copyright and trademark rights irrespective of the injury suffered by rights holders. administrative procedures Criminal procedures Scope of criminal Similar to TRIPS. In addition, criminal Similar to TRIPS. In addition, criminal procedures apply in cases of willful infringements, not only for financial and remedies procedures and procedures apply in case of willful gain, and specifically for knowingly trafficking in counterfeit labels affixed to certain copyrighted works remedies not infringements, not only for a financial (e.g., CDs, software). specifically gain. defined. Source: Adapted and updated from Fink and Reichenmiller 2005. Note: CAFTA–DR, Dominican Republic–Central America Free Trade Agreement; IPR, intellectual property right; TRIPS, Trade-Related Aspects of Intellectual Property Rights; WIPO, World Intellectual Property Organization. a. The U.S.–Chile agreement does not explicitly oblige protection of life forms under the patent system but mandates “reasonable efforts� to develop legislation related to patent protection for plants within four years from entry into force of the agreement. b. A side letter to the U.S.–Morocco agreement specifies the form in which notifications in case of alleged copyright infringement must be made. 393 394 Carsten Fink expensive, and it raises ethical concerns. U.S. FTAs do information. Drugs benefiting from this type of marketing not expressly provide for exceptions to test data exclusiv- exclusivity include not only new patented products but ity, especially when a government grants a compulsory also older generic products for which the patents have license. Exceptions may be justified through other provi- expired, although generic competition for previously sions in the IPR chapter (e.g., on anticompetitive prac- approved uses of such drugs would remain unaffected. tices), although such an interpretation remains legally Sometimes drug regulatory authorities recognize the uncertain (see box 18.1). Thus, test data exclusivity may marketing approval decisions of foreign regulators in pose another obstacle to governments’ effective use of granting marketing approval for the same product at home. compulsory licensing. The intellectual property chapter of the U.S.–Singapore Several U.S. FTAs go further on data exclusivity. When agreement mandates, in this case, that foreign data exclu- pharmaceutical companies seek marketing approval for sivity also applies domestically. In other words, no com- previously unapproved uses of already registered drugs, peting manufacturer is allowed to rely on the test data regulatory authorities typically require the submission of submitted to a foreign regulator in seeking marketing “new� clinical information. The agreements with Bahrain, approval in a PTA member. Morocco, and Oman provide for an additional three-year The agreements with Australia, Bahrain, CAFTA–DR, data exclusivity period triggered by such new clinical and Oman are even more far-reaching with respect to the Box 18.1. Patent-Registration Linkage and Test Data Protection: The Case of Chile The free trade agreement (FTA) between the United States and Chile mandates patent-registration linkage and exclusive rights to pharmaceutical test data (see the summary in table 18.2). Chile has implemented these obligations in a way that has sought to protect public health, promote coherence with patent rules, prevent anticompetitive behavior, and avoid imposing an undue burden on Chile’s public health authority, the Institute of Public Health (Instituto de Salud Pública de Chile). Specifically, Articles 89–91 of Chile’s Industrial Property Law provide for the five-year exclusivity period for pharmaceutical test data stipulated in the FTA but also create several exceptions to the acquisition and maintenance of exclusive rights, as follows: • If the owner of the test data has engaged in conduct deemed contrary to free competition • For reasons of public health, national security, public noncommercial use, national emergency, or other circumstances of extreme urgency • If the pharmaceutical product is the subject of a compulsory license • If the pharmaceutical product has not been marketed in Chilean territory within 12 months from the registration or health authorization issued in Chile • If the application for registration or health authorization of the pharmaceutical product is filed in Chile 12 months or more after the same product was first registered or authorized abroad. With respect to the patent-registration requirement, Resolution 5572 obligates the Institute of Public Health to publish on its website all applications for health registration. In addition, companies that already possess a health registration can request that all new applications submitted to the institute be sent to them by electronic mail. Patent holders can thus monitor incoming applications and, if they believe that a pharmaceutical product for which health registration is sought infringes on a patent, can request an injunction from a Chilean court to prevent the institute from issuing the registration. The U.S. government has not been satisfied with Chile’s implementation of its obligations under the agreement. In 2006 it launched an “out-of-cycle� review of the protection of IPRs in Chile that was sparked by complaints by the Pharmaceutical Research and Manufacturers of America (PhRMA), a U.S. pharmaceutical industry association. Specifically, PhRMA contends that exceptions to test data exclusivity are not allowed under the terms of the FTA and that Chile has not instituted a mechanism whereby the Chilean authorities by themselves prevent the granting of marketing approval of a patent-infringing pharmaceutical product. As a result, PhRMA claims, several copies of drugs protected by patents have obtained health registration from the Institute of Public Health. The Office of the U.S. Trade Representative placed Chile on the U.S. priority watch list in 2007 and 2008 and has raised its concerns in bilateral discussions with the Chilean government, but it has not explicitly accused Chile of violating its FTA obligations, nor has it invoked the FTA’s formal dispute resolution mechanism. It is uncertain how a dispute panel would rule on Chile’s compliance with its FTA obligations. A likely key question would be whether the exceptions to test data rules can be justified by the FTA articles on anticompetitive behavior, by the references in the FTA to the Doha Declaration on Trade-Related Aspects of Intellectual Property Rights (TRIPS) and Public Health, or by other provisions. Another question would probably be whether the issuance of a health authorization by the Institute of Public Health constitutes an act of “marketing approval,� as conceptualized in the FTA’s patent-registration linkage requirement. Sources: Chile, Industrial Property Law (Law 19.039); Resolution 5572 of the Chilean Institute of Public Health; various articles published by Inside U.S. Trade; “Special 301� submission by PhRMA, available at http://members.phrma.org/international/PhRMA_2005_Special_301.pdf; “Special 301� reports, Office of the U.S. Trade Representative, http://www.ustr.gov. Intellectual Property Rights 395 cross-border application of data exclusivity. Even if regula- compulsory licensing and parallel importation to “pro- tory authorities do not recognize foreign marketing mote access to medicines for all.�17 In August 2003, WTO approvals, competing manufacturers are prevented from members created a special mechanism under the TRIPS using test data submitted to a drug regulatory agency in Agreement that allows countries with insufficient manu- another territory for five years, starting from the date of facturing capacity to effectively use compulsory licenses by submission in the foreign country. That is, test data exclu- importing generic drugs (see Fink 2005). sivity applies automatically in all PTA jurisdictions, once Technically, the Doha Declaration and the August 2003 a company submits test data to a drug regulator in one decision by WTO members do not address questions of territory—even outside the PTA area. Furthermore, phar- marketing approval during the patent term or of test data maceutical companies can wait for up to five years to seek exclusivity. However, the provisions of U.S. PTAs on these marketing approval in a PTA member and still benefit matters can still be seen as being at odds with the spirit of from the full five-year exclusivity term in that member. the multilateral accords, to the extent that they preclude The provisions on test data exclusivity are especially the effective use of compulsory licenses. important for small countries. Pharmaceutical companies In side letters to the U.S. agreements with Bahrain, may not seek out patents in every jurisdiction for every ini- CAFTA–DR, Morocco, and Oman, the respective govern- tially promising pharmaceutical compound. The registra- ments shared understandings that the intellectual property tion of patents can be costly and at the stage of patenting, it chapters “do not affect� their ability to “take necessary is uncertain whether the pharmaceutical compound in measures to protect public health by promoting medicines question will turn into a commercial product or how suc- for all. . . .�18 The U.S. government, however, does not view cessful it will be. The incentive to seek out patents in small the side letters as creating any kind of exemption that countries is therefore less. However, even in the absence of would allow parties to the FTAs to ignore obligations in the patent exclusivity, pharmaceutical companies can still ben- agreements’ intellectual property chapters (Fink and efit from five years of market exclusivity when they intro- Reichenmiller 2005). In the U.S. view, the side letters duce a new product because of the rights these companies merely signal the parties’ belief that the intellectual prop- hold to the product’s clinical test data. erty rules of the PTAs will not interfere with the protection Finally, whether the parallel importation of pharmaceu- of public health.19 tical products that have been placed on the market in for- Copyright protection. Most U.S. PTAs require accession eign jurisdictions is permitted is another issue covered by to the World Intellectual Property Organization (WIPO) PTA rule books on IPRs. Parallel importation can be a Copyright Treaty and the WIPO Performances and Phono- means of exerting downward pressure on pharmaceutical grams Treaty. These treaties were concluded in 1996 after prices, if products are sold more cheaply abroad. The the end of the Uruguay Round and are therefore not incor- TRIPS Agreement affords WTO members flexibility in porated in the TRIPS Agreement. They were negotiated to determining whether to permit parallel importation of adapt copyright protection to advances in information and patented drugs.15 By contrast, the U.S. agreements with communication technologies that greatly facilitated the Australia, Morocco, and Singapore allow patent holders to distribution of literary or artistic works in digital form. prevent parallel importation of goods.16 The provisions of the two WIPO treaties go beyond TRIPS The TRIPS+ provisions in U.S. PTAs have attracted in several ways—for example, by making it clear that stor- much criticism from nongovernmental organizations age of protected works in digital form is a reproduction (NGOs) concerned with affordable access to medicines. that can be controlled by copyright owners and by confer- They argue that stronger market exclusivity would curtail ring certain moral rights on performers. (See Abbott, competition from generic producers and would ultimately Cottier, and Gurry 2007 for a detailed discussion of the lead to high drug prices that are out of reach for poor treaties.) population segments (MSF 2004; Oxfam 2006). They also In addition to incorporating the two WIPO treaties, allege that IPR provisions in PTAs are at odds with efforts U.S. PTAs set new substantive protection standards. First, at the WTO to ensure that IPR rules are consistent with except for the agreement with Jordan, they extend the public health objectives. At their 2001 ministerial meeting TRIPS copyright term of life of author plus 50 years by an in Doha, WTO members issued a declaration recognizing additional 20 years, reflecting U.S. practice. the gravity of the public health problems afflicting many Second, most U.S. bilateral PTAs include obligations developing and least-developing countries. Among other to prevent circumvention of technological protection things, the declaration reaffirmed the right of WTO measures—devices and software developed to prevent members to use the flexibilities of TRIPS in the area of unauthorized copying of digital works. This issue is not 396 Carsten Fink covered under TRIPS, and the provisions on the subject in that would permit the circumvention of technological pro- the two WIPO treaties are fairly general. The U.S. Digital tection measures if such action does not result in an Millennium Copyright Act of 1998 set more far-reaching infringement of a copyrighted work.20 Ensuring fair use of standards on circumventing technologies that are designed copyrighted material seems particularly important for to prevent unauthorized copying of digital content. Those accessing educational material. The opportunities and standards have found their way, to varying degrees, into gains from the use of digital libraries, Internet-based dis- seven of the bilateral agreements. Related provisions in six tance learning programs, and online databases would be of the PTAs define the liability of Internet service limited if access to such tools were unaffordable or other- providers when infringing content is distributed through wise restricted by copyright law. their servers and networks. Again, these provisions are Enforcement of IPRs. The TRIPS Agreement—for the based on standards found in the U.S. Digital Millennium first time in an international agreement on intellectual Copyright Act. property—introduced detailed obligations on the enforce- Third, all of the bilateral PTAs place the burden of proof ment of IPRs. Certainly, without judicial enforcement of in copyright infringement cases—that is, of showing that intellectual property laws, rules on patents, copyright, and works are in the public domain—on the defending party. other forms of protection could be seriously undermined. TRIPS does not impose any such obligation. The FTAs thus However, the agreement, recognizing the institutional lim- strengthen the position of copyright holders, as artistic and itations in many developing countries, does not create any literary works are generally considered protected unless obligation “with respect to the distribution of resources as they obviously belong in the public domain. between enforcement of intellectual property rights and Finally, as in the case of pharmaceutical products, the enforcement of law in general.�21 Such a caveat may be TRIPS does not mandate any rule on the permissibility of important. The enforcement of intellectual property rights parallel imports of copyrighted works, such as books or can be a costly exercise in terms of both budgetary outlays musical CDs, that have been lawfully sold in foreign mar- and the employment of skilled personnel. For developing kets. Some countries, such as New Zealand, permit parallel countries that face many institutional deficiencies, there is importation of certain copyrighted products as a way of a risk that stronger enforcement of IPRs would draw away stimulating price competition. By contrast, the U.S. bilat- financial and human resources from other development eral agreements with Jordan and Morocco give copyright priorities. holders the right to block parallel importation. The U.S. agreements with Australia and Jordan do not From an economic perspective, the welfare conse- explicitly allow for the same institutional flexibility. In quences of new and stronger copyright protection stan- these cases, it may be difficult to defend derogations from dards are ambiguous. Most countries have industries that the specific enforcement provisions of the agreements’ IPR rely on copyright protection and that may benefit from chapters on the grounds of inherent institutional con- strengthened protection. Furthermore, new technologies straints, such as limited budgetary or human resources. The that make copying of digital works much easier pose chal- agreements with Bahrain, CAFTA–DR, Chile, Morocco, lenges which policy makers must address. At the same Oman, and Singapore go further, spelling out that resource time, copyright laws have historically sought to strike a bal- constraints cannot be invoked as an excuse for failure to ance between the interests of copyright producers and comply with the agreements’ specific enforcement obliga- those of the general public; for example, so-called fair use tions.22 Indeed, some of these obligations seem to create exemptions allow the copying of protected works for edu- additional institutional requirements. For example, as in cational or research purposes. There are concerns that new the case of TRIPS, the PTAs require customs authorities to rules with respect to the term of protection, technological block trade in counterfeit and pirated goods. But TRIPS protection measures, the liability of Internet services requires such measures only for imported goods, whereas providers, and the burden of proof in case of copyright most PTAs mandate the application of border measures to infringement could diminish the rights of consumers and imported goods, goods destined for export, and even, in the general public (CIPR 2002). some cases, transiting goods. Such concerns have also been voiced in the United Finally, the enforcement rules of the bilateral agree- States itself, not only by consumer rights advocates and ments mandate a stronger deterrent against IPR infringe- academic institutions but also by computer manufacturers ment. For example, TRIPS requires only the imposition of and communications service providers that distribute fines adequate to compensate IPR holders for the mone- copyrighted works. For example, specific amendments to tary damages they suffered. In the case of copyright piracy the Digital Millennium Copyright Act have been proposed and trademark counterfeiting, all the PTAs require the Intellectual Property Rights 397 imposition of fines irrespective of the injury suffered by yet ratified by Congress. In May 2007, these were the free IPR holders. TRIPS mandates criminal procedures only in trade agreements with Colombia, the Republic of Korea, cases of willful trademark counterfeiting or copyright Panama, and Peru. In the case of Korea, continued dis- piracy on a commercial scale. Many PTAs go beyond this agreement between the Congress and the administration broad standard and define more explicitly the scope of acts on automotive trade led to the exclusion of this FTA of infringement subject to criminal procedures—including, from the Bipartisan Agreement.25 In the case of the other for example, copyright piracy with a significant aggregate three agreements, USTR initiated the renegotiation of monetary value that is not necessarily undertaken for the FTA texts with the concerned governments. Only the financial gain. Thus, certain forms of end-user piracy may revised U.S.–Peru FTA has so far gained congressional be considered a criminal offense. approval. Examination of the revised U.S.–Peru text reveals that the elements of the Bipartisan Agreement outlined above The May 2007 Bipartisan Agreement and have found their way into the final IPR chapter: patent the U.S.–Peru FTA term extension is mentioned as a voluntary option, as is As pointed out above, the adoption of TRIPS+ standards in patent-registration linkage, and only certain enforcement U.S. PTAs has received much criticism from NGOs, partic- mechanisms need to be made available to the patent ularly in the area of pharmaceuticals. Over time, these holder. Whereas the FTA provides for five years of pharma- standards have become controversial within the United ceutical test data exclusivity, a special exception to exclusive States, as well as abroad. The U.S. Congress has for many rights is created for measures to protect public health in years considered legislation that would authorize the paral- accordance with WTO rules.26 lel importation of medicines from Canada and other coun- Other TRIPS+ elements of the FTA with Peru were not tries. Concerned that such legislation would violate obliga- affected by the Bipartisan Agreement. These follow tions under PTAs, the Congress inserted language in a 2005 closely the provisions found in the CAFTA–DR agree- appropriations bill that effectively prohibits the Office of ment (see table 18.2). An important exception is that the U.S. Trade Representative (USTR) from negotiating the U.S.–Peru FTA does not feature the provisions on the PTA provisions that would block parallel imports of cross-border application of test data exclusivity described patented pharmaceuticals (see Abbott 2006). above. More fundamentally, with Democrats winning both As a final note, although the immediate impact of the houses of Congress in the November 2006 elections, the bipartisan trade deal has been limited to the FTAs nego- Bush administration and the congressional leadership tiated with Colombia, Panama, and Peru (with the started to negotiate a new trade framework. The resulting former two still awaiting congressional approval), the May 2007 Bipartisan Agreement led to a de facto revision deal marks an important shift in U.S. trade policy toward of the USTR’s negotiating mandate. In the area of IPRs, the more sensitivity to public health concerns in global agreement sets out flexibilities that roll back some of the IPR rules. TRIPS+ provisions outlined above.23 Specifically, it turns the obligation to grant patent term extension for delays in TRIPS+ Provisions in EU PTAs obtaining marketing authorization into a voluntary option that governments “may� choose to adopt. Similarly, drug The European Union’s drive toward concluding PTAs is regulators would no longer be required to deny marketing more recent than that of the United States. In the second approval on the basis of a drug’s patent status. Govern- half of the 1990s, the EU launched PTA negotiations with a ments would only be required to make available certain variety of trading partners, which led to the conclusion of mechanisms to allow patent holders to effectively enforce agreements with Chile, Mexico, and South Africa. Starting their rights.24 Crucially, the agreement creates an express in 1999, however, the EU practiced a de facto moratorium exception to the test data exclusivity rules for measures to on new negotiations. This moratorium was not a formal protect public health “in accordance with the Doha Decla- policy, but it reflected the consensus of the EU member ration and subsequent protocols for its implementation.� states and the European Commission that priority should In other words, it removes any legal uncertainty about be given to the conclusion of a comprehensive multilateral countries’ ability to make effective use of compulsory trading round. The one exception to the moratorium was licensing. the negotiation of economic partnership agreements The Bipartisan Agreement applied only to those U.S. (EPAs) with countries belonging to the Africa-Caribbean- PTAs that had been signed by the administration but not Pacific (ACP) group of former European colonies. In this 398 Carsten Fink case, the negotiation of reciprocal trade agreements was agenda has become substantially more ambitious. In 2004 driven by the expiry of trade preferences granted by the EU the European Commission issued a Strategy for the under the Cotonou Agreement and deemed incompatible Enforcement of Intellectual Property Rights in Third with WTO rules. So far, the EU has concluded a full EPA Countries that calls for “a long-term line of action� to with the Caribbean Forum of ACP States (CARIFORUM) strengthen the protection of IPRs outside EU territory. The and several interim EPAs with other ACP partners. strategy expressly calls for revisiting the approach to the Despite the early setbacks in the negotiations under the IPR chapters of bilateral agreements.28 Doha Development Agenda (DDA), the EU maintained its moratorium until about 2006, when it announced the “Early Generation� Agreements with Chile, negotiation of a PTA with Central America. Since then, Mexico, and South Africa negotiations with a considerable number of additional trading partners have been launched. Among other factors, The IPR sections in the EU agreements with Chile, Mexico, this shift in the EU’s trade policy strategy was driven by and South Africa are short, hardly taking up the space of a two developments. First, the prospects of an ambitious single page. In the case of the Mexico and South Africa and comprehensive Doha agreement became increas- agreements, the provisions are nonbinding and general. ingly remote, especially with respect to some of the EU’s They emphasize the importance of adequate intellectual priority trade concerns—trade facilitation, competition property protection and call for the establishment of con- policy, investment rules, and geographical indications sultation mechanisms should “difficulties� arise in the pro- (GIs). Second, PTAs involving major economies were pro- tection of intellectual property. In the Chile agreement, the liferating rapidly in all regions, putting increased pressure IPR provisions are more ambitious. The agreements incor- on the EU to not lose out on commercial opportunities porate several WIPO conventions, including the WIPO abroad and to negotiate its own set of preferential treaties. Copyright Treaty and the WIPO Performances and Phono- (For a more in-depth discussion of the shift in EU trade grams Treaty. As explained in the previous section, these policy, see Woolcock 2007.) Table 18.3 offers an overview conventions go beyond the copyright rules of the TRIPS of the current EU PTA landscape.27 Agreement, and their incorporation into a PTA therefore Turning to the IPR dimension of EU agreements, the constitutes a TRIPS+ element. (Interestingly, the two “early generation� PTAs with Chile, Mexico, and South WIPO Conventions are not incorporated into the Africa focused primarily on establishing strong and com- U.S.–Chile FTA, as is shown in table 18.2.) prehensive rules for GI protection. For the “new genera- The most far-reaching TRIPS+ provisions are found in tion� PTAs that are currently being negotiated, the EU’s separate agreements on wines and spirits that strengthen Table 18.3. European Union (EU) Preferential Trade Agreements (PTAs) and Economic Partnership Agreements (EPAs) Concluded PTAs and EPAs Provisional EPAs PTAs and EPAs under negotiation Mexico, 2000 West Africa, 2007 (Côte d’Ivoire, Ghana) Southern Cone Common Market South Africa, 2000 Central Africa, 2007 (Cameroon) (Mercosur, Mercado Común del Sur) (Argentina, Brazil, Paraguay, Uruguay) Chile, 2003 Eastern African Community (EAC), 2007 (Burundi, Kenya, Rwanda, Tanzania, Central America (Costa Rica, El Salvador, Caribbean Forum of African, Caribbean, Uganda) Guatemala, Honduras, Nicaragua) and Pacific (ACP) States (CARIFORUM), 2008 (Antigua and Barbuda, Bahamas, Eastern and Southern Africa, 2007 India Barbados, Belize, Dominica, Dominican (Comoros, Madagascar, Mauritius, Association of Southeast Asian Nations Republic, Grenada, Guyana, Haiti, Seychelles, Zimbabwe) (ASEAN) (Brunei Darussalam, Cambodia, Jamaica, St. Christopher and Nevis, Southern African Development Indonesia, Lao PDR, Malaysia, Myanmar, St. Lucia, St. Vincent and the Grenadines, Community (SADC), 2007 Philippines, Singapore, Thailand, Suriname, Trinidad and Tobago) (Botswana, Lesotho, Mozambique, Vietnam) Korea, Rep., 2010 Namibia, Swaziland) Andean Community (Bolivia, Colombia, Pacific, 2007 (Fiji, Papua New Guinea) Ecuador, Peru) Pending EPA negotiations with countries in West Africa, Central Africa, EAC, Eastern and Southern Africa, SADC, and the Pacific Source: Woolcock 2007; European Commission data, http://ec.europa.eu/trade/. Note: This table does not include the EU’s association agreements with Eastern European countries and Euro-Mediterranean partners. Intellectual Property Rights 399 GI protection. In the case of the PTA with Chile, the provi- regional heterogeneity in agricultural production, affords sions on wines and spirits are separate annexes. The EU GI owners stronger exclusive rights than in the United negotiated self-standing agreements on wines (with South States. In addition, EU laws protect “traditional� expres- Africa only) and on spirits (with Mexico and South Africa), sions that describe certain methods of production, such as outside the formal PTA framework. These special agree- “ruby� for port from Portugal. ments were, however, negotiated in parallel with the Unfortunately, little systematic evidence exists to assess broader trade agreements and may have formed part of the whether the more lenient approach toward GIs in the overall package of legal obligations undertaken by the par- Americas has led to confusion on the part of consumers. ties. It is worth pointing out that the EU has negotiated What also complicates an economic assessment of different agreements on wines or spirits (or both) with three devel- protection regimes is that certain products known by GIs oped countries—Australia, Canada, and the United are likely to have significant status value. Economists define States—with which it has not entered into bilateral PTAs. “status goods� as products for which the mere display of a The key substantive element of the EU agreements on particular label confers prestige on their buyers, regardless wines and spirits is the establishment of lists of geographic of the product’s quality. Champagne, Beluga caviar, and names to which signatories must apply the “highest� level Kobe beef may fall into this category. Since use of these of GI protection. This means that nonoriginal producers in GIs by nonoriginal producers may undermine the status the PTA parties are not allowed to use a listed name even if value attached to the original varieties, the price premiums they display the true origin of the good and even if use of of the original producers may suffer, even if consumers are the name is accompanied by expressions such as “kind,� not confused about the true origin of their purchases. “type,� or “style� (e.g., “Champagne-style sparkling wine�). Although the EU’s agreements on wines and spirits are The lists of geographic names are often extensive; the clearly driven by the interests of EU producers of specialty EU–South Africa agreement on wine alone is 120 pages foods, certain producers in EU partner countries may also long. Several of the European GIs listed in the agreements, benefit. For example, the EU–Mexico agreement on spirits especially for spirits, were being used by producers in EU establishes GI protection for the Tequila and Mezcal desig- PTA partner countries, and this use had to be (or will have nations. In addition, the disuse of semigeneric expressions to be) phased out. In most cases, use of the relevant GIs may lead only to a temporary adjustment; evidence from would likely have been among the exceptions in the TRIPS Australia’s wine industry suggests that the rebranding of Agreement.29 The phase-out of these generic names may the affected products and associated marketing efforts have affected exports of third countries to EU partner provided a boost for local producers.30 countries, raising the possibility that the adoption of a GI list leads to a de facto trade preference. “New Generation� EU Agreements In principle, GIs, like trademarks, are important for reducing information asymmetries in markets for “experi- A first impression of the EU’s revised approach to the treat- ence goods�—goods whose quality cannot be easily ment of IPRs in its trade agreements can be gained by observed by simple inspection at the time of purchase. looking at the EPA with CARIFORUM.31 The intellectual From the perspective of consumers, these indications can property chapter of the agreement is comprehensive, cover- lower search costs, particularly for heterogeneous products ing most forms of IPRs, as well as measures for their enforce- such as wines or cheeses. For producers, GIs offer a means ment. As in the EU–Chile PTA, the EU–CARIFORUM of attaching a reputation for quality to a place name that agreement incorporates the WIPO Copyright Treaty and may then be marketed and used on labels. A number of the WIPO Performances and Phonograms Treaty.32 The empirical studies have quantified substantial price premi- EPA also establishes TRIPS+ standards for GIs, requiring ums associated with certain products labeled with pro- CARIFORUM members to establish a dedicated system for tected GIs (see Fink and Maskus 2005 for a review). their protection. This obligation is TRIPS+ in the sense More controversial is the precise scope of exclusive that the TRIPS Agreement allows the implementation of its rights that should be conferred on rights holders. In the GI standards through the trademark system. More sub- United States, Mexico, Chile, and other “New World� coun- stantially, the EPA requires a high level of GI protection for tries, nonoriginal producers have traditionally been all products, preventing nonoriginal producers from using allowed to use GIs for certain products as long as the true a GI even if the true origin of the good is indicated or the geographic origin of the products is made clear. This has name is accompanied by expressions such as “kind,� “style,� given rise to “semigeneric� expressions such as “American- or “imitation.� In the TRIPS Agreement, this high-level made Pecorino cheese.� The EU, with its long tradition of protection standard is limited to wines and spirits. The EU 400 Carsten Fink has been among the leading proponents at the WTO for Similarly, leaked information on the EU’s proposal for the extending strong protection to all products but has PTA with the Andean Community suggests the inclusion encountered resistance from other WTO members, notably on the EU’s wish list of test data protection, as well as a producer countries in the Americas. Discussions on this broadening of criminal sanctions and border measures to topic in the Doha Development Agenda have to date not infringements of all types of IPRs (see Hernández 2009). led to any outcome.33 Of course, it is uncertain how far the EU’s proposals will Extensive new IPR obligations are found in the enforce- be reflected in the final PTA texts. A definite assessment of ment part of the IPR chapter, reflecting the emphasis on the treatment of IPRs in the “new generation� PTAs will strengthened enforcement outlined in the 2004 strategy have to await the conclusion of the latest set of negotiations. document described above. Interestingly, the EPA enforce- ment obligations are rooted in the EU’s internal approach TRIPS+ Provisions in Other PTAs toward upholding IPRs, notably Directive 2004/48, on the enforcement of intellectual property rights, and two regu- In addition to the agreements negotiated by the United lations dealing with border measures.34 In fact, certain States and the EU, numerous PTAs contain provisions on enforcement provisions of the EU–CARIFORUM EPA the protection of intellectual property. For the most part, have been copied word for word from the 2004 Enforce- these are of a general nature, merely calling for the ade- ment Directive.35 quate and effective protection of IPRs. Sometimes they To summarize, the EPA clarifies and expands the TRIPS incorporate certain provisions of the TRIPS Agreement provisions on measures for the preservation of evidence, into the PTA. Although such incorporation may not be the collection of information on the origin and destination seen as TRIPS+, it is still significant, as it may subject of IPR-infringing products, injunctions, and several other TRIPS disciplines to the dispute settlement mechanisms of matters. Probably the most significant TRIPS+ element is bilateral PTAs.38 the extension of border measures to the importation, Several FTAs that do not involve the United States or the exportation, reexportation, and other border movements EU set TRIPS+ protection standards. Most prominently, of goods. In addition, border measures are to apply not the European Free Trade Association (EFTA), with Iceland, only to counterfeit trademark goods and pirated copyright Liechtenstein, Norway, and Switzerland as members, has goods (as in the TRIPS Agreement) but also to goods negotiated a considerable number of PTAs that feature cer- infringing on a design right or a GI. Like many of the U.S. tain TRIPS+ elements. Specifically, the EFTA agreements FTAs, the EU–CARIFORUM EPA does not recognize the with Chile, Colombia, Korea, Lebanon, and Singapore institutional limitations of developing countries, as the incorporate the WIPO Copyright Treaty and the WIPO TRIPS Agreement does. Thus, CARIFORUM members Performances and Phonograms Treaty. The agreements may not be able to invoke resource constraints as a rea- with Chile, Korea, and Singapore also provide for an exten- son for not complying with the agreements’ specific sion of the patent term as compensation for unreasonable enforcement obligations. As pointed out above, this lack delays in regulatory approval processes, along the lines of of institutional flexibility may be a concern, as it is not some of the U.S. agreements. clear that increased government spending on the enforce- The EFTA’s agreements with Chile, Korea, Singapore, ment of IPRs is warranted in environments in which many and Tunisia require pharmaceutical test data exclusivity, public goods are underprovided.36 again following the U.S. approach. In the Korean case, the The EU–CARIFORUM EPA provides a window into the relevant provision also allows for an alternative to exclusive EU’s revised approach to the treatment of IPRs. Future test data rights, whereby any company would be allowed to PTA chapters on IPRs may well be more ambitious, not rely on test data submitted to a regulatory agency if the least because some of the EU’s prospective PTA partners company first submitting these data is adequately compen- are more developed than the Caribbean countries. Early sated. Such a compensatory liability approach has been indications point in this direction. The leaked EU proposal advocated by academic scholars as a superior way of pro- for the IPR chapter in the PTA between the EU and the tecting test data because it allows research-based companies Association of Southeast Asian Nations (ASEAN) shows to recoup investments in clinical trials without inhibiting more far-reaching TRIPS+ obligations, including an exten- market competition (Reichman 2004). Interestingly, the sion of the copyright term, pharmaceutical test data exclu- EFTA–Colombia agreement, signed in 2008, incorporates sivity, rules on the liability of Internet service providers, an exception to exclusive test data rights to protect public and an extension of border measures to all forms of IPRs.37 health in accordance with WTO rules. The language used Intellectual Property Rights 401 in this PTA is identical to that in the revised U.S.–Colombia mandated in Mexico’s PTAs with Nicaragua and Uruguay agreement, illustrating how the U.S. policy shift described and in the Northern Triangle PTA involving El Salvador, above influenced norms in a treaty that does not involve Guatemala, and Honduras. the United States. A second remarkable feature of the EFTA–Colombia IPRs and Bilateral Investment Rules PTA is its incorporation of a mandatory disclosure require- ment in patent applications of the origin or source of In addition to the rules contained in the intellectual genetic material. Several developing-country WTO mem- property chapters of PTAs, IPRs are subject to separate bers have argued that the TRIPS Agreement affords inade- bilateral investment disciplines. Many PTAs contain sep- quate safeguards against “biopiracy�—the acquisition of arate investment chapters in which IPRs are expressly patent rights for biological materials (and related tradi- listed in the definition of what constitutes an investment. tional knowledge) that are simply taken from one country’s This is the case in the U.S. agreements with Australia, biological resources, without inventive effort. These mem- CAFTA–DR, Chile, Morocco, Oman, and Singapore. bers have long advocated the inclusion of a disclosure Moreover, parties to some PTAs that do not feature a sep- requirement in the TRIPS Agreement, but such a move has arate investment chapter have concluded self-standing been resisted by other WTO members, and discussions on bilateral investment treaties (BITs) that include IPRs in the topic in the context of the Doha negotiations have not the definition of investment; examples are the U.S. agree- yet led to any compromise.39 It is interesting that Colom- ments with Bahrain and with Jordan. There are also bia, as one of the leading proponents of a disclosure numerous BITs between countries that have not entered requirement, persuaded EFTA countries to insert such a into a PTA. Since no multilateral agreement on investment clause in their PTA, showing that the negotiation of IPR exists at the WTO or elsewhere, these bilateral investment chapters can advance the interests of developing countries. rules break new ground. In February 2009, Switzerland signed a comprehensive A broad definition of investment that encompasses IPRs economic partnership agreement with Japan that incorpo- makes government measures affecting the intellectual rates the TRIPS+ protection standards found in the EFTA property portfolios of foreign investors subject to the agreements described above: incorporation of WIPO investment disciplines of PTAs or BITs. This raises such treaties, patent term extension, and exclusive test data questions as whether granting a compulsory license is con- rights. As an agreement between two developed countries, sidered an act of expropriation. Many PTA investment it introduces a number of additional obligations. In partic- chapters expressly remove compulsory licenses from the ular, it mandates the highest-level GI protection for all scope of expropriation, as long as such licenses comply goods, along the lines of the EU–CARIFORUM agree- with the obligations of the TRIPS Agreement and the intel- ment.40 In the area of rights enforcement, it broadens the lectual property chapter of the respective PTA. This, how- application of border measures and criminal sanctions to a ever, is not always the case; the U.S. BITs with Bahrain and wider range of IPR infringements, similar to what the EU Jordan do not feature such a safeguard. Thus, if, say, Jordan has negotiated or is reportedly seeking in its “new genera- were to issue a compulsory license in case of a national tion� PTAs.41 emergency, there is a question as to whether the patent Finally, it is worth pointing out that Chile and Mexico holder could challenge such a decision as an act of invest- have signed a considerable number of PTAs that include ment expropriation. lists of specific GIs related to wines and spirits which Questions like this may be important because many receive protection either as set out in the terms of the investment agreements provide for direct investor-to-state respective agreements or according to TRIPS standards. dispute settlement, going beyond the more traditional In the case of Chile, these lists include the names of state-state dispute settlement procedures included in trade Chile’s many wine regions, as well as the Pisco (or Pisco agreements.42 Investor-state dispute settlement may be more Chileno) liquor distilled from grapes. Agreements that attractive to foreign investors, who can seek arbitration have incorporated at least some of these GIs are the awards for uncompensated expropriation. By contrast, Trans-Pacific Strategic Economic Partnership, with Brunei state-state dispute settlement can typically authorize only Darussalam, Chile, New Zealand, and Singapore, and the imposition of punitive trade sanctions. Chile’s bilateral PTAs with Canada, Japan, Korea, and How widespread are investment treaties that have the Mexico. Mexico’s interests are confined to the Tequila potential to discipline government measures relating to the and Mezcal names, and their protection is expressly protection of IPRs? Given the rapid proliferation of PTAs 402 Carsten Fink and BITs, it is difficult to answer this question precisely. papers and positions can be found on the following Web Some 421 trade agreements had been notified to the WTO sites: up to December 2008 and there are estimated to be more than 2,500 BITs in force.43 Not all trade agreements feature • Knowledge Ecology International, http://www.cptech investment disciplines, and some only establish limited .org/ip/health/trade/ disciplines on foreign direct investment or commercial • Oxfam, http://www.oxfam.org.uk/resources/policy/ presence, which does not encompass IPRs.44 Where PTAs health/bp102_trips.html adopt fully fledged investment chapters, intellectual prop- • Médecins Sans Frontières, http://www.doctorswithout- erty rights are typically included in the definition of invest- borders.org/news/issue.cfm?id=2379 ment. The same is true for BITs, including most treaties concluded between developing countries. There is more The following Web sites contain some industry positions variation on the specifics of investor-state arbitration on the inclusion of IPR provision in trade agreements: mechanisms; for example, in some agreements, the initia- tion of arbitration proceedings is subject to the consent of • International Intellectual Property Alliance, http:// the affected governments. www.iipa.com/fta_issues.html As a final note, it is important to point out that the • Pharmaceutical Research and Manufacturers of America reach of investment agreements into the intellectual (PhRMA), http://www.phrma.org property domain remains in many ways legally uncertain (Correa 2004). So far, no arbitration claim related to International agreements referred to in this chapter can government measures in the IPR field has been made. At be accessed at the following Web sites: the same time, given the proliferation of investment dis- ciplines over the past decade, private investors may well • WTO Secretariat, http://www.wto.org/english/tratop_e/ initiate claims if they feel their IPRs have been unduly TRIPS_e/TRIPS_e.htm “expropriated.� In this context, it is worth pointing out • World Intellectual Property Organization, http://www that some arbitral decisions have been criticized for their .wipo.int expansive interpretation of BIT provisions, creating • U.S. Trade Representative, http://www.ustr.gov more burdensome obligations than those originally • European Commission, Directorate General for Trade, intended by the signatories.45 It is therefore important http://ec.europa.eu/trade/issues/index_en.htm for governments to carefully consider all the implications • European Free Trade Association, http://www.efta.int of signing onto investment disciplines that cover meas- ures affecting IPRs. Notes The views expressed are the author’s own and were written during his Annex. Further Information tenure as Professor of International Economics at the University of St. Gallen, Switzerland. They do not necessarily reflect the views of the The IPRs module of the World Bank Institute World Intellectual Property Organization or its member states. 1. U.S. Chamber of Commerce, http://www.uschamber.com/issues/ (WBI)–Columbia University Executive Training Course index/international/ipr.htm. (Fink 2007) offers a basic introduction to the economics of 2. An important caveat is that when trade opening takes the form of IPRs and discusses the main obligations of the TRIPS preferential tariff liberalization, the welfare consequences from increased trade are more ambiguous. In fact, it is possible that preferential tariffs Agreement. For an in-depth legal analysis of the TRIPS will lead to the displacement of imports from nonmember countries to agreement and the multilateral IPR treaties adminis- such an extent that overall welfare may fall. See Panagariya (2000) for a tered by the WIPO, interested readers are referred to review of the standard economics of preferential trade agreements. UNCTAD–ICTSD (2005) and Abbott, Cottier, and Gurry 3. The rationale for the protection of trademarks and geographical indications is different. These two forms of IPRs seek to promote an (2007). Maskus (2000) provides a comprehensive eco- orderly functioning of markets by reducing information asymmetries nomic analysis of the IPR system, including the perspective between buyers and sellers of goods and services. Fink (2007) provides of developing countries. Abbott (2006) and Roffe (2004) more detailed treatment of the various economic arguments for protect- ing IPRs. analyze U.S. FTA chapters in light of U.S. and Chilean 4. For example, would a patent owner in the form of a juridical per- law, respectively; Santa Cruz (2007) and Vivas Eugui and son qualify if that person were established in the territory of one of the Spennemann (2005) review the treatment of IPRs, espe- PTA parties but 50 percent of its equity were owned by another juridical person in a nonmember state? cially GIs, in EU agreements. 5. A possible exception is the inclusion in PTAs of lists of protected A number of NGOs have campaigned against the inclu- geographical indications, which may lead to a de facto trade preference. sion of TRIPS+ provisions in PTAs. Some of their policy See the section “TRIPS+ Provisions in EU PTAs,� in this chapter. Intellectual Property Rights 403 6. Fink and Maskus (2004) review studies that assess econometrically However, the value of these nonderogation clauses in bilateral disputes is the link between the strength of intellectual property protection and the legally uncertain (Abbott 2004). extent of inward FDI. They conclude that countries which strengthen their 20. See the proposed Digital Media Consumers’ Rights Act, intro- IPR regimes are unlikely to experience a dramatic boost in FDI inflows; duced in the U.S. House of Representatives, http://en.wikipedia.org/wiki/ other factors account for most of the variation across countries in the activ- Digital_Media_Consumers’_Rights_Act. ity of multinational enterprises. At the same time, the empirical evidence 21. Article 41.5, TRIPS Agreement, http://www.wto.org. does point to a positive role of IPRs in stimulating cross-border licensing 22. The U.S.–Chile and CAFTA–DR agreements contain language activity and so affecting the nature of formal technology transfers. similar to that of the TRIPS Agreement, acknowledging that no obligation 7. In many recently concluded North-South PTAs, the willingness of is created regarding the distribution of law enforcement resources. How- developing-country governments to enter into the arrangements was ever, the proviso that resource constraints may not be invoked as an arguably not the lure of preferential market access to a large market but excuse for not meeting the agreements’ specific enforcement obligations the threat of losing previously existing unilateral trade preferences. For appears to significantly weaken this flexibility. example, Colombia and Peru faced the expiry of the so-called Andean 23. The Bipartisan Agreement covers not only IPRs but also a number trade preferences extended by the United States, and countries belonging of other trade-related topics: basic labor standards, environment and to the Africa-Caribbean-Pacific (ACP) group (former European colonies) global warming, government procurement, port security, investment, and faced the expiry of trade preferences granted by the EU under the Cotonou worker assistance and training. The text of the agreement is available at Agreement. http://www.cpath.org/sitebuildercontent/sitebuilderfiles/2007_new_trade_ 8. A notable exception is a project by the International Center for policy_details5-10-07.pdf. Roffe and Vivas-Eugui (2007) offer a commen- Trade and Sustainable Development (ICTSD), the World Bank Institute tary on the agreement. (WBI), and the World Health Organization (WHO) that focuses on the 24. For instance, drug regulators need to give general notice of sub- pharmaceutical sector. The aim of the project is to develop a partial equi- missions of applications for marketing approval so that patent holders librium model to assist researchers in assessing the price and expenditure have the opportunity to discover products that may infringe on their effects of prolonged market exclusivity for pharmaceutical products. For patents. further information, see the website on intellectual property rights 25. In the case of the U.S.–Colombia FTA, the Bipartisan Agreement sponsored by the ICTSD and the United Nations Conference on Trade identified additional concerns about violence against trade unionists and and Development, http://www.iprsonline.org/ictsd/Dialogues/2007-05- demanded that this issue be addressed in a satisfactory way—along with 27/2007-05-27_desc.htm. the other proposed revisions set out in the agreement—before the revised 9. In addition to FTAs, the United States has negotiated several bilat- FTA is submitted for congressional approval. eral trade agreements (BTAs) with countries that are not yet members of 26. The same changes have been introduced into the texts of the U.S. the WTO and therefore do not enjoy full MFN treatment by the U.S. gov- FTAs with Colombia and Panama. ernment. Some of these BTAs, notably that with Vietnam, provide for 27. Table 18.2 does not include the EU’s association agreements with obligations on IPRs; for reasons of space, these are not discussed here. For countries in Eastern Europe and the Euro-Mediterranean partners. Some further details, see Fink and Reichenmiller (2005). of these agreements incorporate soft-law provisions that generally call for 10. U.S. Trade Promotion Authority Act of 2002, http://frwebgate the protection of IPRs in conformity with the highest international stan- .access.gpo.gov/cgi-bin/getdoc.cgi?dbname=107_cong_public_laws&docid dards or at a level similar to that existing in the EU. They also make refer- =f:publ210.107. ence to existing multilateral treaties, in particular the TRIPS Agreement 11. Given their older vintage, we do not review the IPR provisions of and selected WIPO conventions, sometimes including the WIPO Copy- the U.S.–Israel FTA or of NAFTA. right Treaty and the WIPO Performances and Phonograms Treaty. For the 12. The CAFTA–DR agreement requires countries that already pro- most part, accession to these treaties is voluntary. See Santa Cruz (2007) vide patent protection for plants to maintain such protection. for further discussion. 13. In addition, the TRIPS provisions on compulsory licensing 28. Official Journal of the European Union, document 2005/C 129/03. require a government to first make efforts to obtain a voluntary license 29. For a more detailed discussion of GI obligations in EU bilateral from the patent holder, although this requirement can be waived in emer- agreements, see Vivas Eugui and Spennemann (2005); Santa Cruz (2007). gency situations or for public, noncommercial use. The obligations of 30. See Battaglene (2005). In addition, Schamel and Anderson (2003) bilateral agreements are similar or identical in this respect. find that regional origin has become a major determinant of prices in the 14. In the case of agrochemical products, most of the bilateral agree- Australian wine industry, with price premiums averaging about 31 per- ments require data exclusivity for 10 years. cent for wines carrying Australian GIs. 15. The permissibility of parallel importation is governed by rules on 31. Although the interim EPAs that have been negotiated so far with the exhaustion of patents. A system of international exhaustion is associ- other ACP partners do not yet contain a full IPR chapter, they do call for ated with free parallel trade, whereas patent holders can restrict parallel the negotiation of such a chapter in the final agreement. importation if patent rights exhaust only nationally. TRIPS Article 6 does 32. The EU–CARIFORUM EPA foresees full implementation of the not mandate a particular exhaustion regime, only nondiscriminatory IPR chapter by 2014. Least-developed CARIFORUM members are application. granted a special transition period until 2021. 16. Governments may, however, limit the right to prevent parallel 33. In fact, there is no consensus among WTO members as to importation to cases in which the patent holder includes territorial whether the Doha Ministerial Declaration of 2001 provides a negotiating restraints in contractual agreements with wholesalers or retailers. mandate for extension of GI protection; see Fink and Maskus (2005). 17. Paragraph 4, Doha Declaration on TRIPS and Public Health, 34. See Council Regulation (EC) 1383/2003, concerning customs http://www.wto.org/english/thewto_e/minist_e/min01_e/mindecl_trips_ action against goods suspected of infringing certain intellectual property e.htm. rights and the measures to be taken against goods found to have infringed 18. The side letters also clarify that the intellectual property chapters such rights, and Commission Regulation (EC) 1891/2004, which lays down of the FTAs “do not prevent the effective utilization� of the August 2003 provisions for the implementation of Council Regulation (EC) 1383/2003. decision by WTO members, described in the text. 35. See, for example, Article 154 of the EPA, on measures for preserv- 19. The agreements with Australia, CAFTA–DR, Chile, and Jordan ing evidence, and Article 155 on the right of information. contain provisions affirming the rights and obligations of member coun- 36. The EU–CARIFORUM EPA foresees the provision of technical tries under the TRIPS agreement. To some extent, these provisions may assistance by the EU for implementing the IPR obligations undertaken by be interpreted as preserving the flexibilities of the TRIPS agreement. CARIFORUM members. Although aid of this type may alleviate the 404 Carsten Fink budgetary pressure associated with strengthened IPR enforcement, a sim- Commission on Intellectual Property Rights established by the U.K. ilar caveat still applies: are the EU’s scarce aid resources best spent in the Secretary of State for International Development, London. IPR field as opposed to, say, education or health? Correa, Carlos M. 2004. “Bilateral Investment Agreements: Agents of a 37. The EU’s draft IPR chapter is available at http://www.bilaterals New Global Standard for the Protection of Intellectual Property .org/article.php3?id_article=14281. Rights?� Briefing paper, GRAIN, Barcelona, Spain, and Los Banos, the 38. For example, Japan’s PTAs with Indonesia, Malaysia, Mexico, Philippines. http://www.grain.org/briefings/?id=186. Singapore, Thailand, and Vietnam fall into this category. The same is true Dolzer, Rudolf, and Christoph Scheuer. 2008. Principles of International for the Japan–Philippines economic partnership agreement, except for Investment Law. New York: Oxford University Press. one notable TRIPS+ obligation: the expansion of criminal sanctions Fink, Carsten. 2005. “Shifting Tides: TRIPS in the Doha Round.� World beyond trademark counterfeiting and copyright piracy to include, as well, Trade Brief. London: Haymarket Management. infringements of patents, utility models, layout designs for integrated cir- ———. 2007. “Intellectual Property Rights: Economic Principles and cuits, and plant variety rights. Trade Rules.� Training module prepared for the World Bank Institute– 39. See Fink (2005) for additional background on this topic. The Columbia University Executive Course on Trade Policy for Develop- EFTA–Colombia PTA also contains strong language on the fair and equi- ment, World Bank, Washington, DC. table sharing of benefits arising from the use of genetic resources and tra- ———. 2008. “Intellectual Property and Public Health: An Overview of ditional knowledge, although it does not make an agreement on benefit the Debate with a Focus on U.S. Policy.� Working Paper 146, Center sharing a precondition for the grant of a patent. General language on the for Global Development, Washington, DC. use of biological resources and traditional knowledge and the equitable Fink, Carsten, and Keith E. Maskus. 2004. Intellectual Property and Devel- sharing of benefits is also found in the agreements between the United opment: Lessons from Recent Economic Research. Washington, DC: States and Peru, the EU and the Caribbean Community (CARICOM), and World Bank; New York: Oxford University Press. Guatemala and Taiwan, China. ———. 2005. “The Debate on Geographical Indications in the WTO.� In 40. The GI provisions of the Japan–Switzerland agreement have two Trade, Doha, and Development: Window into the Issues, ed. Richard other noteworthy features. First, they introduce the concepts of indications Newfarmer. Washington, DC: World Bank. of services and country names, which equally benefit from protection Fink, Carsten, and Patrick Reichenmiller. 2005. “Tightening TRIPS: The against misleading use. Second, the parties exchanged lists of geographic Intellectual Property Provisions of Recent US Free Trade Agreements.� names that serve as a source of information about GIs protected under Trade Note 20, World Bank, Washington, DC. domestic laws in the parties’ respective territories. These lists are not limited Hernández, Xavier Seuba. 2009. “Health Protection in the European and to wines and spirits but include cheeses, meat-based products, bread, pastry, Andean Association Agreement.� Health Action International, Ams- cakes, watches, textile products, pharmaceutical products, and other goods. terdam. http://www.haiweb.org/23032009/18%20Mar%202009%20 41. The Japan–Switzerland EPA extends border measures to goods Policy%20Paper%20EU-CAN%20Association%20Agreement%20 destined for export, as well as transiting goods. FINAL.pdf. 42. An exception is the investment chapter of the U.S.–Australia FTA, Maskus, Keith E. 2000. Intellectual Property Rights in the Global Economy. which only allows for the possibility of investor-state dispute settlement Washington, DC: Institute for International Economics. procedures being negotiated in the future. MSF (Médecins Sans Frontières). 2004. “Access to Medicines at Risk 43. See WTO, http://rtais.wto.org/UI/PublicMaintainRTAHome.aspx; across the Globe: What to Watch Out for in Free Trade Agreements Dolzer and Scheuer (2008). with the United States.� MSF Briefing Note, MSF. Geneva. 44. The EU–CARIFORUM, EU–Chile, Australia–Thailand, Chile– Oxfam. 2006. “Patents versus Patients: Five Years after the Doha Declara- EFTA, and Mexico–EFTA agreements are examples of the latter. tion.� Oxfam Briefing Paper 95, Oxfam, Oxford, U.K. http://www 45. For example, certain interpretations of obligations relating to .oxfam.org/sites/www.oxfam.org/files/Patents%20vs.%20Patients.pdf. “fair and equitable treatment� and measures “tantamount to expropria- Panagariya, Arvind. 2000. “Preferential Trade Liberalization: The Tradi- tion� have been considered to go beyond parties’ original intentions. In tional Theory and New Developments.� Journal of Economic Literature response to some of these arbitral decisions, the United States and Canada 32 (2): 287–331. have included interpretative notes in their BITs and their FTA investment Reichman, Jerome H. 2004. “Undisclosed Clinical Trial Data under the chapters clarifying the scope of the fair and equitable treatment and TRIPS Agreement and Its Progeny: A Broader Perspective.� Pre- expropriation provisions. See Sornarajah (2004). sented at the United Nations Conference on Trade and Develop- ment (UNCTAD)–International Centre for Trade and Sustainable Development (ICTSD) Dialogue, “Moving the Pro-Development IP References Agenda Forward: Preserving Public Goods in Health, Education and Learning,� Bellagio, Italy, November 29–December 3. http:// Abbott, Frederick. 2004. “The Doha Declaration on the TRIPS Agreement www.iprsonline .org/unctadictsd/bellagio/docs/Reichman_Bellagio4. and Public Health and the Contradictory Trend in Bilateral and pdf. Regional Free Trade Agreements.� Occasional Paper 14, Quaker Roffe, Pedro. 2004. “Bilateral Agreements and a TRIPS-Plus World: The United Nations Office, New York. Chile–USA Free Trade Agreement.� TRIPS Issues Paper 4, Quaker ———. 2006. “Intellectual Property Provisions of Bilateral and Regional International Affairs Programme, Ottawa. Trade Agreements in Light of U.S. Federal Law.� International Center Roffe, Pedro, and David Vivas-Eugui. 2007. “A Shift in Intellectual Prop- for Trade and Sustainable Development (ICTSD)–United Nations erty Policy in US FTAs?� Bridges (International Center for Trade and Conference on Trade and Development (UNCTAD) Project on IPRs Sustainable Development) 11 (5): 15–16. http://ictsd.org/i/news/ and Sustainable Development, Issue Paper 12, UNCTAD, Geneva. bridges/4128/. Abbott Frederick, Thomas Cottier, and Francis Gurry. 2007. “International Santa Cruz, Maximiliano S. 2007. “Intellectual Property Provisions in Intellectual Property in an Integrated World Economy.� New York: European Union Free Trade Agreements: Implications for Develop- Aspen Publishers. ing Countries.� Issue Paper 20, ICTSD Programme on Intellectual Battaglene, Tony. 2005. “The Australian Wine Industry Position on Geo- Property Rights (IPRs) and Sustainable Development, International graphical Indications.� Paper presented at the Worldwide Symposium Center for Trade and Sustainable Development, Geneva. on Geographical Indications, Parma, Italy, June 27–29. Schamel, Günter, and Kym Anderson. 2003. “Wine Quality and Varietal, CIPR (Commission on Intellectual Property Rights). 2002. “Integrating Regional and Winery Reputations: Hedonic Prices for Australia and Intellectual Property Rights and Development Policy.� Report of the New Zealand.� Economic Record 79 (246): 357–69. Intellectual Property Rights 405 Sornarajah, Muthucumaraswamy. 2004. The International Law on Vivas Eugui, David, and Christoph Spennemann. 2005. “The Treatment of Foreign Investment. 2nd ed. Cambridge, U.K.: Cambridge University Geographical Indications in Recent Regional and Bilateral Free Trade Press. Agreements.� In The Intellectual Property Debate: Perspectives from UNCTAD-ICTSD (United Nations Conference on Trade and Develop- Law, Economics, and Political Economy, ed. Meir Perez Pugatch, ment–International Center for Trade and Sustainable Development). 305–44. Cheltenham, U.K.: Edward Elgar. 2005. Resource Book on TRIPS and Development. Cambridge, U.K.: Woolcock, Stephen. 2007. “European Union Policy towards Free Trade Cambridge University Press. http://www.iprsonline.org/unctadictsd/ Agreements.� ECIPE Working Paper 03/2007, European Centre for ResourceBookIndex.htm. International Political Economy, Brussels. 19 Environment Anuradha R. V. The linkage between trade and environment is a sensitive countries to address regional environmental concerns, and controversial issue for many countries. Trade, like any including concerns relating to shared natural resources, and other economic activity, has environmental implications, it concludes by discussing practical steps for developing but whether trade as a policy instrument should address countries to consider when negotiating PTAs. environmental concerns is a question to which there are no The analysis relies on a review of those PTAs that incor- unanimous answers. Although practically all countries rec- porate environmental provisions and on secondary litera- ognize the critical importance of the environment, a number ture on the subject. Inclusion of environmental provisions of developing countries do not believe that addressing envi- in PTAs is of fairly recent date, and the literature analyzing ronmental issues as part of trade agreements is necessarily the impact of the linkage is also at a nascent stage. The the best approach. Accordingly, a few developing countries most comprehensive study on environmental provisions in have entered into regional agreements on environmental PTAs was conducted by the Organisation for Economic issues that are distinct from preferential trade agreements Co-operation and Development (OECD 2007). That study (PTAs). Developed countries have been the primary drivers identifies the United States, the European Union (EU), for PTAs comprising environmental provisions. Neither Canada, and New Zealand as the principal proponents of approach has been proved more effective than the other. environmental provisions in PTAs. The study underscores This chapter does not seek to build a rationale for the the paucity of experience with actual implementation of trade-environment linkage or to determine whether envi- environmental provisions and the ongoing learning ronmental issues should be addressed in trade agreements process. It emphasizes the need for preparation and coordi- or in stand-alone environmental agreements. The focus, nation in dealing with environmental issues in PTAs and instead, is on the main contours of the trade and environ- stresses that developing countries often require financial ment debate at the multilateral level and on the increasing support and capacity-building assistance to this end. incorporation of environmental provisions into PTAs, Asia-Pacific Economic Cooperation (APEC) and the especially those involving major developed economies. The Organization of American States (OAS) maintain data- chapter addresses issues relating to (a) the nature of the bases that summarize environmental provisions in PTAs.1 legal obligations emerging from provisions dealing with Studies by research institutions and think tanks have also the environment in PTAs; (b) the potential economic costs been consulted for this chapter. of particular environmental requirements, including obli- The only PTA to have been analyzed in substantive detail gations to maintain specific environmental regulatory is the North American Free Trade Agreement (NAFTA), standards and to adhere to particular environmental sani- which was concluded by the United States, Canada, and tary and phytosanitary measures or technical regulations; Mexico in 1994. Studies have been conducted to evaluate (c) the need for technical assistance and capacity building to NAFTA’s implications for the environment in Mexico, ensure compliance with environmental obligations; (d) the its economic costs, and the effectiveness of using a PTA nature and extent of financial assistance required for to address environmental issues (see Gallagher 2004; implementing such provisions; and (e) the dispute settle- Tiemann 2004; Hufbauer and Schott 2005; Environment ment and enforcement mechanisms for such provisions. It Canada 2007). The findings have been mixed, with some analyzes the extent to which PTAs can be a vehicle for analysts hailing NAFTA’s achievements and others 407 408 Anuradha R. V. highlighting its limitations and unresolved issues, including found that although trade-led growth resulted in a shift the inadequacy of financial resources for combating envi- toward cleaner production by a number of industries, indus- ronmental problems. trial pollution increased overall because nations in the hemisphere “lack the proper policies to stem the environ- mental consequences of trade-led growth.�3 Trade and Environment Linkages The only literature to examine whether, as a policy mat- This section looks at the economic and political-economy ter, trade agreements are an appropriate forum for address- dimensions of PTAs and discusses WTO policies that affect ing environmental issues is a World Trade Organization environmental provisions in trade agreements. (WTO) background note (WTO Secretariat 1997). It observes that trade instruments are not the first-best policy for addressing environmental problems and finds a posi- Economic Perspectives tive correlation between removal of trade restrictions and Initial discussions on the subject of trade and environmental increased availability of environmental goods and services linkages revealed two distinct schools of thought, one argu- and of cleaner technologies. ing for the integration of adequate environmental safe- guards within trade agreements and the other maintaining Political-Economy Considerations that free trade should not be subject to any restrictions. According to the first school, free trade does not necessarily Turning to the political-economy dimension of the trade- lead to efficiencies, and it can bring about a deterioration of environment debate, the divide between developed and environmental standards. Firms whose main concern is to developing countries is distinct. Developed economies maximize profits may be inclined to move their operations such as the United States, Canada, New Zealand, and the to developing countries, where pollution control is inexpen- EU have been among the strongest proponents of environ- sive and lax—this scenario has been termed the “pollution mental measures in trade agreements, both at the multi- haven� hypothesis (Daly 1993a, 1993b). Porter (1991), draw- lateral level and in PTAs. In fact, the domestic policies of ing on experience in northern Europe and Japan, concludes these countries require that environmental assessments of that stronger environmental regulation is not only necessary PTAs be carried out and that adequate provisions on the but could also make firms more competitive. environment be included in the agreements. U.S. Public On the other side of the spectrum is the argument that Law 107-210, the Trade Promotion Act (TPA) of 2002, there should be no environmental or other restrictions on requires, among other things, that countries signing trade free trade, on the grounds that free trade will eventually agreements with the United States ensure that their envi- lead to economic growth, increased income levels, and, as a ronmental laws are enforced. The EU mandates sustain- consequence, investment in higher environmental stan- ability impact assessments (SIAs) of PTAs (European dards (Bhagwati 1993). The argument of this group is not Commission 2006). New Zealand’s Framework for Inte- that free trade has no environmental implications; it is, grating Environment Issues into Free Trade Agreements rather, that trade should not be used as a tool to impose emphasizes the harmonization of the “objectives for trade environmental standards because the welfare implications and for the environment, with both serving the overarch- of free trade are independent of those standards (Bhagwati ing objective of promoting sustainable development.�4 and Srinivasan 1996; Bhagwati 2004). Developing countries, by contrast, have been skeptical More recent literature explores the subject from other about the trade-environment linkage, their concern angles and leads to findings somewhere between the first two being that environmental issues might become a means schools. Empirical evidence on the pollution haven hypoth- of imposing protectionist measures (see, for example, esis has been mixed. Studies have demonstrated that lower Khan et al. 2004; Rodriguez 2009). India, among other environmental regulations do not necessarily lead to a race countries, has emphasized that harmonization may be to the bottom and that environmental regulations are not easier for developed countries and that among develop- the only factors that guide investment decisions (see, for ing countries, with their widely differing environmental example, Friedman, Gerlowski, and Silberman 1992; Mani, standards in accordance with their national priorities, Pargal, and Huq 1997). At the same time, empirical research harmonization would be both difficult and inadvisable demonstrates that although trade liberalization can promote (India 2000). However, countries such as Chile, China, economic growth, it may have adverse environmental conse- and Mexico have been incorporating environmental pro- quences, which it would be necessary to address simultane- visions into their PTAs, although the scope and depth of ously by strengthening environmental regulation (see Mani such provisions are not as elaborate as those proposed by and Jha 2006).2 A study of PTAs in Latin America similarly developed countries. Environment 409 Trade and Environment at the WTO increasingly agreed to environmental obligations in PTAs with these countries. With the exceptions of perhaps Chile Any discussion of trade and environment is perhaps best and Mexico and, more recently, China, developing coun- appreciated when set in the context of the multilateral dis- tries have not themselves initiated environmental provi- cussions at the WTO. Countries entering into PTAs typi- sions in PTAs. cally focus on the rights and obligations under a PTA that Proponents of environmental provisions in PTAs either are additional to or different from those under the WTO’s incorporate such provisions into the main text of the PTA multilateral system. As noted above, not all WTO mem- or place them in separate side agreements. Such provisions bers agree on whether and to what extent environmental typically pertain to any or all of the following matters: obli- provisions should be reflected in trade agreements. The gations relating to “high standards� in domestic environ- WTO agreements, however, do confirm the right of WTO mental laws; mechanisms for resolving disputes involving members to protect the environment, provided that cer- environmental provisions; principles of cooperation on tain conditions are met. Environmental provisions find a environmental issues; and provisions on technical assis- place within the preamble of the Agreement Establishing tance and capacity building. These obligations are a blend the WTO, in general exceptions to trade obligations, and of legally binding and nonbinding provisions. (The annexes in agreements on technical barriers to trade (TBTs) and to this chapter summarize environmental provisions in U.S. sanitary and phytosanitary measures (SPS).5 At the Doha and EU PTAs.) ministerial conference, WTO members reaffirmed their Since the conclusion, in 1994, of NAFTA and a side commitment to environmental protection and agreed to agreement, the North American Agreement on Environ- embark on a new round of trade talks, including negoti- mental Cooperation (NAAEC), all PTAs negotiated by the ations on certain aspects of the linkage between trade United States have included environmental considera- and environment. tions both in environmental chapters and in separate There have been several prominent disputes at the WTO instruments that focus mainly on environmental cooper- concerning the trade and environment interface.6 Some ation. Examples include the agreements with Singapore, basic propositions emerge from the jurisprudence that has Chile, Australia, Bahrain, and Morocco and the PTA with been developed so far: the five countries of the Central America Free Trade Agreement, plus the Dominican Republic (CAFTA–DR).7 • WTO law does not exist in clinical isolation from inter- These agreements explicitly establish an obligation by the national law and international developments, including parties to effectively enforce their environmental laws, environmental concerns. However, environmental meas- and they include provisions to ensure enforcement of this ures to restrict trade can be adopted only under certain commitment through mechanisms for dispute settlement strict conditions. and public submissions. They provide for environmental • Multilateral solutions to environmental issues are pre- cooperation between the parties and are accompanied by ferred, and WTO members should therefore make seri- an environmental cooperation agreement or memoran- ous efforts to negotiate such solutions. If, despite such dum of understanding that establishes a framework for efforts, an agreement cannot be concluded, unilateral such cooperation. Canada’s PTAs also contain compre- measures for protection of the environment may be hensive provisions on the environment, similar to those taken, even outside the country’s jurisdiction. in U.S. PTAs. • Adequate scientific evidence and risk assessment must The EU’s early PTAs, such as those with Mexico, Chile, underlie any action under the WTO’s SPS Agreement, and the Mediterranean countries, contain fewer and more which allows for measures to protect human, animal, broadly worded provisions on the environment.8 It is only and plant life and health. in the recently concluded economic partnership agreement • Tests of necessity and of the availability of less trade (EPA) between the EU and the Caribbean Forum of restrictive measures are to be applied prior to applica- African, Caribbean, and Pacific (ACP) States (CARIFO- tion of any trade restriction on environmental grounds. RUM) that the EU has devoted a separate chapter to envi- ronmental provisions. The most significant difference between U.S. and EU Overview of Approaches to Environmental PTAs lies in the dispute settlement mechanism. All U.S. Concerns in PTAs PTAs except the one with Jordan prescribe remedies in the As has been noted, the primary proponents of environ- form of monetary compensation for noncompliance with mental provisions in PTAs have been the United States, the environmental provisions. In the event that such compen- EU, Canada, and New Zealand. Developing countries have sation is not paid by a party, then—as a last resort—tariff 410 Anuradha R. V. concessions may be suspended. EU PTAs leave the issue of Regional Public Goods “compliance measures� to the judicial mechanism hearing What motivates countries in a region to address a com- the dispute. Remedies may include monetary compensa- mon environmental concern? Agreements between tion, but not suspension of concessions. Several U.S. PTAs neighboring countries, whether PTAs or stand-alone also provide for public submission of complaints to the regional environmental agreements, often provide a plat- institutional authority responsible for implementing the form for addressing environmental concerns with trans- PTA; this provision is absent in EU PTAs. boundary implications. The OECD study on regional New Zealand has so far addressed environmental con- trade agreements observes that when partners share cerns in side agreements on environmental cooperation, ecosystems, there is risk of “environmental blowback� using the softer language of “intent� and “endeavour� to from unmanaged growth across the border (OECD 2007, achieve environmental objectives and ensure environmen- 83). For instance, concern in U.S. border states about the tal cooperation. Japan’s approach, like New Zealand’s, seems growth of polluting industries in Mexico is perceived to to be focused on principles of cooperation to achieve speci- have provided a strong impetus for building environ- fied environmental goals. More recently, Japan has been mental elements into the NAFTA relationship (ibid). introducing environmental provisions, primarily worded as Another example is the ASEAN–China PTA, which, in principles of cooperation, into its PTAs. Australia, so far, is Article 7.1(3), emphasizes Mekong River Basin develop- the only major developed country to maintain a stand that ment as a priority area for cooperation.10 Environment, environmental issues need to be addressed separately from fisheries, forestry, and forestry products are among the trade agreements. In its PTA with the United States, how- many policy areas in which the parties commit to coop- ever, Australia has agreed to environmental provisions that erate (Article 7.2). are similar to those in other U.S. PTAs. PTAs are not the only vehicle for addressing regional A number of regional economic groups of developing environmental issues (and in any case, PTA partners are countries recognize environmental issues as an important not necessarily geographically proximate). As was seen in matter for regional cooperation. They include the South- the preceding section, several regional groups of develop- ern Cone Common Market (Mercosur, Mercado Común ing countries have addressed environmental concerns in del Sur); the Andean Community; the Association of stand-alone environmental agreements rather than in Southeast Asian Nations (ASEAN) Free Trade Area PTAs. For example, (AFTA); the South Asian Association for Regional Cooper- ation (SAARC); the Caribbean Community (CARICOM); the East African Community (EAC); and the Southern • ASEAN countries have concluded an agreement that African Development Community (SADC).9 These groups provides for work on transboundary haze, nature con- deal with environmental issues in separate agreements or servation and biodiversity, the coastal and marine envi- understandings, not as part of a PTA. Their agreements ronment, global environmental issues, and cross-cutting contain no provisions on linkages between trade and environmental initiatives. environment. Although economic cooperation is one of • The SAARC Plan of Action on Environment sets the the pillars of these groups, environmental issues are not parameters for enhanced action at the regional level and regarded as part of the economic relationship. Environ- has led to the establishment of two Regional Centers of ment and economics are treated as distinct and separate Excellence in the field of environment.11 elements of the regional interrelationships between • The SADC Wildlife Programme of Action promotes the countries. coordinated action by member states for protection of Among the major developing economies, Brazil, the wildlife in the region. Russian Federation, India, and China have supported and • The Caribbean Environment Program focuses on envi- participated in all major multilateral environmental ronmental management, training, and awareness pro- agreements. Their PTAs typically contain references to grams in relation to regional issues such as coastal the environment as part of the preamble or the general zone management, biodiversity, coral reef manage- objectives. These are worded as statements of intent, rather ment, protected areas, and wildlife protection. than as binding legal obligations. Recent PTAs entered into by China contain environmental provisions of a more Experience so far indicates that countries in a region elaborate nature. These are primarily couched as “soft� are most likely to address a problem that is transboundary obligations in its agreements with ASEAN, New Zealand, in nature and affects all of them. The framework of a trade and Singapore. agreement such as a PTA is not a necessary precondition. Environment 411 It is also important to note that such agreements empha- higher environmental standards. Box 19.1 lists a number of size principles of cooperation, rather than legally binding considerations for developing countries in negotiating obligations. environmental issues in PTAs. PTAs, and other regional agreements between countries The United States, Canada, and New Zealand have that are geographically proximate, are more likely to suc- undertaken pre-PTA environment impact assessments cessfully address regional environmental concerns than (EIAs) of the environmental impacts of PTAs in their own global issues—for which multilateral cooperation would countries. The EU approach is slightly different; sustain- be necessary. PTAs are probably less likely to tackle global ability impact assessments (SIAs) are carried out for both environmental problems such as climate change, declining EU member states and the trading partners with which the biodiversity, depletion of ocean fisheries, and overexploita- PTA is being negotiated. No study has yet been done on tion of shared resources. There are, however, a few PTAs the extent to which the findings from these assessments that seek to address environmental issues that are not are reflected in the drafting of PTA provisions, but such purely regional. An example is the Japan–Mexico PTA, assessments are likely to be valuable tools for designing which contains provisions on environmental cooperation effective environmental provisions that can have develop- on activities to implement the Clean Development Mecha- mental benefits. It would stand to reason that the broader nism and the Kyoto Protocol. The U.S.–Peru PTA includes scope of an EU-type SIA inquiry into environmental and detailed annexes on biodiversity protection and conserva- sustainable development impacts in all the trading part- tion and on forestry governance, and U.S. PTAs with ners to a PTA is a conceptually superior method of Colombia and Costa Rica also address biodiversity protec- addressing economic development in the trading parties. tion. These PTAs exhibit a mix of legally binding and non- binding approaches that seek to complement multilateral Regional Environmental Policy in a efforts on similar issues. Multilateral World One of the main criticisms of PTAs is that they lead to pro- Benefits of the Alternative Approaches for Economic liferation of diverse rules and to increased complexity in Development and the Environment trade relations between countries. Several visually descrip- The above discussions reveal three broad approaches: tive phrases have emerged in trade linguistics to describe this effect of PTAs, such as “spaghetti bowl,� “noodle bowl,� and • Environmental provisions worded as binding commit- more recently, “termites in the trading system� (Bhagwati ments and obligations and incorporated in PTAs 2008).12 The essential point, as summarized in a recent • Environmental provisions included in PTAs as non- WTO publication, is that the proliferation of PTAs has cre- binding soft obligations, in the form of arrangements ated an array of criss-crossing arrangements, with little for cooperation and statements of intent attention to coherence among agreements or to the impli- • Environmental agreements or understandings, mostly cations of so many regimes for trade costs and efficiency worded as soft obligations, outside the purview of PTAs. (Baldwin and Low 2009). Against the above argument, it is sometimes asserted No study has been carried out on the relative effective- that multiplication of PTAs need not necessarily be ness, from the perspective of promoting sustainable devel- viewed as enhancing complexities and that liberalization opment, of these alternatives. through PTAs may in fact lead to greater multilateral lib- The effectiveness of any approach will hinge on the eralization over time. It is still early for a definitive assess- nature of commitments by member countries to a specific ment on this issue. The discussion here asks whether agreement and on the political and financial resources for environmental provisions offer potential for the exten- implementing such commitments. Technology sharing and sion of open regionalism and whether they discriminate transfer, technical assistance and capacity building, and against nonparties. financing of environmental initiatives are critical determi- nants of the economic development benefits of these Open Regionalism approaches. This is borne out by experience with imple- mentation of environmental provisions, as discussed One of the main issues concerning the use of PTAs as below. These experiences underscore that environmental against a multilateral approach to trade negotiations is provisions in PTAs need to be accompanied by sustained whether PTAs can eventually lead to open regionalism—that support for regulatory and institutional building to promote is, to external liberalization by regional trading blocs. 412 Anuradha R. V. Box 19.1. Considerations for Developing Countries in Negotiating PTAs In view of the growing tendency to include environmental provisions in PTAs, it is important for developing countries to (a) assess the likelihood that a trading partner will raise environmental issues as part of the PTA agenda, (b) analyze the potential implications of such provisions, and (c) develop a proactive agenda of their own concerns and desired positions. Some specific considerations and recommendations follow. 1. Preparedness. Examine the potential partner’s existing PTAs to assess beforehand what environmental provisions the other party is likely to suggest. Flexibilities offered in other PTAs entered into by the other country should be explored to the fullest. 2. Assistance for pre-PTA impact assessments. If sustainable impact assessment (SIA) in both parties to the PTA is to be part of the negotiating process, seek technical, financial assistance, and capacity-building assistance from the party that has more experience with such assessments. 3. Key substantive issues. Keep in mind the following issues in negotiating environmental provisions: (a) the legal obligations emerging from the provisions; (b) the potential economic costs of environmental requirements, including obligations to maintain specific environment regulatory standards and requirements relating to adherence to any environmental SPS measures or technical regulations; (c) the need for technical assistance and capacity building to ensure compliance with environmental obligations; (d) the financial assistance required; and (e) dispute settlement and enforcement mechanisms. 4. Implications for domestic law and institutions. Evaluate the nature, extent, and feasibility of regulatory and legal changes that may be required at the domestic level to ensure compliance with the proposed environmental provisions in the PTA. 5. Economic costs. Carefully assess the economic costs of implementing environmental provisions, including the cost of new regulations and new institutional mechanisms for enforcement and the effects of more stringent environmental standards on export competitiveness. Capacity-building and technical assistance may be needed to ensure adherence to such requirements, and adequate and sustained support for such purposes should be built into the PTAs. Pre-PTA assessments would be valuable tools for understanding the costs. 6. Provisions for resolving conflicting norms. Determine whether the PTA contains provisions for resolving conflicts between different aspects of the PTA—for example, between trade and investment priorities and environmental priorities. Adequate mechanisms for dialogue between the institutional frameworks responsible for these norms are necessary. 7. Specific environmental concerns. Developing countries’ environmental agendas when negotiating PTAs need not always be reactive; identify and raise specific environmental matters of concern. Examples may include SPS measures or technical regulations faced by exporters, or exporters’ need for access to environmentally friendly technology. Cooperation and technical assistance may be required under the PTA to facilitate market access to the developed country’s market. 8. Preferential market access. Identify provisions that could be built into the PTA concerning preferential access to clean technologies and to renewable and energy-efficient goods and services. Requirements for technical assistance and capacity building to enhance domestic capacity for developing environmental goods and services and clean technologies could also be considered. 9. Technical, financial, and capacity-building assistance. Determine whether proposed provisions on environmental obligations are binding. Legally binding obligations would need to be accompanied by binding commitments from developed-country partners to provide technical, financial, and capacity-building assistance to the developing partner to support the enactment and enforcement of environmental regulations. Specific obligations by a developing country to enact and maintain environmental regulations should be made conditional on actual development assistance and capacity-building support from the developed-country partner. Clear benchmarks for monitoring implementation of such provisions are necessary. Specific areas in which cooperation and assistance are required should be prioritized, and the associated work programs need to be monitored. 10. Dispute settlement provisions. Consider negotiating dispute settlement mechanisms that emphasize consultation and cooperation. Binding dispute resolution mechanisms may be theoretically superior for ensuring the effectiveness of a law, but the implementation of environmental provisions depends on other factors, such as the availability of technical and financial resources. In view of this, it may be advisable to opt for binding dispute resolution only if all the elements for securing effective implementation of the provisions, including the technical and financial assistance required for implementing any new environmental laws and standards, are built into the PTA. Several working definitions of the phrase have been Broadly, as described above, developed actors such as the proposed, differing from each other as to the extent of United States, the EU, Canada, and New Zealand are the openness necessary if the trade policy of a regional bloc is primary drivers of stronger environmental provisions in to qualify as open regionalism.13 PTAs, exerting pressure that a developing country partner The concept of open regionalism in the context of envi- may find difficult to resist during the trade negotiations. ronmental provisions in PTAs is interpreted here as an PTAs entered into by these countries are likely to replicate assessment of whether countries adhering to environmen- similar provisions, which may create a certain degree of tal obligations in one PTA are replicating similar provi- harmonization between agreements over time. sions in other PTAs. In other words, is there a possibility A developing country may agree to environmental pro- that common environmental standards will be applied visions in a PTA with a developed country but may not through replication of regional approaches? choose to replicate such provisions in its PTAs with other Countries have taken disparate approaches toward countries, whether developed or developing. Even when it incorporating environmental provisions into their PTAs. does, the environmental provisions are likely to be milder Environment 413 (i.e., couched as softer obligations) than those negotiated goods and services if liberalization is confined to parties to with developed-country partners. An illustration is the the PTA and is not undertaken on a most favored nation approach adopted by Chile and Mexico, both of which (MFN) basis. adhered to comprehensive environmental obligations for Liberalization of trade in environmental goods and the first time in their PTAs with the United States and services (EGS) was made part of the WTO’s agenda under Canada. Subsequent PTAs concluded by Chile and Mexico the Doha Round negotiations.15 It remains a contentious with other countries incorporate environmental provi- issue.16 To the extent that countries agree on preferential sions, but these are not as elaborate as the provisions in treatment of EGS at the bilateral level, discrimination their PTAs with the United States and Canada. Chile has against third parties could result. However, the provisions signed a memorandum of environmental cooperation with on EGS in PTAs have so far only been in the form of broad China, and it has also concluded PTAs with Japan, Mexico, commitments to cooperate, rather than of concrete obliga- Peru, and the Republic of Korea that contain some provi- tions to liberalize trade. sions on the environment.14Apart from NAFTA, Mexico’s Some examples of provisions in PTAs dealing with EGS only significant PTA to incorporate environmental provi- will illustrate the point: sions is the one with Japan. Consistent with Japan’s approach so far, the agreement emphasizes soft principles • The agreement between the EU and CARIFORUM of environmental cooperation. includes specific provisions dealing with the commit- The overall picture of environmental provisions in ment of the parties “to make efforts to facilitate trade in PTAs, as it stands today, thus seems to be one of disparate goods and services which the Parties consider to be provisions that depend on the parties involved and their beneficial to the environment. Such products may inter-se relationships. The OECD study of regional trade include environmental technologies, renewable and agreements highlights the diversity of environmental com- energy-efficient goods and services and eco-labelled mitments in PTAs as a matter of potential concern. It notes goods.�17 that countries are likely to be faced with an increasingly • The CAFTA–DR agreement contains provisions for complex problem of managing various levels of environ- cooperative action for developing and promoting envi- mental commitments, and different types of environmen- ronmentally beneficial goods and services.18 tal cooperation programs, under a variety of PTAs to which • The Japan–Mexico PTA refers to cooperation in the they are parties and that this situation may require closer field of “encouragement of trade and dissemination of attention in the near future (OECD 2007, 35). environmentally sound goods and services.�19 To sum up, although open regionalism is a theoretically • The U.S.–Morocco PTA states, “Parties recognize that desirable goal of PTAs, experience to date with these provi- strengthening their co-operative relationship on envi- sions does not permit conclusions as to whether the goal is ronmental matters can encourage increased bilateral achievable through PTAs in general or, more specifically, in trade in environmental goods and services.�20 the context of environmental provisions in PTAs. Whether implementation of these provisions could result in discrimination against third parties has not yet Discrimination under Environmental Provisions in PTAs been tested. Do environmental provisions under PTAs discriminate against nonparties? At the outset, it needs to be emphasized Experiences with Implementation that a PTA, by its very nature, gives preferential market of Environmental Provisions access to the parties to it, as opposed to nonparties. Envi- ronmental provisions in a PTA are structured as obliga- Since most PTAs are very recent, there are not many empiri- tions to adhere to certain standards, but such adherence is cal studies examining the implementation of their provi- not directly linked to preferential market access (except in sions. The OECD study reports that in some countries, the a few U.S. PTAs that provide for suspension of conces- negotiation of a PTA which included environmental com- sions as a consequence of nonadherence to environmen- mitments was a driver of reform or led to the acceleration of tal provisions). Environmental provisions in PTAs are internal reform processes. In Morocco and Chile, the con- by themselves unlikely to result in discrimination against clusion of PTAs with the United States accelerated the adop- third parties in terms of preferential market access. tion of several environmental laws, as well as the overhaul Concerns about preferential access for parties and dis- and codification of environmental legislation (OECD 2007, criminatory treatment of nonparties could arise in the 48). There is, furthermore, anecdotal evidence that the com- context of PTA provisions dealing with environmental mitments in the U.S.–Singapore PTA influenced Singapore 414 Anuradha R. V. to more effectively enforce its domestic laws regarding illegal • Its support for research and symposiums dedicated to wildlife transshipment and to enact a new endangered understanding the effects of trade on the environment species act that brought transshipment under the country’s in North America enforcement mandate and increased penalties tenfold. The • The establishment of public submission mechanisms study observes, however, that the causality for Singapore’s whereby civil society organizations and citizens can law is difficult to determine and that the country’s environ- petition the commission on issues relating to compli- mental efforts may also be related to the ASEAN-wide push ance with the agreement’s environmental obligations to cooperate on stronger implementation and enforcement (Articles 14 and 15 of the NAAEC) of measures against illegal wildlife trade (OECD 2007, 125). • The initiation of pilot funding programs to build the The effectiveness of PTAs in promoting the enforce- environmental capacity of small and medium-size ment of environmental laws has not been definitively enterprises and civil society organizations. studied. Domestic enforcement of environmental laws is often influenced by such factors as availability of adequate Other studies have highlighted some limitations in resources and the presence or absence of effective institu- implementing environmental provisions under the tional mechanisms. Although several PTAs entered into NAAEC. Inadequate resources are blamed for the short- by the United States and the EU specify that the devel- comings. For instance, Gallagher (2004, 74–76) reasons, oped-country partner would provide assistance through using persuasive data and analysis, that environmental financial resources and technical capacity-building sup- conditions worsened in Mexico in the post-NAFTA port, effective enforcement of environmental laws has not period and that Mexico failed to steer the benefits of eco- necessarily resulted. A recent report by the U.S. General nomic integration toward increased environmental pro- Accountability Office (U.S. GAO 2009) studies U.S. PTAs tection. Gallagher points out that the NACEC was not with Chile, Singapore, Morocco, and Jordan and con- “designed to significantly reverse environmental conse- cludes that although the trading partners have improved quences of economic growth in Mexico� and argues that their environmental laws, enforcement has been a chal- the NACEC, with its annual budget of US$9 million, was lenge and that U.S. assistance has been limited in this “insufficient� to make a dent in the problems that cost the regard. The report cites lack of reliable funding as one of Mexican economy US$40 million annually. Similarly, the main reasons for lack of progress on environmental Hufbauer and Schott (2005, 159–60) find that the budget- provisions, and it identifies the absence of effective ele- ary allocation for the NAAEC was inadequate for its man- ments for monitoring and managing environmental proj- date and represents an “insignificant fraction of resources ects as a shortcoming. dedicated to the environment in North America.� Audley and Ulmer (2003, 3) evaluate U.S. trade policy and capac- ity-building initiatives in the context of PTAs and con- NAFTA and the NAAEC clude that it is “challenging to translate good intentions Some of the significant studies and analyses of the effective- into effective policy.� ness of the implementation of environmental provisions under a PTA have looked at the NAAEC, the environmental Technical and Financial Assistance side agreement to NAFTA. These studies highlight some advantages and practical limitations of implementing envi- Technical and financial assistance is intrinsically linked to ronmental provisions under a PTA. A recent paper from the issue of implementation of environmental provisions. the Office of the U.S. Trade Representative (USTR) empha- The financial costs of implementation, and the costs of sizes the contribution of NAFTA and the NAAEC in pro- framing new legal and institutional frameworks, could be viding a clean and healthy environment for residents along especially high for small and medium-size enterprises in the U.S.–Mexican border.21 Independent studies (for developing countries (India 2000). example, Markell and Knox 2003) have commended the Two studies for the World Bank emphasize that to technical cooperation efforts of the North American Com- ensure implementation of the environmental provisions of mission for Environment Cooperation (NACEC) that was a PTA, it would be more rational to address costs and established under the agreement and have identified a capacity-building requirements than to resort to trade number of achievements: sanctions in the event of noncompliance. Wheeler (2000) shows that there is a cost-benefit rationale for requiring • The commission’s role in establishing a pollution release stricter environmental controls while fostering economic and transfer registry (PRTR) in Mexico relationships between countries but that sustained support Environment 415 for building regulatory and institutional capacity is neces- pattern. Translating these provisions into an assessment of sary. Wilson, Otsuki, and Sewadeb (2002) conclude that the economic costs of implementation, and benchmarking imposing higher environmental regulatory standards as the implementation of the provisions, are steps that would part of trade agreements would mean higher costs for non- need to be built into PTA negotiations. OECD countries than for OECD countries and would affect Another area with implications for economic costs and the former’s export competitiveness; adequate policy instru- capacity-building requirements in developing countries is ments are needed to offset that effect. More important, both adherence to developed countries’ product standards and studies conclude that use of trade sanctions is not an effective regulatory requirements, which may be justified on “envi- method of enforcing environmental provisions. They recom- ronmental� grounds. Several of the EU’s SIAs include a mend, instead, sustained support for environmental provi- finding on training and capacity-building requirements in sions and a coordinated approach by the parties involved. this respect. Funding for such activities is sometimes Despite such findings, few PTAs deal with the issues of addressed as part of the preparatory process or during the financial assistance, technical assistance, and capacity implementation of the PTA, but this is not specified as an building, and any such provisions are typically worded in a obligation under the PTA. An example is the Trade Invest- nonbinding and open-ended manner (see OECD 2007, ment Development Program between the EU and India, 86–87). This is true of both U.S. and EU PTAs. (See the which includes a component on capacity-building initia- annexes to this chapter for an overview of relevant provi- tives aimed at equipping Indian laboratories to meet EU sions.) The scope and depth of the provisions vary, from requirements for product testing and certification.22 general statements on cooperation, to support for capacity There has, however, been no comprehensive exercise to building. Although important initiatives for assistance and assess the costs of implementation of such requirements capacity building have been undertaken under most of the and the monitoring and evaluation thereof, although U.S. and EU PTAs, the agreements themselves do not these elements would be necessary for any effective com- incorporate any language on linkages between assistance pliance with such requirements under the framework and capacity building, on the one hand, and legally binding of the PTA. obligations for maintenance of environmental standards under the PTA, on the other hand. Inadequate funding and Conflicting Priorities lack of effective structures for monitoring and managing environmental projects are seen as the main reasons for A curious and interesting issue in relation to implemen- ineffective implementation of environmental provisions tation of environmental standards under PTAs is that of (U.S. GAO 2009). conflicting norms within a PTA, when some provisions The EU–CARIFORUM EPA refers to “development promote and others restrict the space for environmental cooperation� as an imperative; cooperation can take both policy. Studies have indicated that provisions dealing financial and nonfinancial forms (Article 7.1). The provi- with investment under PTAs tend to constrict countries’ sions on cooperation on environment in Article 28.2(a) of policy space for channeling foreign investment into their the EU–Chile agreement refer to the recognition of the territories and also carry high environmental costs relationship between poverty and environment. No studies (Working Group 2008). PTAs that deal with investment have assessed the actions taken as a consequence of such often incorporate provisions on investor-state disputes generic statements. in the investment chapters. In several such disputes, Parties to a PTA need to give greater consideration to regulatory measures enacted for environmental purposes the construction of adequate linkages between commit- have been interpreted as resulting in “expropriation,� ments by a lower-income trading partner to adhere to cer- leading to awards for compensation for private tain environmental standards and the obligation of the investors.23 Some of these disputes challenging environ- higher-income trading partner to commit to capacity- mental regulations as infringing on investors’ rights have building and resource assistance to ensure compliance with arisen under NAFTA, even though that agreement has, in such obligations. An important precedent is the 2004 U.S. the NAAEC, perhaps the strongest side agreement on PTA with CAFTA–DR, which sought to benchmark and environmental issues. The Council for Environmental monitor provisions relating to environmental cooperation. Cooperation (CEC), which is the main institution During the negotiations, each Central American country charged with the task of implementing the NAAEC, is submitted a capacity-building report identifying its priori- required to assist NAFTA’s Free Trade Commission (FTC) ties. The current EU negotiations of EPAs with African, in environment-related matters. Implementation of this Caribbean, and Pacific (ACP) countries follow a similar provision has, however, been limited, and in reality, the 416 Anuradha R. V. CEC and the FTC have been reported to have little con- The Dynamics of Binding and Nonbinding tact (Hufbauer and Schott 2005, 158). The result has been Provisions on Environment the treatment of environmental issues as subsidiary to investor-related concerns. The U.S., Canadian, and EU approaches to environmental The prospect of investor-state arbitration regarding provisions in PTAs represent a blend of binding obligations environmental measures (often referred to as “regulatory and nonbinding principles of intent and cooperation. takings� in investment literature) could be a potential hur- A key area covered by these provisions is domestic environ- dle in the implementation of environmental provisions mental policy in the trading partners. This is a sensitive under a PTA. A study by the United Nations Commission subject for most countries because international agree- on Trade and Development (UNCTAD) finds that it could ments rarely make incursions into domestic policy space. result in a “regulatory chill� because concern about liability U.S. and EU PTAs refer to the obligation of a party to exposure might lead host countries to restrict the ambit of ensure adherence to “high levels of environment protec- or soften a necessary regulation (UNCTAD 2005; see also tion� in its domestic law and policy, but this obligation is Cosbey et al. 2004). often accompanied by an explicit recognition of the right Some recent PTAs, such as the U.S.–Singapore agree- of parties to establish their “own levels of environmental ment, have attempted to address this concern through side protection and environment development policies.�24 This letters dealing with the interpretation of “expropriation� language seems to create a tenuous balance between “high and specifying that, except in “rare� circumstances, nondis- standards� and sovereign rights. The approach using bind- criminatory regulatory measures undertaken for “public ing legal obligations appears to create some flexibility for welfare benefits,� including “environment,� will not consti- parties while holding them to “high� standards, by allowing tute expropriation. How such “rare� circumstances would them to determine their “own level of protection.� be interpreted in an actual dispute has yet to be tested. Because the environmental obligations are broadly worded, scrutiny of specific environmental measures in a country under these provisions would have to be very case Hard- and Soft-Law Approaches to specific. None of the provisions has as yet led to any dis- Environmental Provisions pute resolution situation. Hence, to date, there is no Although it is perhaps too early to assess the implementa- jurisprudence on, for instance, the exact nature of stan- tion of environmental provisions and their effect on multi- dards that would be identified as “high,� or on whether a lateral relations, it is useful to examine in greater detail the particular environmental measure seeks to encourage trade nature of environmental provisions in PTAs and the basis by “lowering levels� of environmental protection, or on for their enforcement. In this section, we examine both whether the resulting trade in goods and services is sup- legally binding (“hard-law�) and nonbinding (“soft-law�) portive of the environment. provisions in PTAs; various approaches to resolution of U.S. PTAs contain provisions referring to the “right of disputes concerning environmental provisions; and the each party to retain discretion with respect to investigatory, effect of the chosen approach on environmental policy in prosecutorial, regulatory, and compliance matters and to members of the PTA. make decisions regarding the allocation of resources to Theoretically, legally binding provisions could be enforcement with respect to other environmental matters expected to be more effective than provisions that are determined to have higher priorities.�25 How this provi- expressed in nonbinding language. In the context of inter- sion is interpreted will need to be examined through the national relations, legally binding language in an agree- development of case law. It is significant that this discre- ment also provides insights into the political will and tion is not unfettered; it is circumscribed by statements intent of parties to be bound by such obligations. At the that the exercise of discretion is to be “reasonable� or that same time, nonbinding provisions do not necessarily indi- the exercise of discretion is to be based on “a bona fide cate a lack of legal intent to implement the provisions, and decision regarding allocation of resources.� The standards they often create significant pathways for dialogue and for determining “reasonable,� and the factors in consider- assessment to enable implementation. Moreover, the ing whether a decision is “bona fide,� have yet to evolve implementation and impact of such provisions do not through jurisprudence. solely depend on the legally binding or nonbinding lan- U.S. PTAs also typically oblige a party to ensure the guage in which the provisions are formulated; preexisting availability of judicial, quasi-judicial, or administrative conditions and the availability of adequate financial and proceedings that are fair, equitable, and transparent and to technical assistance are also important factors. provide for appropriate administrative and procedural Environment 417 protections in accordance with its law, for the enforcement consultations. With regard to state-state mechanisms, the of the party’s environmental laws. What is “fair, equitable PTA provides for consultations, appointment of experts, and transparent,� and what would qualify as “appropriate,� formal dispute resolution, and enforcement. are not defined in the PTAs. Most U.S. PTAs provide that the panel constituted for Each of the approaches discussed above affords a trad- dispute resolution can recommend appropriate remedies ing partner the possibility of examining the domestic con- for noncompliance with environmental obligations tent and implementation of environmental laws in another under the PTA. These remedies can include monetary trading partner, whether or not there is any effect on trade compensation, to be deposited into a fund for environ- between the countries because of such laws. The exact mental initiatives.26 If the losing party does not provide parameters of such scrutiny, and how the balance between such compensation, then, as a matter of last resort, the sovereign discretion and obligations will play out in the other party can suspend tariff concessions. PTAs with event of a dispute, are yet to be tested. differing provisions include the U.S. agreements with Japan’s agreement with Mexico on environment coop- Korea and Israel. The PTA with Korea provides that the eration (a side agreement to the Japan–Mexico PTA) amount of compensation is to be determined through con- represents a variation in that it mandates parties to coop- sultations between the parties rather than by the panel erate. Article 147 of the agreement states, “The Parties, (Article 22.13). The U.S.–Israel PTA states that the recom- recognizing the need for environmental preservation and mendations of the panel on settlement of a dispute is improvement to promote sound and sustainable develop- nonbinding; the winning party has the right to take ment, shall cooperate in the field of environment� (italics “appropriate measures� (Article 19). added). The description of cooperative activities is, how- Several U.S. PTAs also provide for a public submission ever, drafted in nonbinding, inclusive language, allowing process. NAFTA and NAAEC and the CAFTA–DR, scope for the parties to develop further agreements on U.S.–Panama, and U.S.–Peru agreements contain detailed implementation in the future. provisions allowing for members of the public to submit An example of a purely nonbinding approach is the New complaints of noncompliance with environmental obliga- Zealand–Thailand Environmental Cooperation Agreement. tions to an environmental commission or secretariat or Section 4.1 states, “The Arrangement . . . represents a polit- another institutional mechanism constituted for the pur- ical commitment between New Zealand and Thailand but pose.27 This is not a uniform feature and is not found in the does not legally bind either country.� U.S. PTAs with Chile and Singapore.28 Under PTAs that do provide for this process, the commission or secretariat is required to prepare a factual record and forward it to the Approaches to Dispute Resolution environment affairs council, which can make recommen- The remedies available, and the mechanisms for resolution dations. The recommendations are not binding on the of disputes in the event of contravention of environmental state parties to the PTA; the parties would need to trigger obligations, are fundamental indicators of the nature of formal state-state consultations and go through the dis- hard- and soft-law approaches in a PTA. The approaches to pute resolution process under the PTA to secure any dispute resolution for environmental provisions can be enforceable remedies. categorized as follows: The OECD study on regional trade agreements (OECD 2007) analyzed the extent to which public submission pro- • Binding dispute resolution with remedies in the form of cedures under U.S. PTAs have been exercised and con- compensation and possible suspensions of concessions cludes that these procedures have been used to a greater (found in U.S. PTAs) extent than have the state-state mechanisms. The greatest • Binding dispute resolution with remedies in the form of usage of the public submission process has been under the compensation (typical of EU agreements) NAAEC, which has received 68 submissions since 1995.29 The • Dispute resolution with an emphasis on principles of study finds some evidence that in at least two cases, one cooperation and adoption of a nonbinding approach against Mexico and the other against Canada, the process (in PTAs involving New Zealand and Japan). has brought about improved environmental protection, although the exact nature of such improvements was not dis- These approaches are discussed in detail below. cussed (OECD 2007, 127; see also Kirton 2004). Other stud- U.S. PTAs. PTAs entered into by the United States pro- ies (for example, Hufbauer et al. 2000) have criticized these vide for two means of addressing disputes on environ- public submission processes for not being effective enough. mental provisions: state-state procedures, and public The costs and time expended in the NAAEC submission 418 Anuradha R. V. process, and the absence of monitoring of enforcement of the sanctions fail to discriminate between clean and dirty firms recommendations, are regarded as shortcomings. in the affected countries and because they would inevitably EU PTAs. Like U.S. PTAs, PTAs entered into by the EU penalize workers in such countries by reducing opportuni- emphasize consultation and cooperation in environmental ties for jobs and higher wages (Wheeler 2000). Such instru- matters. Under the EPA with CARIFORUM, if consulta- ments would also not be effective in achieving the desired tions fail, the matter can be referred to a committee of environmental goals. experts formed under the EPA. The committee’s findings As a theoretical principle, the availability and enforce- are provided to the consultative committee responsible for ability of binding dispute resolution mechanisms are criti- implementing the EPA. If there is no resolution within nine cal for ensuring the effectiveness of any law. In the case of months from the initiation of consultations, an arbitration environmental provisions, effective implementation of the panel for dispute resolution can be requested by the com- provisions depends on several other factors, such as the plaining party. Measures for compliance may include com- technical and financial resources committed for imple- pensation, but unlike the case of U.S. PTAs, remedies for mentation. This issue is particularly significant for devel- violation of the environmental provisions of the PTA can- oping countries, which need to take into account the not include suspension of concessions. financial and technical costs of implementing environ- New Zealand. The New Zealand–Thailand arrange- mental provisions. Accordingly, it may be advisable to opt ment on environmental cooperation provides that if any for binding dispute resolution as an enforcement measure differences about the arrangement arise between the par- only when all the elements for securing effective implemen- ties, the environment committee set up under the agree- tation of the provisions are built into the PTA, including the ment will attempt to resolve them through consultation technical and financial assistance required for implement- (Section 3). The Trans-Pacific Strategic Economic Part- ing new environmental laws and standards (Wheeler 2000). nership (TPSEP), of which New Zealand is a member, goes Dispute settlement mechanisms emphasizing consulta- a step further, providing that if consultations between the tion and cooperation may therefore be a better option for parties fail, any interested party may refer the issue to the dealing with noncompliance with environmental regula- TPSEP Commission for discussion and that the report tions than binding dispute settlement and sanctions. A emerging from such discussions would need to be imple- phased approach may be desirable: in the first few years, mented (Article 15.6). disputes would be handled through a cooperative approach, Japan. Japan’s PTA with Mexico, which contains princi- and then, after experience with environmental provi- ples of environmental cooperation, specifically states that sions under the PTA has been gained, stronger remedies, the provisions for dispute resolution under the PTA do not such as monetary compensation or trade sanctions, could apply to the chapter containing those principles (Articles be considered. 147 and 148). Conclusions Devising a Suitable Dispute Resolution Mechanism The findings that have emerged from this analysis of key None of the provisions on state-state dispute resolution PTAs that incorporate environmental provisions are described above have ever been resorted to in practice, and summarized here. so their effectiveness, in terms of remedies and conse- quences, has not been tested. The OECD study notes that 1. Approaches to environmental concerns in international “countries may simply hesitate to incur the costs—financial, agreements. Certain developed economies (the United political, and other—of initiating a dispute leading to impos- States, Canada, the EU, New Zealand, and, more ing penalties on another country, even if the letter of the recently, Japan) have pushed for the incorporation of agreement would entitle them to do so� (OECD 2007, 124). environmental provisions into PTAs. Regional eco- From a developing-country perspective, any specific nomic groups of developing countries, as well as approach to dispute settlement under a PTA will need to be Australia, have tended to address environmental issues determined in light of the overall framework of the PTA, in separate agreements or understandings rather the nature of the legally binding obligations it imposes, and than within PTAs. In addition, environmental provi- the resources required to implement those obligations. As sions are sometimes addressed in separate side discussed earlier, studies have shown that the use of trade agreements. No one approach has proved superior in sanctions to enforce environmental obligations in a poorer delivering environmental and economic benefits. country would have unjust consequences because such The choice of side agreements does not necessarily Environment 419 imply that environment is only an incidental issue; tal standards have cost implications. Environmental some of the strongest environmental provisions are standards also affect the export competitiveness of contained in the NAAEC, a side agreement to NAFTA. countries that are required to raise their environmen- 2. Increasing use of environmental provisions in PTAs. tal standards. Yet the costs of compliance are not typi- Environmental provisions are increasingly being cally taken into account in the PTAs in the form of incorporated into PTAs. Developing countries should financial assistance. A fair amount of literature exists plan for negotiations accordingly, as described in on the economic implications for Mexico of imple- “Considerations for Developing Countries in Negoti- menting the trade and environmental provisions aris- ating PTAs,� below. ing from NAFTA. Concrete studies and assessments 3. Pre-PTA environmental impact assessments. The United have not yet been carried out for other PTAs. States, Canada, and New Zealand regularly conduct 7. Implementation of environmental provisions. Since assessments of the potential environmental impacts of environmental provisions in PTAs are of fairly recent PTAs within their own countries. (A few recent U.S. date, their implementation has not been studied in PTAs have considered the trading partner as well.) The detail. In a few instances, stronger environmental laws EU’s approach is to conduct sustainability impact have been enacted and implemented after the coun- assessments that focus on the potential economic, tries entered PTAs. Such developments are also influ- social, and environmental effects of the agreement on enced by other internal and external factors, and it both the EU and the trading partner. Conceptually, the may therefore not always be possible to state defini- EU’s approach offers a better framework for address- tively that the environmental measures were taken ing environmental and sustainable development con- solely or primarily because of the PTA. cerns in both parties to the PTA, but no studies have as Implementation of environmental provisions may yet monitored to what degree the findings of such sometimes conflict with other provisions of the PTA, assessments are translated into the provisions of the such as those relating to investment. Not all PTAs have PTA or their effectiveness in furthering intended envi- adequate linkages for addressing these conflicts. ronmental and development goals. 8. Technical assistance, capacity building, and financial 4. Binding and nonbinding provisions. Both legally bind- assistance. Technology sharing and transfer, technical ing and nonbinding environmental provisions appear assistance and capacity building, and financing of in U.S., Canadian, and EU PTAs. PTAs entered into by environmental initiatives are critical determinants New Zealand and Japan have so far contained mainly of how effectively environmental provisions can be nonbinding language on environmental matters. implemented. Some PTAs have incorporated provi- There has been to date insufficient experience for sions relating to technical and capacity-building assis- making an assessment as to the relative effectiveness of tance; fewer contain provisions on financial assistance. binding and nonbinding principles. Some countries Provisions on these forms of support are not worded have implemented environmental provisions even in legally binding terms. While several important ini- under a nonbinding legal obligation. tiatives for such support have been undertaken under 5. Regional public goods. Countries within a region are both U.S. and EU PTAs, further studies are required to likely to address problems that are transboundary in evaluate their adequacy. nature and affect them all. PTAs, or any other regional 9. Dispute settlement and remedies. Dispute settlement agreements between countries that are geographically provisions in PTAs typically consist of consultations proximate, are more likely to successfully address followed by recourse to formal dispute settlement regional environmental concerns than global issues, mechanisms. Most U.S. PTAs prescribe remedies in the for which multilateral cooperation between many form of monetary compensation for noncompliance countries would be necessary. Nevertheless, some with environmental provisions, with suspension of tar- PTAs between geographically distant countries do seek iff concessions as a last resort. EU PTAs leave the issue of to address environmental issues that are not purely compliance measures to the judicial mechanism hearing regional, such as climate change and protection of bio- the dispute; remedies may include monetary compensa- diversity. Agreements of this kind can complement the tion but not suspension of concessions. A few U.S. PTAs multilateral framework addressing similar issues. contain provisions allowing public submissions to the 6. Economic costs. Enactment of new laws, establishment environmental committee constituted under the PTA. of new institutions, or changes in manufacturing prac- Such submissions result in fact-finding reports but do tices undertaken to ensure adherence to environmen- not lead to any binding rulings. Examples of nonbinding 420 Anuradha R. V. approaches to dispute resolution are found in PTAs ance matters and to make decisions regarding the entered into by Japan and New Zealand, which empha- allocation of resources to enforcement or to other size consultations and a noncontentious approach to environmental matters determined to have higher pri- resolving environmental concerns. orities. This discretion is not unfettered; it is circum- None of the provisions on dispute resolution has scribed by language stating that a party is in compliance been resorted to in practice, and so their effectiveness, with this provision provided that the exercise of discre- in terms of remedies and consequences, has yet to be tion is “reasonable� or that the exercise of discretion tested. Studies have, however, pointed out that trade results from “a bona fide decision regarding allocation sanctions against a poorer country to enforce environ- of resources.� mental compliance do not offer real solutions for envi- • Commitment by each party to strive to ensure that, in ronmental issues and would only exacerbate problems the process of encouraging trade and investment, it does for workers in such countries by limiting employment not waive or derogate from its environmental laws in a opportunities and the prospect of higher wages. manner that weakens or reduces the protections afforded 10. Overall conclusions. Environmental provisions in PTAs in those laws. could potentially be useful in addressing specific envi- • Obligation of each party to ensure that judicial, quasi- ronmental concerns of either party, but only insofar as judicial, or administrative proceedings for the enforce- they are preceded by assessment of the specific envi- ment of its environmental laws are fair, equitable, and ronmental and developmental concerns, are couched transparent and to provide for appropriate administrative in clear language, and are backed by adequate technical and procedural protections in accordance with its law. and financial support for implementation. The PTAs with Chile and Korea clarify that the chapter on environment shall not be construed to empower a Annex A. Provisions on Environment party’s authorities to undertake environmental law in U.S. PTAs enforcement activities in the territory of the other party. The provisions cited here have been extracted mainly from The PTA with Peru allows for such enforcement only in the NAFTA and the NAAEC, the CAFTA–DR agreement, and context of the annex on forest governance. U.S. PTAs with Australia, Bahrain, Chile, Colombia, Morocco, Oman, Panama, and Peru. Provisions found in Environmental Cooperation, Technical only some U.S. PTAs are also described. Assistance, and Capacity Building U.S. PTAs differ in the scope and content of provisions General Principles dealing with environmental cooperation. Environmental • Obligation of both parties to cooperate in the field of cooperation typically pertains to provisions for technical environment assistance and capacity building in relation to environ- • Obligation to ensure that trade and environmental poli- mental standards. Provisions on environmental coopera- cies are mutually supportive tion in NAFTA’s side agreement on environment, the NAAEC, for instance, focus on environmental impacts on trade, as well as labeling, financing, purchase of envi- Right to Establish Levels of Environmental Protection ronmentally friendly products, and so on. The provisions • Right of each party to establish its own levels of envi- on environmental cooperation in the U.S.–Peru PTA ronmental protection and environmental development pertain to protection of biological diversity and indige- policies nous knowledge. The provisions on environmental cooperation in the U.S.–Chile PTA highlight the following areas in which the Provisions with Respect to Domestic United States is committed to cooperate with Chile: Environmental Laws • Obligation of each party to strive to ensure that its laws • Development in Chile of a pollutant release and trans- and policies provide for and encourage high levels of fer register (PRTR)—a publicly available database of environmental protection. chemicals that have been released to air, water, or land • Right of each party to retain discretion with respect to or transferred offsite for further waste management investigatory, prosecutorial, regulatory, and compli- • Reduction of mining pollution in Chile Environment 421 • Improvement of environmental enforcement and com- enforcement of provisions regarding domestic environ- pliance through training and exchange of information mental laws in each party to the PTA. If consultations • Training to help reduce pollution from agricultural between the parties fail, the PTAs provide for engagement practices in Chile of experts to advise on the matter. If this step does not lead • Cooperation to reduce methyl bromide emissions to resolution of the dispute, formal dispute settlement pro- • Improvement of wildlife protection and management in ceedings under the PTA can be commenced. Remedies for Chile. noncompliance with domestic environmental obligations can include monetary compensation, to be deposited in a fund for environmental initiatives. If the losing party does TBTs and SPS Product Standards not provide such compensation, then, as a matter of last Most U.S. PTAs contain provisions relating to formation of resort, the other party can suspend tariff concessions. a specific committee to oversee implementation of SPS and Such provisions have not been appealed to under any of TBT standards and to assist in trade facilitation. The the U.S. PTAs, and so there has been no experience with CAFTA–DR agreement and the U.S. PTAs with Panama possible outcomes. and Peru include provisions on the responsibility of this Public submission process. NAFTA, the CAFTA–DR agree- committee to make recommendations for capacity build- ment, and the U.S.–Panama and U.S.–Peru PTAs include ing in trade and for the design of programs of technical detailed provisions allowing members of the public to sub- assistance to support adherence to SPS and TBT standards. mit complaints of noncompliance with environmental These provisions contain no references to financial assis- obligations to the environmental commission or secre- tance. The U.S.–Korea PTA has a separate provision on tariat set up under the agreement. That body is required to automotive standards and regulations that may be adopted prepare a factual record and provide it to the environmen- for environmental or other reasons. tal affairs council (EAC) constituted under the PTA, which can make recommendations. The recommendations are not binding on the parties to the PTA. Public Participation Most U.S. PTAs contain fairly detailed language, ranging Annex B. Environmental Provisions from binding obligations to best-efforts provisions, on pub- in EU PTAs lic participation. The CAFTA–DR agreement mandates that each party provide for the receipt and consideration of pub- The outline below draws on key provisions of the EU EPA lic communications on matters related to the chapter on with CARIFORUM and on the EU agreements with Chile environment. Each party is also required to convene a and Mexico. national consultative or advisory committee, made up of members of its public, to provide views on matters related to General Principles implementation. The U.S. PTAs with Morocco and Singa- pore contain a less onerous obligation to put in place “proce- • Emphasis on the objectives of sustainable development dures� for dialogue with the public and to make “best and protection of the environment. efforts� to respond to requests by members of the public with regard to implementation of environmental provisions. Provisions with Respect to Domestic Environmental Laws • Right of the parties to regulate in order to achieve their Institutional Arrangements own levels of domestic environmental and public health All U.S. PTAs provide for a separate body, termed a joint protection and their own sustainable development pri- committee or environmental affairs council, that is orities and to adopt or modify their environmental laws responsible for overseeing implementation of the provi- and policies accordingly sions of the agreement and providing recommendations • Obligation of parties to seek to ensure that their own for implementation. environmental and public health laws and policies pro- vide for and encourage high levels of environmental and public health protection and to strive to continue to Dispute Resolution improve those laws and policies State-state consultations. All U.S. PTAs except that with Jor- • Recognition that “special needs and requirements of dan specify a separate dispute resolution mechanism for CARIFORUM States shall be taken into account in the 422 Anuradha R. V. design and implementation of measures aimed at pro- The EU–CARIFORUM agreement sets forth detailed tecting environment and public health that affect trade provisions with specific relevance for environmental between the parties� standards: • Obligation not to encourage trade or foreign direct investment or to enhance or maintain a competitive • Technical assistance to help producers meet relevant advantage by lowering the level of protection pro- product and other standards applicable in EU markets vided by domestic environmental and public health • Promotion and facilitation of private and public volun- legislation or by derogating from or failing to apply tary and market-based programs, including relevant such legislation. labeling and accreditation schemes. Dispute Resolution Obligations for Technical Assistance and Capacity Building EU PTAs emphasize consultation and cooperation in The EU–CARIFORUM EPA and the EU’s agreements with environmental matters. The EPA with CARIFORUM Chile and Mexico contain provisions on “facilitating sup- provides for consultations. If consultations fail, the mat- port� to ensure technical assistance for implementation of ter can be referred to a committee of experts formed the provisions of the agreement, including the provisions under the EPA, and the committee’s findings are then relating to environmental obligations. The EU–Chile agree- provided to the consultative committee responsible for ment also emphasizes cooperation on projects to reinforce implementing the EPA. If no resolution is achieved Chile’s environmental structures and policies. within nine months of the initiation of consultations, an Article 190 of the EU–CARIFORUM agreement con- arbitration panel for dispute resolution can be requested tains the most detailed provisions. Examples include the by the complaining party. Measures to enforce compli- parties’ obligations to ensure cooperation for “facilitating ance may include compensation, but not suspension of support� with respect to concessions. • Capacity building for environmental management in tourism areas at the regional and local levels Notes • Technical assistance to help producers meet relevant The author thanks Jean-Pierre Chauffour and Muthukumara Mani of the product and other standards in EU markets World Bank for their invaluable comments and guidance. The author also expresses her sincere thanks to John Strand, Charles Di Leva, and Sachiko • Technical assistance and capacity building, in particular Morita for their comments, which were very helpful in refining the con- for the public sector, for the implementation and tents and focus of the study. Thanks are also due to Sumiti Yadava and enforcement of multilateral environmental agreements Deepak Raju for their research assistance. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the • Facilitation of trade between the parties in natural views of the reviewers or the World Bank. Any errors or oversights are resources (including timber and wood products) from attributable only to the author. legal and sustainable sources 1. A pilot database that enables a comparison of the provisions of the 30 PTAs analyzed in the APEC study is available at www.mincetur • Assistance to producers to develop or improve produc- .gob.pe/apec_fta. The OAS database is at http://www.oas.org/dsd/ tion of goods and services that the parties consider to be EnvironmentLaw/EnvlawDB/Default.htm. Thirty-five countries in the beneficial to the environment Americas are OAS members. • Promotion and facilitation of public awareness and 2. Mani and Jha (2006) explore the impact of trade liberalization in Vietnam and the nature of regulatory interventions that would be education programs related to environmental goods required to address environmental harm. and services in order to foster trade in such products 3. Working Group (2004). The trade policy context covered in the between the parties. report includes NAFTA, the Free Trade Area of the Americas (FTAA), the Central America Free Trade Agreement (CAFTA), the U.S.–Chile Free Trade Agreement, and negotiations toward a pact between the United States and the Andean nations. TBT and SPS Product Standards 4. New Zealand Ministry of Foreign Affairs and Trade, “Framework for Integrating Environment Issues into Free Trade Agreements,� http:// The EU’s PTAs with Chile and Mexico, and the EPA www.mfat.govt.nz/Trade-and-Economic-Relations/0—Trade-archive/WTO/ recently negotiated with the CARIFORUM countries, 0-environment-framework.php#tradepolicies (accessed May 5, 2009). 5. The general exceptions are embodied in Article XX of the General include provisions for technical assistance to help produc- Agreement on Tariffs and Trade (GATT) and Article XIV of the General ers meet relevant product and other standards. Agreement on Trade in Services (GATS). Environment 423 6. United States–Standards for Reformulated and Conventional member countries. A stricter view (Bhagwati 2006) is that open regional- Gasoline, WT/DS/1 (May 20, 1996); United States–Import of Certain ism would require members of a regional group to undertake trade liber- Shrimp and Shrimp Products, WT/DS/58 (November 6, 1998); United alization in concert and to extend it worldwide on a most favored nation States–Import of Certain Shrimp and Shrimp Products–Recourse to basis. The concept of open regionalism has been discussed to a large Article 21.5 of the Dispute Settlement Understanding, WT/DS/58 extent in the context of the APEC; see for example, Bergsten (1997). (November 21, 2001). 14. For the Chile–China agreement (Chile y China Firmaron Acuerdo 7. The documents, respectively, are the U.S.–Singapore Free Trade Ambiental), see http://www.sice.oas.org/TPD/CHL_CHN/CHL_CHN Agreement (FTA) (May 6, 2003), http://www.ustr.gov/Trade_Agreements/ _e.ASP (accessed May 5, 2009). The joint statement by Chile and Japan, Bilateral/Singapore_FTA/Final_Texts/Section_Index.html (accessed May March 27, 2007, is available at http://www.direcon.cl/documentos/ 5, 2009); the U.S.–Chile FTA (June 6, 2003), http://www.ustr.gov/ japon/joint.pdf (accessed May 5, 2009). The other agreements are, respec- Trade_Agreements/Bilateral/Chile_FTA/Final_Texts/Section_Index.html tively, the Chile–Mexico Free Trade Agreement (April 17, 1998), http://www (accessed May 5, 2009); the U.S.–Australia FTA (May 18, 2004), .sice.oas.org/trade/chmefta/indice.asp (accessed May 5, 2009); Acuerdo de http://www.ustr.gov/Trade_Agreements/Bilateral/Australia_FTA/Final_Text/ Libre Comercio Chile–Perú (August 22, 2009), http://www.sice Section_Index.html (accessed May 5, 2009); the U.S.–Bahrain FTA .oas.org/Trade/CHL_PER_FTA/Index_s.asp (accessed May 5, 2009); Chile– (September 14, 2004), http://www.ustr.gov/Trade_Agreements/Bilateral/ Korea Free Trade Agreement, (February 15, 2003), http://www.sice. Bahrain_FTA/final_texts/Section_Index.html (accessed May 5, 2009); the oas.org/Trade/Chi-SKorea_e/ChiKoreaind_e.asp (accessed May 5, 2009). U.S.–Morocco FTA, http://www.moroccousafta.com/studies.htm; and the 15. Paragraph 31(iii) of the Doha Declaration is focused on the reduc- CAFTA–DR FTA, http://www.ustr.gov/trade-agreements/free-trade- tion or, as appropriate, the elimination of tariff and nontariff barriers to agreements/cafta-dr-dominican-republic-central-america-fta. The mem- environmental goods and services. bers of CAFTA are Costa Rica, El Salvador, Guatemala, Honduras, and 16. For a good overview on the EGS negotiations at the WTO, see Nicaragua. World Bank (2008, ch. 4). 8. EU–Mexico Economic Partnership, Political Coordination and 17. Article 183(5), EU–CARIFORUM EPA. Article 190 refers to pro- Cooperation Agreement (December 8, 1997), http://www.fco.gov motion and facilitation of public awareness and education programs with .uk/resources/en/pdf/pdf13/fco_ref_ts13-01a_coop_eco (accessed May 5, respect to environmental goods and services to foster trade in such prod- 2009); EU–Chile Association Agreement (November 18, 2002), http:// ucts between the parties. trade.ec.europa.eu/doclib/html/111620.htm (accessed May 5, 2009); 18. Article 17.9(3), CAFTA–DR FTA (August 5, 2004), http://www Euro-Mediterranean Agreement Establishing an Association between .ustr.gov/trade-agreements/free-trade-agreements/cafta-dr-dominican- the European Communities and their Member States, of the One Part, republic-central-america-fta. and the State of Israel, of the Other Part (November 20, 1995), http:// 19. Article 147(1)(c), Agreement between Japan and the United europa.eu/eur-lex/pri/en/oj/dat/2000/l_147/l_14720000621en00030156.pdf, Mexican States for the Strengthening of the Economic Partnership (accessed May 5, 2009); Euro-Mediterranean Interim Association (September 17, 2004), http://www.mexicotradeandinvestment.com/ Agreement on Trade and Cooperation between the European Commu- agreement.html (accessed May 5, 2009). nity, of the One Part, and the Palestine Liberation Organization for the 20. Article 17.3(7), U.S.–Morocco Free Trade Agreement (June 15, Benefit of the Palestinian Authority of the West Bank and the Gaza Strip, 2004), http://www.ustr.gov/Trade_Agreements/Bilateral/Morocco_FTA/ of the Other Part (1997), http://eur-lex.europa.eu/LexUriServ/LexUriServ FInal_Text/Section_Index.html (accessed May 5, 2009). .do?uri=CELEX:21997A0716(01):EN:HTML (accessed May 5, 2009); 21. Office of the United States Trade Representative, “NAFTA Facts,� Euro-Mediterranean Agreement Establishing an Association between the March 19, 2008, http://www.ustr.gov (accessed April 21, 2009). According to European Communities and their Member States, of the One Part, and the note, as of March 2008, nearly US$1 billion had been provided for 135 the Arab Republic of Egypt, of the Other Part (June 25, 2001), environmental infrastructure projects, with an estimated cost of US$2.89 http://ec.europa.eu/external_relations/egypt/aa/06_aaa_en.pdf (accessed billion, and US$33.5 million in assistance and US$21.6 million in grants had May 5, 2009). been allocated to more than 450 other border environmental projects. The 9. The following are the members of the organizations named: note also emphasizes that the Mexican government has made substantial new Mercosur—Argentina, Brazil, Paraguay, and Uruguay; Andean Commu- investments in environmental protection, increasing its federal budget for the nity—Bolivia, Colombia, Ecuador, and Peru; AFTA—Brunei Darussalam, environmental sector by 81 percent between 2003 and 2008. The USTR refers Cambodia, Indonesia, Lao People’s Democratic Republic, Malaysia, the to two funds set up for this purpose: the NACEC Fund for Pollution Preven- Philippines, Singapore, Thailand, and Vietnam; SAARC—Bangladesh, tion Projects in Mexican Small and Medium Enterprises (FIRPEV), and the Bhutan, India, the Maldives, Nepal, Pakistan, and Sri Lanka; CARICOM— North American Fund for Environmental Cooperation (NAFEC). Antigua and Barbuda, Bahamas, Barbados, Belize, Dominica, Grenada, 22. See “Delegation of the European Union to India,� http:// Guyana, Haiti, Jamaica, Montserrat, St. Kitts and Nevis, St. Lucia, St. Vincent www.delind.ec.europa.eu/en/eco/tidp.htm#WHAT%20DOES%20TIDP% and the Grenadines, Suriname, and Trinidad and Tobago; EAC—Kenya, 20AIM%20TO%20ACHIEVE (accessed May 29, 2009). Tanzania, and Uganda; SADC—Angola, Botswana, Lesotho, Malawi, 23. See, for example, Metalclad Corporation v. United Mexican States, Mozambique, Swaziland, Tanzania, Zambia, and Zimbabwe. International Centre for Settlement of Investment Disputes (ICSID) 10. The Mekong River runs 4,800 kilometers, from the Tibetan Tribunal Decision, August 30, 2000. The case concerned the alleged plateau through China’s Yunnan Province and on through Myanmar, expropriation caused by an ecological decree of Mexico that interfered Thailand, Cambodia, the Lao PDR, and Vietnam. Development of the with the operation of a hazardous waste landfill constructed by a sub- basin presents a variety of transboundary environmental issues for the sidiary of the complainant. The tribunal found that indirect expropriation riparian states. had occurred and awarded compensation under NAFTA, Chapter 11. See 11. These centers are the Coastal Zone Management Center in the also Higgins (1982); OECD (2004). Maldives and the SAARC Forestry Center in Bhutan. See the SAARC 24. For example, Article 3, NAAEC; Article 17.1, U.S.–Panama PTA; website, http://www.saarc-sec.org/?t=2.5. Article 17.1, U.S.–Australia PTA; Article 17.1, U.S.-Morocco PTA; Article 12. Bhagwati popularized the term “spaghetti bowl� in his earlier 184.1, EU–CARIFORUM EPA. writings. 25. For example, Article 17.3, U.S.–Panama PTA; Article 18.2, U.S.– 13. For instance, an International Monetary Fund working paper CAFTA–DR PTA; Article 19.2, U.S.–Australia PTA. (Frankel and Wei 1998) proposes that the degree of liberalization for 26. For example, Article 20.17, in the CAFTA–DR, U.S.– Panama, imports from nonmembers to a PTA need not be as high as that for and U.S.–Colombia agreements. 424 Anuradha R. V. 27. Articles 14 and 15, NAAEC; Article 17.7, CAFTA–DR PTA; Article the Use and Potential of International Investment Agreements. Win- 14, Canada–Chile PTA. nipeg, Manitoba, Canada: International Institute for Sustainable 28. In the case of the Chile–U.S. PTA, each party is required to make Development. best efforts to accommodate requests for consultations by persons or Daly, Herman. 1993a. “The Perils of Free Trade.� Scientific American 269 organizations (Article 19.4.2). (5, November): 24–29. 29. See the Commission for Environmental Cooperation website, ———. 1993b. “Problems with Free Trade: Neoclassical and Steady State http://www.cec.org/citizen/status/index.cfm?varlan=english (accessed April Perspectives.� In Trade and the Environment: Law, Economics, and 21, 2009). Policy, ed. Durwood Zaelke, Paul Orbuch, and Robert F. Housman. Washington, DC: Island Press. Environment Canada. 2007. “Evaluation of Canada’s Participation in the Bibliography Commission for Environmental Cooperation (CEC).� Environment Canada, Gatineau, Quebec, Canada. http://www.ec.gc.ca/doc/ae-ve/ Anríquez, Gustavo. 2002. “Trade and Environment: An Economic Litera- CEC-CCE/toc_eng.htm. ture Survey.� Working Paper 02-16, Department of Agricultural and European Commission. 2006. “Handbook for Trade Sustainability Impact Resource Economics, University of Maryland, College Park, MD. Assessment.� European Commission, External Trade, Brussels. http://ageconsearch.umn.edu/bitstream/28598/1/wp02-16.pdf. Frankel, Jeffrey, and Shang-Jin Wei. 1998. “Open Regionalism in a World APEC (Asia-Pacific Economic Cooperation). 2008. “Identifying Conver- of Continental Trade Blocs.� IMF Working Paper WP/98/10, Interna- gences and Divergences in APEC RTAs/FTAs.� Report 2008/AMM/ tional Monetary Fund, Washington, DC. 010, presented at the 20th APEC Ministerial Meeting, Lima, Peru, Fredriksson, Per G., ed. 1999. Trade, Global Policy, and the Environment. November 19–20. World Bank Discussion Paper 402. Washington, DC: World Bank. ASEAN (Association of Southeast Asian Nations). 1997. “ASEAN Vision Friedman, Joseph, Daniel A. Gerlowski, and Johnathan Silberman. 1992. 2020.� ASEAN Secretariat, Jakarta. http://www.aseansec.org/1814.htm. “What Attracts Foreign Multinational Corporations? Evidence from Accessed April 21, 2009. Branch Plant Location in the United States.� Journal of Regional Science ———. 2007. “Combating Haze in ASEAN: Frequently Asked Questions.� 32 (4, November): 403–18. ASEAN Secretariat, Jakarta. http://www.aseansec.org/Fact%20Sheet/ Gallagher, Kevin P. 2004. Free Trade and the Environment: Mexico, NAFTA ASCC/2007-ASCC-001.pdf. and Beyond. Stanford, CA: Stanford University Press. ———. 2009. “Minerals and Energy.� ASEAN Secretariat, Jakarta. http:// Harrison, Glenn W., Thomas F. Rutherford, and David G. Tarr. 2001. www.aseansec.org/4948.htm. “Chile’s Regional Arrangements and the Free Trade Agreement of the Audley, John, and Vanessa Ulmer. 2003. “Strengthening Linkages Americas: The Importance of Market Access.� Policy Research Work- between U.S. Trade Policy and Environmental Capacity Building.� ing Paper 2634, World Bank, Washington, DC. Carnegie Paper 40, Carnegie Endowment for International Peace, Heal, Geoffrey. 1998. “New Strategies for the Provision of Global Public Washington, DC. Goods: Learning from International Environmental Challenges.� Baldwin, Richard, and Patrick Low. 2009. “Introduction.� In Multilateral- Columbia Graduate School of Business Working Paper, Columbia izing Regionalism: Challenges for the Global Trading System, ed. University, New York. Richard Baldwin and Patrick Low. Cambridge, U.K.: Cambridge Uni- Higgins, R. 1982. “The Taking of Property by the State: Recent Develop- versity Press. ments in International Law.� Recueil des Cours (Académie de Droit Bergsten, C. Fred, ed. 1997. Whither APEC? The Progress to Date and International) 176: 276–77. Agenda for the Future. Washington, DC: Institute for International Hufbauer, Gary Clyde, Daniel C. Esty, Diana Orejas, Luis Rubio, and Economics. Jeffrey J. Schott. 2000. NAFTA and the Environment: Seven Years Bhagwati, Jagdish. 1993. “The Case for Free Trade.� Scientific American 269 Later. Policy Analyses in International Economics 61. Washington, (5, November): 41–57. DC: Peterson Institute for International Economics. ———. 2004. In Defense of Globalization. New York: Oxford University Hufbauer, Gary Clyde, and Jeffrey J. Schott, assisted by Paul L. E. Grieco Press. and Yee Wong. 2005. NAFTA Revisited: Achievements and Challenges. ———. 2006. “Why Asia Must Opt for Open Regionalism on Trade,� Washington, DC: Peterson Institute for International Economics. Financial Times (Asian edition), November 3: 12. Hufbauer, Gary, and Yee Wong. 2005. “Prospects for Regional Free Trade ———. 2008. Termites in the Trading System: How Preferential Agreements in Asia.� Working Paper Series WP 05-12, Peterson Institute for Inter- Undermine Free Trade. Oxford, U.K.: Oxford University Press. national Economics, Washington, DC. Bhagwati, Jagdish N., and T. N. Srinivasan. 1996. “Trade and Environment: IISD (International Institute for Sustainable Development) and UNEP Does Environmental Diversity Detract from the Case for Free Trade?� (United Nations Environment Programme). 2005. Environment In Fair Trade and Harmonization: Economic Analysis, ed. Jagdish N. and Trade: A Handbook. 2nd ed. Winnepeg, Manitoba, Canada: Bhagwati and Robert E. Hudec. Cambridge, MA: MIT Press. IISD. Colyer, Dale. 2008. “Green Trade Agreements: Comparison of Canada, India. 2000. “The Study of the Effects of Environmental Measures on U.S. and WTO.� Presented at Joint Annual Meeting of the Canadian Market Access.� WT/CTE/W/177, submission to the Doha Ministerial Agricultural Economics Society and Northeastern Agricultural and Conference, World Trade Organization, Geneva, October 27. Resource Economic Association, Quebec City, June 30–July 1. http:// ———. n.d. “India and the World Trade Organization.� Ministry of Com- ageconsearch.umn.edu/bitstream/37329/2/NAREA%20Paper%202008% merce and Industry, New Delhi. http://commerce.nic.in/trade/ 20revised.pdf. international_trade_enviro_te.asp. Cosbey, Aaron. 2007. “Regional Arrangements.� In Trade and Develop- Khan, S. R., et al., 2004. “Southern Agenda for Trade and Environment.� ment: A Resource Book, ed. Adil Najam, Mark Halle, and Ricardo Background paper prepared for ICTSD/IISD/RING Southern Agenda Meléndez-Ortiz, 169–78. International Institute for Sustainable Phase II Project, Regional Consultation for South/Southeast Asia, Development (IISD), International Centre for Trade and Sustainable Kandalama, Sri Lanka, January. Development (ICTSD) and the Regional and International Networking Kirton, John. 2004. “NAFTA Dispute Settlement Mechanisms: An Group (The Ring). http://www.iisd.org/pdf/2007/trade_and_env.pdf. Overview.� Prepared for Experts’ Workshop on “NAFTA and Its Impli- Cosbey, Aaron, Howard Mann, Luke Eric Peterson, and Konrad von cations for ASEAN’s Free Trade Area,� Asian Institute, Munk Centre Moltke. 2004. Investment and Sustainable Development: A Guide to for International Studies, Toronto, May 27. Environment 425 Krugman, Paul. 1997. “What Should Trade Negotiators Negotiate About?� Rodríguez, Enrique Valerdi. 2009. “The European Union Free Trade Journal of Economic Literature 35 (March): 113–20. Agreements: Implications for Developing Countries.� Working Paper Langton, Danielle. 2008. “United States–Southern African Customs 8/2009, Real Instituto Elcano, Madrid. Union (SACU) Free Trade Agreement Negotiations: Background and SAARC (South Asian Association for Regional Cooperation). “Areas of Potential Issues.� CRS Report for Congress, RS21387, Congressional Cooperation.� SAARC Secretariat, Kathmandu. http://www.saarc-sec Research Service, Library of Congress, Washington, DC. .org/?t=2#. Less, Cristina Tébar, and Joy Aeree Kim. 2008. “Checklist for Negotiators of ———. “Regional Centres.� SAARC Secretariat, Kathmandu. http://www Environmental Provisions in Regional Trade Agreements.� OECD Trade .saarc-sec.org/Regional-Centers/12/. and Environment Working Paper 2008-2, Organisation for Economic Tiemann, Mary. 2004. “NAFTA: Related Environmental Issues and Initia- Co-operation and Development, Paris. http://fiordiliji.sourceoecd.org/ tives.� CRS Report for Congress, Congressional Research Services, vl=27461473/cl=19/nw=1/rpsv/cgi-bin/wppdf?file=5kzc06k3c823.pdf. Library of Congress, Washington, DC. http://assets.opencrs.com/rpts/ Low, Patrick, ed. 1992. International Trade and the Environment. World 97-291_20040323.pdf. Bank Discussion Paper 159. Washington, DC: World Bank. UNCTAD (United Nations Conference on Trade and Development). Mani, Muthukumara, and Shreyasi Jha. 2006. “Trade Liberalization and 2005. Investor-State Disputes Arising from Investment Treaties: A the Environment in Vietnam.� Policy Research Working Paper 3879, Review. UNCTAD Series on International Investment Policies for World Bank, Washington, DC. Development. New York: United Nations. Mani, Muthukumara, Sheoli Pargal, and Mainul Huq. 1997. “Does Envi- U.S. GAO (General Accountability Office). 2009. “Four Free Trade Agree- ronmental Regulation Matter? Determinants of the Location of New ments GAO Reviewed Have Resultved in Commercial Benefits, but Manufacturing Plants in India in 1994.� Policy Research Working Challenges on Labor and Environment Remain.� Highlights of GAO- Paper 1718, World Bank, Washington, DC. 09-439 (July), GAO, Washington, DC. Markell, David L., and John H. Knox, eds. 2003. Greening NAFTA: The USTR (Office of the U.S. Trade Representative). 2008. “NAFTA—Myth vs. North American Commission for Environmental Cooperation. Stanford, Facts.� March, USTR, Washington, DC. http://www.ustr.gov/sites/ CA: Stanford University Press. default/files/NAFTA-Myth-versus-Fact.pdf. Maskus, Keith E., Tsunehiro Otsuki, and John S. Wilson. 2005. “The Cost ———. 2009. “The President’s Trade Policy Agenda.� USTR, Washington, of Compliance with Product Standards for Firms in Developing DC. http://www.ustr.gov/sites/default/files/uploads/reports/2009/ Countries: An Econometric Study.� Policy Research Working Paper asset_upload_file810_15401.pdf. 3590, World Bank, Washington, DC. Wheeler, David. 2000. “Racing to the Bottom? Foreign Investment and Air Morici, Peter. 2002. Reconciling Trade and the Environment in the World Trade Pollution in Developing Countries.� Policy Research Working Paper Organization. Washington, DC: Economic Strategy Institute. http://www 2524, World Bank, Washington, DC. .econstrat.org/images/ESI_Research_Reports_PDF/reconciling%20 Wilson, John S., Tsunehiro Otsuki, and Mirvat Sewadeb. 2002. “Dirty trade %20and%20the%20environment%20in%20the%20world%20 Exports and Environmental Regulation: Do Standards Matter to Trade?� trade%20organization%20%28peter%20morici%29.pdf. Policy Research Working Paper 2806, World Bank, Washington, DC. Najam, Adil, Mark Halle, and Ricardo Meléndez-Ortiz, eds. 2007. Trade Working Group (Working Group on Development and the Environment and Development: A Resource Book. International Institute for Sus- in the Americas). 2004. Globalization and the Environment: Lessons tainable Development (IISD), International Centre for Trade and from the Americas. Washington, DC: Heinrich Böll Foundation North Sustainable Development (ICTSD) and the Regional and Interna- America. tional Networking Group (The Ring). http://www.iisd.org/pdf/ ———. 2008. Foreign Investment and Sustainable Development: Lessons 2007/trade_and_env.pdf. from the Americas. Washington, DC: Heinrich Böll Foundation North OECD (Organisation for Economic Co-operation and Development). 2004. America. “‘Indirect Expropriation’ and the ‘Right to Regulate’ in International World Bank. 2008. International Trade and Climate Change: Economic, Investment Law.� Working Papers on International Investment 2004/4, Legal, and Institutional Perspectives. Washington, DC: World Bank. Directorate for Financial and Enterprise Affairs, OECD, Paris. WTO (World Trade Organization) Secretariat. 1997. “Environmental Ben- ———. 2007. Environment and Regional Trade Agreements. Paris: OECD. efits of Removing Trade Restrictions and Distortions.� WT/CTE/W/67, Porter, Michael E. 1991. “America’s Green Strategy.� Scientific American WTO, Geneva. 264 (4, April): 168. ———. 2004. “Trade and Environment at the WTO: Background Docu- PricewaterhouseCoopers. 2007. “Sustainability Impact Assessment of ment.� WTO, Geneva. the EU–ACP Economic Partnership Agreements: Key Findings, Zaelke, Durwood, Paul Orbuch, and Robert F. Housman, eds. 1993. Trade Recommendations and Lessons Learned.� Prepared for the European and the Environment: Law, Economics, and Policy. Washington, DC: Commission, PricewaterhouseCoopers, Paris. Island Press. 20 Labor Rights Kimberly Ann Elliott Within international trade rules, the only reference to • Effective abolition of child labor labor rights is not a positive obligation, but an exception • Elimination of discrimination in employment.1 that permits member countries to ban imports produced with prison labor. Despite increased attention to the About the same time as NAFTA was signed, U.S. nego- issue with the spread of globalization, member countries tiators began pushing for discussion of labor rights at the have not agreed to any new provisions on labor rights in WTO. Developing countries, however, have successfully the World Trade Organization (WTO), primarily because opposed creation of a study group or the inclusion of of the collective opposition of developing countries. By labor rights issue in the Doha Round of multilateral contrast, such provisions are becoming increasingly com- trade negotiations. mon in bilateral negotiations, particularly those between Except for Chile, developing countries generally do large, powerful developed countries and smaller, poorer not include labor provisions in their agreements with developing countries. one another, either. With one recent exception, enforce- Demands to include labor rights in trade agreements able labor rights provisions are absent from agreements began with the United States and the negotiation of the that involve Asian countries and do not include the North American Free Trade Agreement (NAFTA) with United States. Mexico and Canada in the early 1990s. NAFTA represented Traditionally, agreements to which the European Union a departure from past U.S. practice in two respects. It was (EU) is a party often have language addressing human the first U.S. preferential trade agreement (PTA) with a rights, not worker rights specifically, but that may be developing country (and only the third U.S. PTA of any changing. Canada is the only country besides the United kind), and, to ensure legislative approval, it included side States that includes labor provisions in its agreements with agreements to protect worker rights, as well as the environ- developing countries, but this is always done in side agree- ment. Every U.S. PTA since then has incorporated legally ments, and with institutional mechanisms that emphasize binding and enforceable provisions on labor rights in the cooperation rather than sanctions. text of the agreement. These provisions, in U.S. PTAs and Labor standards language, where it exists outside U.S. in those of other countries taking similar approaches, are trade agreements, is usually either hortatory or only generally based on the Declaration on Fundamental Prin- nominally binding, with no or only weak enforcement ciples and Rights at Work that was issued by the Interna- measures. Nor do most agreements, including those that tional Labour Organization (ILO) in 1998. (U.S. and involve the United States, provide much in the way of Canadian agreements refer, in addition, to “acceptable meaningful cooperation or technical assistance for conditions of work� relating to wages, hours, and health improving labor standards. Moreover, those U.S. PTAs that and safety.) contain enforceable labor provisions and binding dispute The ILO document lists four core labor standards as settlement mechanisms generate few formal complaints, deserving of universal application. They are, in brief, and no PTA has yet come close to the use of trade or other sanctions to enforce the labor provisions. The • Freedom of association and “effective recognition� of potential for exerting leverage is, in fact, greatest during the right to collective bargaining the negotiation phase, when American trade negotiators • Elimination of forced labor sometimes condition the conclusion of an agreement on 427 428 Kimberly Ann Elliott changes in labor laws to bring them closer to international There is, however, little compelling evidence in the norms (Hornbeck 2009). empirical literature either of a race to the bottom in labor This chapter begins with a review of the debate over the standards or of protectionist abuse of trade-labor linkages economics of linking trade and labor standards. It then where they exist. The effect of higher standards on compet- examines actual practice with respect to labor rights in itiveness is more complicated than is usually assumed, and PTAs and surveys how those provisions have been imple- any higher costs of compliance that do materialize are mented and enforced. The discussion concludes with les- often offset by higher productivity (Freeman and Medoff sons from this experience for developing countries that are 1984). A meta-survey of the literature published by the considering such agreements. World Bank in 2002 found “little systematic difference in performance between countries that enforce [union rights] and countries that do not� (Aidt and Tzannatos The Economics of Labor Rights in PTAs 2002, 4). Nor does the experience of nearly two decades The policy debate and the accompanying literature on offer support for the concern that trade sanctions to whether trade and labor standards should be linked, and if enforce labor standards are simply protectionism in dis- so, how, is extensive.2 One strand of the literature focuses guise. A 2003 review of trade-labor standards linkages, on the relationship, if any, between labor laws and regula- particularly under the generalized system of preferences tions, on the one hand, and international trade and invest- (GSP) program, found little evidence that labor rights ment flows, on the other. Alongside the discussion of the have been introduced or used for protectionist purposes economics of trade and labor markets, there is a political- (Elliott and Freeman 2003, ch. 4). Since that review, sev- economy debate about the appropriate role for the WTO eral new PTAs have been implemented, and several more in enforcing international labor standards. The central years of experience with the NAFTA agreement have been issue in this debate, which is relevant to the inclusion of amassed, and no trade or other sanction has yet been labor provisions in PTAs, is whether labor standards imposed because of violations of labor standards embed- should be enforced through trade sanctions or whether ded in trade agreements. that possibility would lead to protectionist abuse (see, An alternative analysis suggests that labor standards and e.g., CITEE 2004). globalization are complements, with each compensating In broad terms, free trade advocates argue that trade for weaknesses in the other.3 Globalization can contribute encourages growth, which, in turn, will bring about higher to growth, and the jobs it creates are generally better than wages and better working conditions. In this view, no spe- those in agriculture or the informal sector, but it is also cial attention to labor standards in trade agreements is associated in some cases with increased income inequality. needed. Some argue that pushing developing countries to Those arguing that no special attention to labor standards adopt higher standards would make them uncompetitive is needed also tend to ignore the disproportionate influence in export markets and lead to fewer jobs and worse condi- that multinational corporations have on trade negotiations tions. On the other side, labor standards advocates, led by and the possibilities that exist for improving conditions unions and many human rights groups, maintain that for workers in developing countries at low cost and with- competition to attract foreign investment or to capture a out jeopardizing economic growth. Promoting global larger export market share causes countries to suppress labor standards simultaneously with trade could spread labor standards, or at least not to raise them, leading to a the benefits of globalization more broadly, discourage the “race to the bottom.� worst abuses of workers, and increase public support for Many of those who assert a negative link between trade trade agreements. and labor rights also view the ILO with skepticism, regard- One key to realizing the complementarities and avoid- ing it as toothless and ineffective. The 1995 launch of the ing potential negative effects is to distinguish “core� labor WTO, with a legally binding dispute settlement system and standards from others. Some labor standards—for exam- sanctions for noncompliance, seemed an attractive alterna- ple, wages and health and safety regulations—will clearly tive to those concerned that globalization would lead to a have to vary with the countries’ level of development and race to the bottom. Most developing-country governments the local standards of living. However, the four core labor and free traders viewed more forceful international standards set forth in the 1998 ILO Declaration are meant enforcement of labor standards as anathema, convinced to be upheld by all countries, regardless of their level of that it would hinder developing countries’ development development. As noted at the beginning of this chapter, the and poverty reduction prospects and would be used by core standards are freedom of association and the right to developed countries for protectionist purposes. organize and bargain collectively; freedom from forced Labor Rights 429 labor; the effective abolition of child labor, beginning with with export sectors that are highly price competitive, the “worst forms�; and nondiscrimination.4 require few skills, and are geographically mobile, such as These standards are part of the framework rules that clothing and footwear. govern labor market transactions; they do not specify par- This scenario suggests a role for international promo- ticular outcomes, such as wages. They are comparable to tion of core labor standards to overcome potential collec- the rules that protect property rights and freedom of trans- tive action problems among poor countries competing actions in product markets, which most economists view as against one another in similar markets. The incorporation necessary if market economies are to operate efficiently. in trade agreements of measures against trade-related vio- The core labor standards are also fundamental elements of lations of labor standards could help build support for well-functioning democracies. And, just as the universality such agreements in rich countries. Poor countries could of property rights and freedom of market transactions also point to such provisions to reassure consumers (and does not imply identical laws or institutions among coun- international buyers serving them) that imported goods tries, universality of these core labor standards does not are produced under decent conditions (box 20.1). imply uniformity in the details of protection or in the insti- tutions that implement it. The Practice of Labor Rights in PTAs Even if improved implementation of core labor stan- dards in developing countries can contribute to economic, The formal treatment of labor standards in PTAs has political, and social development, there is still a question as evolved significantly over the past two decades, both within to the appropriate scope and treatment of those standards countries and among them, but implementation generally in trade agreements. Extensive research into these ques- remains weak. Language on labor standards in PTAs tions finds no systematic evidence that globalization is entered into by the United States—the leading proponent leading generally to a race to the bottom, or—the flip side of the linkage—has become more intrusive and nominally of this argument—that higher labor standards negatively more enforceable, but neither vigorous enforcement nor affect comparative advantage. There is also little question significant capacity building is consistently part of the that the worst conditions in most developing countries are process. A few other countries or regions are increasingly in the nontraded agricultural and services sectors. including labor standards in trade agreements, but legal The results of this research into the alleged race to the approaches vary greatly, and, again, little attention is bottom have been used to argue that there is no need to accorded to implementation or capacity building. incorporate minimum labor standards into trade agree- ments. But repression of those standards could give a Labor Rights Provisions in PTAs short-run competitive edge to particular firms or sectors, and there certainly is evidence that many firms and a few Unlike other parts of PTAs, labor provisions do not in governments perceive competitive disadvantages associ- themselves raise questions about the compatibility of the ated with higher labor standards. Thus, while there is no obligations with multilateral trade rules. The reason is that evidence of a race to the bottom driving down standards the WTO says virtually nothing about labor standards in rich countries, there could be a race to the bottom from (box 20.2). The General Agreements on Tariffs and Trade the bottom among low-wage countries—especially those (GATT), now incorporated into the WTO, contains only Box 20.1. Sweatshop Scandal Insurance for Brand-Name Buyers Better Work, a joint project of the International Labour Organization (ILO) and the International Finance Corporation (IFC), is designed to improve workers’ conditions and firms’ productivity, build government capacity to enforce labor standards, and reassure brand-name firms that sweatshop scandals in their supply chains are less likely to occur if factories participate in the program. The program is based on Better Factories Cambodia, a program that stemmed from an agreement in which the United States offered additional market access for Cambodian apparel exports under the Multifibre Arrangement (MFA) if factories were in “substantial compliance� with local labor laws and the ILO core standards and if ILO monitors were put in place to verify compliance. (See Elliott and Freeman 2003; Polaski 2006.) When the MFA was phased out at the beginning of 2005, Cambodia negotiated an extension of the ILO program as part of its strategy for attracting brand-name buyers and maintaining competitiveness in the post-quota situation. A Better Work project, the successor to Better Factories, was recently launched in Jordan, and one may begin soon in Vietnam. Source: ILO website, http://www.ilo.org/wow/Articles/lang—en/WCMS_094381/index.htm. 430 Kimberly Ann Elliott Box 20.2. Labor Rights and the WTO The issue of trade and labor standards has been with the WTO since its birth. At the ministerial conference of the General Agreement on Tariffs and Trade (GATT), held in Marrakesh in April 1994 to sign the treaty that formed the World Trade Organization (WTO), nearly all ministers expressed a view on the subject. The chairman of the conference concluded that there was no consensus among member governments at the time and thus no basis for agreement on the issue. At the 1996 Singapore ministerial conference, the United States again pushed for the WTO to address the issue and suggested formation of a working group to study the linkages between trade and core labor standards. Most other members rejected that proposal, however, and defined the WTO’s role on this question, as follows: We renew our commitment to the observance of internationally recognized core labour standards. The . . . ILO is the competent body to set and deal with these standards, and we affirm our support for its work in promoting them. We believe that economic growth and development fostered by increased trade and further trade liberalization contribute to the promotion of these standards. We reject the use of labour standards for protectionist purposes, and agree that the comparative advantage of countries, particularly low-wage developing countries, must in no way be put into question. In this regard, we note that the WTO and ILO Secretariats will continue their existing collaboration. At the time, there was no collaboration between the institutions, but the secretariats of the WTO and the ILO subsequently began to work together on technical issues under the banner of “coherence� in global economic policy making. No work on this subject is currently being undertaken in the WTO’s councils and committees, and labor standards are not subject to WTO rules and disciplines. Source: WTO. one article on the subject—Article XX(e), which permits enforcement measures from monetary fines to include the import bans for products made with prison labor. Still, a possibility of trade sanctions. Since then, only the agreement violation of international rules could arise if trade sanc- with Peru has been implemented, but, after renegotiation tions were used to enforce labor provisions under a PTA. If of some issues related to auto trade, the PTA with the Repub- a sanctions-related tariff increase should lead to an applied lic of Korea is expected to be approved by Congress by the tariff rate higher than what is bound in a country’s WTO summer of 2011. PTAs with Colombia and Panama remain schedule, that would be a violation which could be chal- in limbo, in part because of concerns by some in Congress lenged at the WTO. that the labor provisions are still not strong enough. Labor standards in PTAs also raise no issues of discrimi- Although respect for workers’ rights has been a condi- nation against third parties because the focus is on ensuring tion of U.S. unilateral trade preference programs since that a country’s labor laws are consistent with international 1984, it became an issue in reciprocal trade negotiations standards and are adequately enforced. There could even be only with the decision to pursue a deep integration agree- positive spillovers for other, nonmember countries in a ment with Mexico in the early 1990s (see Mayer 1998 for a region if conditions for their own migrant workers in the thorough discussion). Since then, U.S. negotiators have PTA member country improve as a result of better compli- routinely failed to persuade developing countries to discuss ance with labor standards under the PTA. There is some labor standards in regional or multilateral negotiations. In evidence of this happening in Jordan under its PTA with bilateral bargaining, however, they have insisted that trade the United States. agreements include labor standards. The inclusion of labor provisions in U.S. PTAs is pri- marily the result of the need to gain congressional Evolution of Worker Rights Provisions in U.S. PTAs approval for these agreements.5 Trade agreements with Starting with NAFTA in the mid-1990s, and continuing small, distant countries with relatively high standards, until the conclusion in May 2007 of a bipartisan agreement good working conditions, and few sensitive exports, such intended to resolve intra-U.S. political differences, U.S. as Australia, Chile, and Singapore, or agreements driven PTAs focused on requiring that countries effectively by foreign policy (e.g., Jordan), have received relatively enforce their labor laws, while “striving to ensure� that strong congressional support. But gaining congressional those laws were consistent with international standards. approval of agreements with countries where labor stan- The agreement reached between President George W. Bush dards are perceived to be inadequate has proved more and the leaders of both parties in the U.S. Congress broad- difficult, especially if the countries also export import- ened the scope of labor standards enforceable in U.S. PTAs, sensitive products such as apparel, sugar, or certain other put more emphasis on international norms, and expanded agricultural products. Labor Rights 431 Over the years, the approach to worker rights in U.S. flexibility. Consequently, the content of the labor chapters PTAs has fluctuated with the distribution of political (indeed, of most chapters) of the PTAs is remarkably simi- power in Washington, and the debate has become increas- lar, regardless of the level of development in the partner ingly partisan and rancorous. Four distinct periods in country. The labor chapter in the U.S. PTA with Australia, the treatment of labor rights in U.S. PTAs can be identi- for example, is virtually identical to those in all the other fied: the period leading to the NAFTA agreement, with PTAs negotiated between 2002 and 2006, except that it its side agreements on labor (and the environment); the does not contain an annex on cooperation and technical 2000 “Jordan standard� agreement; the PTAs negotiated assistance. And, despite concerns over violations of worker and ratified during 2002–06; and the period following the rights in Colombia, U.S. negotiators were constrained conclusion of the May 2007 Bipartisan Agreement. when Colombian negotiators suggested strengthening the Despite differences in the specific standards addressed labor chapter by adding the ILO core standard on nondis- and in dispute resolution mechanisms (see also Porges, crimination. U.S. negotiators at the time concluded that ch. 22 in this volume, for a discussion), all the agreements in adding this standard to the labor chapter would result in the first three periods share a common approach that the loss of more votes from labor standards opponents in requires the parties to enforce their own labor laws, without the Congress than there would be gains from supporters. a clear, legally binding commitment to ensure that those Overall, to the extent that there is variation in the template laws are consistent with internationally agreed labor stan- used, as is discussed next, it has been largely driven by dards. The “enforce-your-own-laws� standard reflects, in changing political balances in the United States, not condi- part, the fact that the U.S. government has ratified only two tions in or preferences of the partner-country government. of the eight fundamental ILO conventions—those on forced NAFTA side agreement on labor. NAFTA, negotiated by labor and on the worst forms of child labor. Some American Canada, Mexico, and the United States in the early 1990s, constituencies are concerned that a legally binding reference was the first U.S. PTA with a low-wage developing country. to ILO standards in a U.S. trade agreement might be used to The agreement was signed by President George H. W. Bush challenge the United States’ own labor laws and practices. but had not been ratified by Congress by the time he left A key implication of the “enforce-your-own-laws� office. In an attempt to assuage the concerns of workers approach for potential U.S. partner countries has become and unions about the agreement, President Bill Clinton that their labor laws may not diverge very far from interna- opted to negotiate a side deal on labor (and another on the tional standards; otherwise, this approach in the PTAs environment) before submitting the agreement to the could have the perverse effect of encouraging enforcement Congress for approval (Mayer 1998). of laws that undermine worker rights. Thus, during negoti- The resulting North American Agreement on Labor ations, or even before they are formally launched, U.S. Cooperation (NAALC) is 44 pages long and is unique in negotiators often identify changes to a partner country’s several respects. First, it was negotiated before the ILO had labor laws that need to be made before the agreement can agreed on the Declaration on Fundamental Principles and be concluded. Countries such as Bahrain, Chile, Guatemala, Rights at Work, which created a consensus on core labor and Morocco undertook major labor law reforms, at least in standards. The NAFTA negotiators developed their own part in anticipation of negotiations on PTAs with the list of labor standards, based on the worker rights condi- United States. Oman had a labor law like others in Middle tions in the U.S. GSP program. Second, the NAALC has the East that prohibited or tightly restricted labor organizing; it most elaborate institutional arrangements of any PTA, had to make major changes in its law as a condition for including a Labor Cooperation Commission, a governing completing PTA negotiations. Labor Council, and a secretariat to manage daily opera- In sum, having labor laws that are largely in compliance tions. There is also a mechanism for ministerial consulta- with the ILO fundamental principles is now a de facto con- tions, to deal with accusations that one of the parties has dition for negotiating a PTA with the United States. not adequately enforced its labor laws. Moreover, if recent pressure on Panama and Peru to make The full range of dispute settlement mechanisms is extensive changes to their labor laws is an indication of available in cases relating to alleged violations of technical things to come, the emerging condition for concluding a labor regulations on minimum wages or occupational PTA with the United States may be full compliance with health and safety, or involving child labor. In these cases, if the technical details of core ILO conventions—despite the neither consultations nor an expert evaluation resolves United States’ own gaps in compliance.6 the problem, the parties can appoint an arbitral panel, The careful political balancing act required for con- and the panel may ultimately impose a monetary fine if gressional approval sharply constrains U.S. negotiators’ the situation is not rectified.7 Allegations of forced labor 432 Kimberly Ann Elliott and discrimination are subject to evaluation by a panel of from enforcement, as it is in later agreements. But “striving independent experts, but no monetary or other penalties to ensure� something is a vague standard and would be dif- may be imposed if the issues are not resolved. Finally, com- ficult to enforce in practice. plaints involving violations of union rights are limited to U.S. PTAs negotiated between 2002 and 2006. Although ministerial consultations. In addition, disputes are referred there are minor differences from one text to the next, the for consultation or further dispute settlement only if there PTA labor chapters negotiated during the period 2002–06 is a “persistent pattern� of failure to enforce relevant labor retain the core approach from the Jordan agreement and laws and if the violations are in trade-related sectors. differ mainly in their enforcement provisions. In addition, The “Jordan standard.� The U.S.–Jordan PTA, negotiated they add more detailed provisions for cooperation on labor in late 2000, established a new precedent by including a (far matters, and they pay greater attention to transparency and less elaborate) section on labor in the main text of the due process. Several of them add articles encouraging agreement and by making it subject to the same dispute efforts within countries to promote public awareness of settlement procedures and remedies as the rest of the labor laws and enforce procedural fairness in their imple- agreement. Thus, if consultations, a dispute settlement mentation. None of these provisions, however, are subject panel, and the joint committee created to implement the to dispute settlement. agreement as a whole do not resolve a dispute, the com- In these agreements, the parties’ reaffirm their commit- plaining party is authorized “to take any appropriate and ment to the ILO Declaration. As in other agreements, they commensurate measure� (emphasis added). also make a commitment to “strive to ensure� that their laws The U.S.–Jordan labor text is also the first to reference are consistent with internationally recognized labor rights the 1998 ILO declaration defining core labor standards— and not to lower or weaken labor laws to encourage trade or although achieving compliance with the ILO standards is investment involving the other party. But these provisions stated as an aspiration rather than an obligation. Whereas are hortatory and are explicitly excluded from dispute set- the ILO core labor rights include nondiscrimination, as tlement. Each of the seven PTAs negotiated during the well as freedom of association and prohibitions on forced period—those with Australia, Bahrain, Chile, Morocco, labor and child labor, the agreement with Jordan, like those Oman, Singapore, and the five countries of Central America that followed up to 2007, continues to use the previous U.S. plus the Dominican Republic (CAFTA–DR)—includes an definition of “internationally recognized labor rights,� article providing that the “enforce-your-own-laws� provi- which excludes nondiscrimination and includes “acceptable sion is the only provision subject to dispute settlement. conditions of work� with respect to minimum wages, hours These agreements did retain the Jordan practices of of work, and occupational safety and health. incorporating the labor chapter into the main text and In addition to calling on parties to “strive to ensure� that using the same general dispute settlement process for labor domestic laws are consistent with “internationally recog- as for other parts of the agreement. But, as in the NAFTA nized labor rights,� as the United States defined them, the side agreement, the agreements limit the potential penalty agreement also discourages the parties from “waiv[ing] or for violations of the labor chapter to imposition of a “mon- otherwise derogat[ing] from . . . [labor] laws as an encour- etary assessment,� and this “fine� can be no higher than agement for trade.�As in NAFTA, the only binding obliga- US$15 million (to be adjusted for inflation), regardless of tion is that a party “shall not fail to effectively enforce its the income level or size of the partner. If the party charged laws� on a sustained basis in a way that affects trade. Other with the violation does not pay the assessment as ordered paragraphs preserve the discretion of governments to by an arbitral panel, the complaining party can collect it by adopt, modify, and enforce labor laws and regulations. suspending trade benefits as a last resort, as in NAFTA. Vio- Thus, according to Article 6, Section 4(b), of the agreement, lations of provisions in other chapters of these agreements a party will be in compliance with its labor obligations can lead to trade sanctions that are generally commensurate under the agreement if “a course of action or inaction [in with the injury suffered. enforcing labor laws] reflects a reasonable exercise of such These agreements, although they contain more detailed discretion, or results from a bona fide decision regarding the provisions on institutional mechanisms and cooperation allocation of resources.� than does the Jordan PTA, do not have as elaborate or com- The standard for determining that a violation has plex a structure as NAFTA (and few have any funding occurred is thus set rather high. The language calling on attached). Indeed, dissatisfaction with the perceived lack of parties to “strive to ensure� that labor laws are consistent enforcement under the NAFTA labor side agreement led with ILO core principles and that those laws are not weak- then-candidate for president Barack Obama to call for ened for competitive reasons is not explicitly excluded reopening NAFTA, perhaps by incorporating the side Labor Rights 433 agreements into the main text. During the presidential pri- dispute settlement procedures—and potential penalties— mary campaign, Obama even threatened to “use the ham- as other PTA chapters. mer of a potential opt-out as leverage to ensure that we Whether President Obama will follow or amend the actually get labor and environmental standards that are May 2007 template for labor chapters in PTAs, or what enforced.�8 As is discussed below, however, the enforce- the template means for negotiations at the WTO, remains ment provisions in agreements in which labor provisions to be seen. As of early 2011, none of the three pending are contained in the main text are untested, and so there is PTAs had been submitted to Congress for approval, no basis on which to conclude that one model is better although submission of the Korean agreement appeared than the other. Ultimately, the impact of these agreements imminent at the time of writing. What is clear is that the depends on the desire of the parties to see them effectively issue will not disappear, as President Obama’s first trade enforced, not the details of the formal procedures. policy agenda report in 2009 emphasized the role of Bipartisan agreement on labor rights in PTAs. U.S. PTA worker rights in making “support for global markets sus- negotiations with Colombia, Korea, and Panama, which tainable,� and the policy’s second priority, after support for were at various stages, came to a standstill after the 2006 a rules-based system, is to “advance social accountability� midterm elections. Many in the newly elected Demo- in U.S. trade policy (USTR 2009). cratic majority in Congress opposed all three negotia- tions for various reasons, including, in the case of Worker Rights Provisions in Non-U.S. PTAs Colombia, concern over violations of the human rights of union organizers and members. More generally, the Recent studies examining the array of provisions in various new congressional majority regarded the “enforce-your- trade agreements found few non-U.S. PTAs that include own-laws� approach as too weak and insisted on changes language on labor rights.9 Canada is the exception, as it to the PTA template, as well as in the three pending typically includes side agreements on labor in its PTAs with agreements (Elliott 2007). developing countries. Chile often, but not systematically, In May 2007, President Bush and the leaders of both incorporates references to labor issues either in its PTAs or parties in Congress reached an agreement on how to in associated memoranda of understanding. Traditionally, address labor issues, as well as environmental, intellectual the European Union had some language on human rights property, and other matters, in U.S. PTAs. The compromise in its PTAs, but that may be changing to include specific on labor standards requires parties “to adopt and maintain references to worker rights, as well. The recently concluded in its statutes and regulations, and practices thereunder,� economic partnership agreement (EPA) between the EU the labor rights, “as stated� in the 1998 ILO Declaration. and the Caribbean region contains provisions on worker A footnote specifies that these obligations refer “only to the rights, but it remains to be seen whether this will become a ILO Declaration,� since some in the U.S. Congress did not consistent template for these agreements in other regions. want the United States to refer in reciprocally binding trade An additional review by the author of PTAs involving these agreements to ILO conventions it had not ratified. The countries, as well as Australia, New Zealand, Japan, and agreement also revises the definition of the standards cov- China, found very few additional examples of agreements ered to include nondiscrimination, as in the ILO Declara- that contain language promoting improvements in worker tion, but it retains the “acceptable conditions of work� rights.10 standard from U.S. practice. Canada. After the United States, Canada most consis- The May 2007 compromise also makes the nonderoga- tently includes legally binding labor provisions in its trade tion article, which prohibits parties from lowering labor agreements. One difference from U.S. practice is that standards for competitive advantage, legally binding and Canada only insists on labor provisions in PTAs with enforceable, and it appears to reduce countries’ flexibility developing countries, whereas they are part of the U.S. in choosing how to allocate resources to enforcement and template regardless of the level of development of the part- prosecution of labor law violations. Although the new ner country. Thus, U.S. agreements with Australia and template states that decisions regarding the “distribution Singapore include labor chapters, but Canada’s agreements of enforcement resources shall not be a reason for not with Israel and the European Free Trade Association complying with the provisions of this Chapter,� it retains (EFTA) contain no specific provisions on labor issues. As of the language about parties having “the right to the reason- mid-2010, Canada had implemented three additional PTAs able exercise of discretion and to bona fide decisions with with developing countries (with Chile, Costa Rica, and regard to the allocation of resources.� Finally, the new Peru), in addition to NAFTA, and it had signed three more, framework makes the labor chapter subject to the same with Colombia, Jordan, and Panama.11 These agreements 434 Kimberly Ann Elliott appear to be modeled broadly on the NAFTA precedent, meet its obligations under the [labor cooperation agree- incorporating provisions on labor and environmental ment].�13 The website for the International Trade and cooperation in side agreements, along with separate con- Labour Grants and Contributions Program, which “is sultation and dispute resolution mechanisms that do not designed to help the Government of Canada meet its com- include trade sanctions. mitment to address the labor dimensions of international Canada’s PTA with Chile was explicitly and closely mod- trade and economic integration,� has not been updated for eled on the NAFTA side agreement in order to facilitate more than a year, and a notice at the top of the page Chile’s accession to that agreement at a later date. (Instead, declares that the program is not currently accepting appli- the United States subsequently negotiated a separate agree- cations for grants.14 ment with Chile, as described above.) Subsequent Canadian In sum, Canada has followed a path regarding labor PTAs retained the model employing a labor side agreement standards in trade agreements with developing countries because, under Canada’s federal system, national policy that is adapted to its constitutional structure and political makers cannot negotiate international treaties that limit needs, but with language on labor standards that is similar provincial jurisdiction over labor law without the provinces’ to that in U.S. PTAs. The key differences are not so much in explicit acquiescence. the legal obligations that each country seeks to promote as Canada’s agreement with Chile, like the later ones with in the enforcement measures and the relative role of coop- Colombia and Peru, follows the NAFTA precedent in pro- eration in promoting worker rights. The similarities could viding for fines for noncompliance with the labor side facilitate coordination of approaches toward linking trade agreement. The agreement with Costa Rica, by contrast, and labor rights, if the two countries so chose. provides no penalties for violations, other than “the mod- Chile. Chile is unusual among developing countries in ification of cooperative activities.� All the agreements that that it frequently seeks to include labor provisions in its followed the one with Chile incorporate references to the agreements with other developing countries—although ILO core standards, as defined in the 1998 Declaration, these provisions are often general and aspirational in and the agreements with Peru and Colombia hew closely nature, rather than legally binding. For example, Chile’s to the language used in the U.S. agreements with those PTA with China includes a memorandum of understand- countries. These agreements also restrict reviews of com- ing calling for both parties to “enhance their communica- pliance to the ILO core standards only, not implementa- tion and cooperation on labor, social security and the envi- tion of other labor laws. Canadian agreements do retain ronment.� The Trans-Pacific Strategic Economic language calling for attention to improvement of other Partnership, which includes (in addition to Chile) Brunei working conditions, similar to the U.S. definition of Darussalam, New Zealand, and Singapore, also includes a “acceptable conditions of work.� memorandum of understanding that affirms the parties’ Perhaps because of the nature of the parties to the two commitment to the ILO core standards and calls on them agreements that have been implemented long enough to be to promote better understanding of one another’s labor analyzed—those with Chile and Costa Rica, both of them systems and practices, provide a forum for discussion of democratic countries with relatively good labor practices— labor issues, and, if questions over the interpretation of the there has been little activity under Canada’s labor coopera- memorandum of understanding arise, to consult and per- tion agreements. There have been no public communications haps discuss them in joint meetings. The agreement under either agreement, and the website where public com- between Chile and the EU includes an article on social munications and information about cooperative activi- cooperation stating that the parties “shall give priority to ties under particular agreements is posted has not been domestic measures aimed at developing and modernising updated since 2005.12 Several technical workshops, public labour relations, working conditions, social welfare and conferences, and seminars were held in the early years of employment security.� But there are no further details or implementation of the Chile PTA, but none has occurred stipulations on how to promote or monitor these activities in several years. The only cooperative activity specified (Bourgeois, Dawar, and Evenett 2007, 27). under the agreement with Costa Rica is a technical assis- The most far-reaching agreement is Chile’s PTA with tance program to be carried out by the ILO to strengthen Colombia, signed in 2006, which is the only PTA found labor administration, to which Canada contributed that involves only developing countries and includes a sep- US$1 million. As for the agreement with Colombia, the arate chapter in the main text addressing labor standards. Canadian Trade Ministry website says only that Canada The language is similar to that contained in PTAs involving “has offered to provide Colombia with labour-related the United States, with which Colombia was negotiating technical assistance which will allow Colombia to better at the time, but it contains no enforcement provisions Labor Rights 435 beyond consultation and cooperation, and both labor and agreement, which contains a provision stating that domes- the environment are excluded from the dispute settlement tic labor laws should not be weakened as a means of chapter that applies to the rest of the agreement. attracting investment. Article 103 is similar to the language European Union. The EU has not traditionally included on labor rights in U.S. PTAs in affirming the parties’ recog- labor rights in its PTAs, but it may be moving in that direc- nition that it is “inappropriate� to weaken labor laws in tion, at least in the economic partnership agreements that order to encourage investment and that each “shall strive to it is negotiating with former African, Caribbean, and ensure� that this does not happen “in a manner that weak- Pacific (ACP) beneficiaries of unilateral preferences under ens or reduces adherence to the internationally recognized the Cotonou Agreement. Horn, Mavroidis, and Sapir labor rights.� Not only is this language strikingly similar to (2010) examined 14 EU trade agreements going back that included in the labor chapters of U.S. PTAs, but the list decades and found that only the recently concluded EPA of “internationally recognized rights� included in the with the Caribbean Forum of African, Caribbean, and agreement also mimics the list contained in U.S. agree- Pacific States (CARIFORUM) included an article on labor ments rather than the core labor standards reflected in the rights. Prior to that, EU agreements focused on broad ILO Declaration of 1998. Under the Japan–Philippines human rights and did not single out worker rights for agreement, if neither consultation nor arbitration resolves special attention. a dispute arising under Article 103, the complaining party The relevant section in the CARIFORUM agreement can temporarily suspend its obligations under the agree- reaffirms the parties’ commitment to internationally rec- ment, preferably in the same sector—that is, foreign direct ognized labor standards, as defined in the ILO Declaration, investment—affected by the violation. None of Japan’s but it also recognizes the right of the parties to “establish other PTAs that have been notified to the WTO includes their own social regulations and labor standards in line such language, including subsequent agreements with with their own social development priorities.� The parties Brunei Darussalam and Indonesia. then agree that they should not “encourage trade or foreign To sum up, developing countries negotiating PTAs direct investment to enhance or maintain a competitive with the United States, Canada, Chile, and, perhaps, the advantage� by weakening labor laws. Complaints related to EU can expect to confront demands for inclusion of labor this chapter are to be investigated under the normal dis- standards in some form. At this stage, only U.S. negotia- pute settlement procedures of the agreement, but “com- tors insist on trade sanctions as a potential enforcement pensation or trade remedies [may not] be invoked against a measure—a key concern for developing countries. But Party’s wishes� (EU 2008). In other words, although the whatever the model followed, and whether the emphasis provision is nominally enforceable, it carries no penalties is on sanctions or on cooperation and capacity building, for violations, other than scrutiny. As this agreement is implementation of labor provisions does not seem to be a quite recent, it is not yet known how it will be implemented high priority in these PTAs. or whether technical assistance will be provided to improve labor standards. Implementation and Enforcement of Whether the labor standards language used in the Labor Rights Provisions in PTAs CARIFORUM agreement will be repeated in other EPAs is unclear; no other agreements have been finalized. To the extent that worker rights have been incorporated Interim agreements have been initialed with a number of in PTAs, most countries have used a “soft-law� approach. countries, but they are mostly limited to addressing mar- The language is usually hortatory or, if it is nominally ket access for goods. For example, the interim agreement binding, many agreements explicitly exclude the labor with Cameroon, which is intended to lead eventually to a provisions from sanctions for noncompliance. The U.S. full EPA with Central Africa, says only that the parties approach, however, has been quite different. Although “agree to ensure that sustainability considerations are specific commitments and procedures differ from agree- reflected in all titles of the EPA and to draft specific chap- ment to agreement, every U.S.-negotiated PTA since ters covering environmental and social issues.� Thus, large NAFTA has incorporated into the main text legally bind- questions loom as to whether the EU will now place ing provisions protecting worker rights that are subject to greater attention on worker rights in its trade agreements dispute settlement and that include some form of sanc- and, if so, how much effort it will put into implementing tion for noncompliance. But despite the striking differ- those provisions. ences in attention to enforcement issues, implementation Japan. Finally, the only other obligation to protect labor across countries has not been as different as the legal tem- rights in a PTA seen thus far is in the Japan–Philippines plates imply. 436 Kimberly Ann Elliott U.S. negotiators are constrained in how they treat labor United States. That program, however, was developed in issues in PTAs by the need to obtain congressional response to a report on exploitative conditions involving approval. The role of dispute settlement procedures and, migrant workers, not as part of a strategy for implementing especially, trade sanctions in enforcing labor provisions has the PTA or as a result of systematic monitoring. A key been a key source of disagreement between the United problem is that the congressional committees that oversee States and its trading partners, but the partner countries trade agreements are not the same as those that appropri- have little influence over labor language because the ate funds for capacity building, and close collaboration is approach is settled in negotiations among various U.S.- often lacking. based constituencies, the two major political parties, and Although the details differ, the dispute settlement process the executive and legislative branches. Although these for labor violations under U.S. PTAs is generally as follows. internal political battles have brought about changes in the The agreements do not provide for a private right of action, language of the labor provisions, depending on the distri- but they do require parties to designate a national contact bution of political power among the various interests, point to accept submissions from citizens or groups request- enforcement of the labor provisions in U.S. PTAs has var- ing consultations over potential problems under the agree- ied surprisingly little. ment. It is then up to the receiving government to decide Between 1994 and 2008, two U.S. presidents, each from whether to request consultations with the other government. a different political party, presided over the implementa- If consultations do not resolve the problem, the government tion of PTAs. Both were restrained in their approaches to alleging a violation can request appointment of a dispute enforcement, with neither coming close to invoking sanc- settlement panel to investigate. If the panel agrees that a vio- tions of any sort. The big unknown today is how aggressive lation exists, the responding country will have a certain President Obama will be on this issue, given the statements amount of time to bring its practices into compliance and if in his official trade agenda on the crucial role that worker it does not, penalties—either fines or trade sanctions, rights should play in trade policy (USTR 2009), and the depending on when the agreement was negotiated—may be decision on July 30, 2010, to formally request consultations authorized (see table 20.1). with Guatemala over alleged violations of the CAFTA–DR Thus far, only three of the nine U.S. PTAs in force have labor chapter.15 seen any enforcement activity, and most of that has been On the other side, technical and financial assistance to under NAFTA. There have been only two other requests for partner governments to strengthen their capacity to imple- consultations regarding labor rights violations, involving ment and enforce labor standards has also been neglected. Jordan in 2006 and Guatemala in 2008. Other than the agreement between the United States and In the Guatemala case, first filed in 2008, the U.S. CAFTA–DR, there has been no dedicated program of Department of Labor (U.S. DOL 2009) investigated the capacity-building assistance related to the signing and issues raised in the submission from the AFL-CIO trade implementation of PTAs. The U.S. Agency for Interna- union, affirmed that problems existed, and recommended tional Development (USAID) is providing technical and actions the government of Guatemala could take to financial assistance to address labor rights violations in address them. In early 2009, the Department of Labor special export zones in Jordan that export clothing to the opted not to recommend formal consultations under the Table 20.1. Sanctions Authorized for Labor Violations in U.S. Preferential Trade Agreements Agreement partners, and year approved Enforcement mechanisms permitted Canada, Mexico, 1993 In Canada, fines only; for Mexico and the United States, monetary assessments that may be collected by suspending tariff concessions if not paid Jordan, 2002 “Any appropriate and commensurate measure�—the same as in other parts of the agreement Chile, 2003; Singapore, 2003; Australia, 2004; Morocco, 2004; Monetary assessments, capped at US$15 million (adjusted for Bahrain, 2005; Oman, 2005; CAFTA–DR, 2005; Korea, Rep., inflation) pending Peru, 2007; Panama, pending; Colombia, pending Compensation, fines, or trade sanctions—the same as for other parts of the agreement Source: Office of the U.S. Trade Representative, http://www.ustr.gov. Note: CAFTA–DR, Dominican Republic–Central America Free Trade Agreement. Labor Rights 437 PTA but promised to reassess the situation within six labor-market conditions.� Another view, however, is that months and “determine whether further action is war- the NAALC and activity under it has been helpful in ranted. . . .�16 The review was extended again when that social mobilization, as “part of larger strategies involving deadline passed with little movement. Finally, on July 30, public education on the issues, media, legislative lobby- 2010, the U.S. trade representative and the secretary of ing and mobilization� across borders (Buchanan and labor jointly sent a letter requesting formal consultations Chapparo 2008, 33). For example, although problems under the trade agreement to their counterparts in remain, pressures on Mexico to improve respect for free- Guatemala, “the first labor case the United States has ever dom of association have led to important changes brought against a trade agreement partner.�17 If those con- requiring that secret ballots be used in union-organizing sultations do not resolve U.S. concerns, a dispute settle- elections (ibid.). ment panel that might lead to sanctions could be The submissions made in the United States against appointed in coming months. Mexico have generally been of two types. One common A recent U.S. General Accountability Office investiga- pattern involves complaints by U.S. unions against U.S. tion of the implementation of four U.S. PTAs, with Chile, multinational companies with operations in Mexico, alleg- Jordan, Morocco, and Singapore, concluded that ing violations of the right to organize. The other common type involves complaints by human rights groups and [free trade agreement] negotiations spurred some labor reforms in each of the selected partners, according to U.S. worker rights activists against Mexican federal or state gov- and partner officials, but progress has been uneven and U.S. ernments, alleging a failure to adequately enforce labor engagement minimal. An example cited was Morocco’s laws, or to provide fair and impartial procedures for the enactment of a long-stalled overhaul of its labor code. consideration of labor disputes under Mexican law. In However, partners reported that enforcement of labor laws addition to problems with union rights, these complaints continues to be a challenge, and some significant labor also often involve occupational safety and health issues and abuses have emerged. In the [free trade agreements] we gender discrimination. examined, [the U.S. Department of] Labor provided mini- It is difficult to separate out the impact of the NAFTA mal oversight and did not use information it had on part- dispute settlement process on any improvements that ner weaknesses to establish remedial plans or work with occurred in these cases because so many other factors that partners on improvement. (U.S. GAO 2009, ii) could have had an effect were operating at the same time. An escalation in the Guatemala case could send impor- Perhaps most important, the tradition of single-party con- tant signals about whether there will be a change in the trol in Mexican politics ended in this period, opening space aggressiveness with which labor standards are enforced for independent unions to more effectively challenge the and effective implementation pursued in U.S. PTAs. traditional corporatist union structure in Mexico. Interest- ingly, some of the most tangible benefits of implementa- tion of the NAALC accrued not to Mexican workers in NAFTA and “Enforcement� of Labor Standards Mexico, but to migrant Mexican workers in the United As of mid-2010, 37 submissions had been made under the States. Several of these cases appear to have resulted in at NAALC, of which 6 were filed in Canada, 9 in Mexico (all least temporary improvements in conditions for the work- against the United States and mostly involving complaints ers directly involved. More broadly, they highlighted the about mistreatment of migrant Mexican workers), and 22 fact that U.S. federal laws protect the rights of workers, in the United States (all but two against Mexico).18 regardless of their legal status in the United States. One Roughly two-thirds of the submissions alleged violations case in particular, described in box 20.3, contributed to of freedom of association or of the right to organize and changing the practices of U.S. agencies with respect to pro- bargain collectively. According to a detailed analysis by tecting migrant rights in the workplace. Hufbauer and Schott (2005), only 14 cases resulted in reports, and 13 of these recommended ministerial consul- U.S.–Jordan PTA: Capacity Building as Enforcement tations. Most of the consultations led to public outreach seminars, studies or consultants’ reports, or government-to- In contrast to NAFTA, protection of labor rights in the government discussions. Most others were rejected for U.S.–Jordan agreement was included in the main text review (11 cases), and a handful was withdrawn or settled of the agreement, and enforcement was the same as for before the review was completed.19 all other parts of the agreement. But in order to gain Hufbauer and Schott (2005, 128) conclude that “the sufficient political support among labor standards skeptics NAALC has had practically no impact on North American in the U.S. Congress, the United States and Jordanian 438 Kimberly Ann Elliott Box 20.3. NAFTA as a Tool for Promoting Rights of Mexican Migrants in the United States In 1998 the Yale Law School Workers Rights Project, along with roughly 20 other nongovernmental organizations concerned with immigration and worker rights, filed a submission under the labor agreement with Mexico alleging that a 1992 memorandum of understanding between the U.S. Department of Labor and the U.S. Immigration and Naturalization Service (INS) had led to a failure to enforce U.S. labor laws with respect to migrant workers. Under the memorandum of understanding, the Labor Department was expected to inspect employment eligibility verification forms when investigating complaints regarding violations of wage laws and to report any discrepancies that might indicate the presence of illegal immigrants to the INS. The petition alleged that this practice had a chilling effect on migrant workers, who would fear being deported if they reported underpayment of wages in violation of U.S. labor laws. Shortly after the submission was accepted for review, the U.S. agencies concerned signed a new memorandum of understanding stating that the Labor Department would no longer inspect employment eligibility forms when investigating complaints under its jurisdiction. This revision, in effect, restored the migrant workers’ rights under U.S. federal law to be protected from illegal behavior on the part of employers, regardless of the workers’ legal status in the country. After additional consultations between the U.S. and Mexican ministries of labor, the parties agreed that the U.S. Labor Department would produce and disseminate materials in Spanish and English explaining the rights of migrant workers. Source: U.S. Department of Labor website, http://www.dol.gov/ilab/programs/nao/status.htm#iib5; National Immigration Law Center, “Law Students File Petition under North American Free Trade Agreement (NAFTA) Challenging Collaboration between Department of Labor and INS,� Immigrants’ Rights Update 21 (8), December 21, 1998. authorities exchanged letters stating that they did not of Representatives by only a slim margin, mainly because anticipate using the dispute resolution mechanism for of concerns about the adequacy of labor standards in labor issues.20 the region. In order to build support for the agreement in Then, in September 2006, the AFL-CIO filed a com- the United States, a package of technical and financial plaint with the Office of the U.S. Trade Representative assistance to support capacity building for improved claiming that Jordan’s labor law was inadequate in several implementation and enforcement of labor standards in areas regarding freedom of association and that Jordan the region was negotiated. The initial appropriation for was not adequately enforcing its labor laws. The AFL-CIO the program was US$57 million for fiscal years 2005–07; asked that the administration formally request consulta- an additional US$15 million was appropriated for fiscal tions on the matter, as called for under the PTA’s dispute 2008 (U.S. DOL 2009). settlement provisions, but the administration declined to In April 2005 the labor ministries from the six partner do so. Instead, the U.S. trade representative turned to the countries, with support from the Inter-American Develop- joint committee created to implement the trade agree- ment Bank, issued a white paper, “Building on Progress: ment to “address these issues as a matter of priority,� and Strengthening Compliance and Enhancing Capacity.� The the committee, in turn, created a Working Group on ILO, as part of a verification project funded by USAID, Labor Affairs. The formal dispute settlement procedures issued a baseline report for assessing implementation of under the PTA were never invoked, but the two govern- the white paper recommendations in 2007. In January ments collaborated with the ILO and the International 2009 the International Affairs Bureau of the U.S. Depart- Finance Corporation to create a Better Work program to ment of Labor issued its first report to Congress on monitor labor standards in the clothing sector and to beef progress in implementing the capacity-building provisions up the capacity of Jordan’s Labor Ministry to enforce its under the labor chapter of the PTA, concluding that results laws (box 20.4). had been mixed (U.S. DOL 2009). It is difficult to assess the degree to which working conditions have really changed because the progress report mainly cites measures of CAFTA–DR: Capacity Building and inputs—budget increases, inspectors hired or trained, Coalition Building workshops held, and radio advertisements aired. It seems The agreement signed in August 2004 between the United likely that economic conditions in Central America— States and CAFTA–DR (the five Central American coun- increased competition from China and the rest of Asia after tries of Costa Rica, El Salvador, Guatemala, Honduras, expiration of the Multifibre Arrangement quotas on tex- and Nicaragua, plus the Dominican Republic) was the tiles and clothing, and a global economic recession—have most politically controversial PTA since NAFTA. When had greater, and negative, effects on jobs, wages, and work- finally approved by Congress in 2005, it passed the House ing conditions. Labor Rights 439 Box 20.4. Responding to a Sweatshop Scandal through Capacity Building and Monitoring In early 2006, the National Labor Committee (NLC), a New York-based nongovernmental organization that investigates labor abuses around the world, released a report alleging serious violations of worker rights in Jordanian garment factories employing mainly migrant workers from South Asia and China and exporting to the United States. The AFL-CIO, the main U.S. union federation, took up the case, filing a petition under the U.S. preferential trade agreement (PTA) with Jordan and pointing out a major gap in Jordanian labor laws, which require workers to be citizens in order to be eligible to join unions. The Jordanian government responded immediately and, in coordination with the U.S. Agency for International Development (USAID), ordered an independent investigation, which confirmed many of the NLC’s allegations. The government also took a number of immediate steps between May 2006 and the end of the year, which included raising the minimum wage, increasing the Ministry of Labor budget by 80 percent, beefing up inspections and closing some factories, creating a multilingual hotline for worker complaints, launching a review of its labor laws with the aim of bringing them closer to international standards, and reaching agreement with the ILO on developing a Decent Work country program. In February 2008, Jordan, the ILO, and the International Finance Corporation (IFC) launched the Better Work Jordan Project, to be jointly funded by the Jordanian government and USAID. In addition to independent monitoring and transparent reporting on factory conditions to encourage improvements in labor standards compliance, the project also has an explicit objective of improving “enterprise performance in global supply chains in developing countries.� The factory assessments will be entered in a database that can be made available to buyers or others as desired by participating factories. Public reports will be issued containing aggregated data on trends and the key issues uncovered, as well as documents naming individual factories and providing indicators of performance in key areas. It is hoped that the system will be credible enough that international buyers will forgo their own factory audits, as Wal-Mart and Sears/Kmart have agreed to do, thereby lowering costs for both factory and buyer. Source: ILO and IFC 2007; Jordan 2008; NLC 2006. Conclusions Fourth, whatever the details of the language on labor standards in trade agreements, there has been little atten- In recent years, there appears to have developed, at least tion to implementation in most cases, and therefore little among a few major rich countries, a degree of convergence impact on the developing countries that sign these agree- that worker rights should be addressed in trade agree- ments. Even under U.S. PTAs, which contain the strongest ments and that they should be based on the principles in language on paper, enforcement is rare, and sanctions have the 1998 ILO Declaration. Beyond that, four main conclu- never been applied. Unfortunately, financial and technical sions emerge from this review of experience with labor capacity-building assistance to improve labor standard provisions in PTAs. implementation is also rare. First, enforceable provisions for labor standards are a The experience analyzed here suggests that developing- condition for negotiating a PTA with the United States and country governments need not fear the inclusion of labor are likely to remain so. Moreover, recent experience sug- standards in PTAs, although the evolution of the U.S. gests that developing countries contemplating negotiation trade policy approach bears careful watching. Indeed, gov- of a PTA with the United States should be prepared to ernments seeking to strengthen their economies and undertake reforms to bring their laws into broad conform- societies by, among other things, spreading the benefits of ity with the eight ILO conventions associated with the four globalization and growth more broadly, may welcome core labor standards. rather than resist pressures concerning labor standards. Second, labor provisions in PTAs not involving the Moreover, if the issue remains important to the United United States or Canada are unusual and are almost always States, developing countries should bargain for something hortatory, rather than legally binding or enforceable. In the they want—perhaps assistance in creating or strengthen- Canadian case, the provisions are in side agreements, with ing safety nets or other adjustment programs to address a separate dispute settlement process and a focus on coop- the costs borne by workers or firms negatively affected by eration, rather than negative sanctions—although the lat- the PTA. ter are not always ruled out. Third, the EU also appears to be joining the trend, if the Notes model used in its economic partnership agreement with the Caribbean is replicated in other similar agreements. How- 1. For the text of the declaration, along with the follow-up reports, see http://www.ilo.org/declaration/(accessed June 17, 2010). ever, the language on labor, although nominally binding, 2. This section draws heavily on Elliott and Freeman (2003, ch. 1) and includes no sanctions for noncompliance. the references cited therein. 440 Kimberly Ann Elliott 3. The evidence and analysis are presented in detail in Elliott and www.ustr.gov/about-us/press-office/speeches/transcripts/2010/july/remarks- Freeman (2003); Elliott (2004) applies the analysis to the U.S. PTA with ambassador-ron-kirk-enforcement-alleghn. Central America. 18. See the website of the U.S. contact point, http://www.dol.gov/ 4. The declaration applies to all ILO members, whether or not ilab/programs/nao/naalc.htm. they have ratified the associated conventions. Thus, all members are 19. An updated list of submissions and their status may be found at expected to promote those broad principles, but those that have not the NAALC website, http://www.naalc.org/userfiles/file/NAALC-Public- ratified a convention, with the exception of Conventions 87 and 98, Communications-and-Results-1994-2008.pdf. dealing with freedom of association and the right to collective bargain- 20. Bolle (2001, 6); this report also provides a useful summary of the ing, cannot be challenged for failing to implement all of the detailed debate over labor standards in trade agreements as of the early 2000s. obligations contained in the conventions. Freedom of association is considered so fundamental that there is a special review and com- plaints mechanism for those conventions. The other conventions are References Conventions 29 and 105, concerning prohibition of forced labor; Conventions 138 and 182, calling for the eradication of harmful child Aidt, Toke, and Zafiris Tzannatos. 2002. Unions and Collective Bargaining: labor; and Conventions 100 and 111, covering nondiscrimination and Economic Effects in a Global Environment. Washington, DC: World equal remuneration in wages. Bank. 5. Hafner-Burton (2009) compellingly analyzes the political versus Bolle, Mary Jane. 2001. “Jordan–U.S. Free Trade Agreement: Labor Issues.� the moral and other motivations for including human rights provisions in Report RS20968, Congressional Research Service, Library of Con- trade agreements. gress, Washington, DC. 6. See, for example, the issues of Inside U.S. Trade for January 19, Bourgeois, Jacques, Kamala Dawar, and Simon J. Evenett. 2007. “A Com- 2009, on Peru, and April 10, 2009, on Panama. Information on all these parative Analysis of Selected Provisions in Free Trade Agreements.� trade agreements, including, in many cases, fact sheets on the labor provi- Prepared for the Directorate General for Trade, European Union, sions and on how countries have responded to demands for labor law Brussels. reform, can be found on the Office of the U.S. Trade Representative web- Buchanan, Ruth, and Rusby Chaparro. 2008. “International Institutions site, http://www.ustr.gov/trade-agreements/free-trade-agreements. See and Transnational Advocacy: The Case of the North American Agree- also Rogowsky and Chyn (2007). ment on Labor Cooperation.� UCLA Journal of International Law and 7. In the case of a bilateral dispute between the United States and Foreign Affairs 13(1): 129–60. Mexico, bilateral tariff concessions can be withdrawn to the extent neces- CITEE (Centre for International Trade, Economics and Environment). sary to collect the value of the fine, but this provision is not regarded as 2004. “Trade-Labour Debate: The State of Affairs.� Research Report, authorizing trade sanctions. Elliott (2001) analyzes the fine mechanism CUTS (Consumer Unity & Trust Society) International, Jaipur. and how it might work if implemented. Destler, I. M. 2005. American Trade Politics. 4th ed. Washington, DC: 8. “Clinton, Obama Pledge to Withdraw from NAFTA unless Renego- Peterson Institute for International Economics. tiated,� Inside U.S. Trade, February 29, 2008. Elliott, Kimberly Ann. 2001. “Fin(d)ing Our Way on Trade and Labor 9. Horn, Mavroidis, and Sapir (2010) examine 14 U.S. and 14 EU Standards?� International Economics Policy Brief PB01-5, Peterson PTAs. Bourgeois, Dawar, and Evenett (2007), in a study for the European Institute for International Economics, Washington, DC. http://www Commission, examine 27 PTAs involving Western Hemisphere, Asian, .piie.com/publications/pb/pb.cfm?ResearchID=73. and European Free Trade Association (EFTA) countries. ———. 2004. “Labor Standards, Development, and CAFTA.� International 10. The discussion that follows is based on an examination of Economics Policy Brief PB04-2, Peterson Institute for International agreement texts identified on national government websites; the Economics, Washington, DC. http://www.piie.com/publications/pb/ WTO database on regional trade agreements, http://rtais.wto.org/UI/ pb04-2.pdf. Also published as “Trading Up: Labor Standards, Develop- PublicMaintainRTAHome.aspx; and the Organization of American ment, and CAFTA,� CDG Brief 3 (2, May) (Washington, DC: Center for States database, http://www.sice.oas.org/agreements_e.asp. Global Development). 11. This discussion draws on the agreement texts and related infor- ———. 2007. “Appendix A. Treatment of Labor Issues in US Bilateral mation posted on the website of the Canadian Department of Foreign Trade Agreements.� In “American Trade Politics 2007: Building Affairs and International Trade, http://www.international.gc.ca/trade- Bipartisan Consensus,� ed. I. M. (Mac) Destler, Policy Brief PB07-5, agreements-accords-commerciaux/index.aspx?lang=eng. Peterson Institute for International Economics, Washington, DC. 12. See the links to the individual agreements at the Human Elliott, Kimberly Ann, and Richard B. Freeman. 2003. Can Labor Standards Resources and Skills Development Canada website, “International Labour Improve under Globalization? Washington, DC: Peterson Institute for Affairs,� http://www.hrsdc.gc.ca/eng/labour/labour_globalization/ila/index International Economics. .shtml (accessed June 22, 2010). EU (European Union), Directorate General for Trade. 2008. 13. Human Resources and Skills Development Canada website, “CARIFORUM–EU Economic Partnership Agreement.� Information “Canada–Colombia Agreement on Labour Cooperation,� http://www Paper, EU, Brussels. .hrsdc.gc.ca/eng/labour/labour_agreements/ccalc/index.shtml. Freeman, Richard B., and James L. Medoff. 1984. What Do Unions Do? 14. Human Resources and Skills Development Canada website, New York: Basic Books. “Overview: The International Trade and Labour Grants and Contribution Hafner-Burton, Emilie M. 2009. Forced to Be Good: Why Trade Agreements Programme,� http://www.hrsdc.gc.ca/eng/lp/ila/Overview.shtml (accessed Boost Human Rights. Cornell, NY: Cornell University Press. June 25, 2010). Horn, Henrik, Petros C. Mavroidis, and Andre Sapir. 2010. “Beyond the 15. The letter to Guatemala requesting consultations may be found at WTO? An Anatomy of EU and US Preferential Trade Agreements.� the USTR website, http://www.ustr.gov/trade-agreements/free-trade- World Economy 33 (11): 1565–88. agreements/cafta-dr-dominican-republic-central-america-fta/kirk-solis-le. Hornbeck, J. F. 2009. “Free Trade Agreements: U.S. Promotion and 16. Office of Trade and Labor Affairs, Bureau of International Labor Oversight of Latin American Implementation.� Inter-American Affairs, U.S. Department of Labor, Public Report of Review under the Development Bank Policy Brief, Inter-American Development Bank, Dominican Republic–Central America–United States Free Trade Agree- Washington, DC. ment, Chapter 16: “Labor,� January 16, 2009. Hufbauer, Gary Clyde, and Jeffrey J. Schott, assisted by Paul L. E. Grieco 17. U.S. Trade Representative Ron Kirk, “Remarks on Enforcement at and Yee Wong. 2005. NAFTA Revisited: Achievements and Challenges. Allegheny Technologies, Washington, Pennsylvania,� July 30, 2010, http:// Washington, DC: Peterson Institute for International Economics. Labor Rights 441 ILO (International Labour Office) and IFC (International Finance Corpo- Washington, DC. http://www.usitc.gov/publications/332/journals/ ration). 2007. “Better Work Jordan: Project Document.� Prepared for trade_law_ftas.pdf. Ministry of Labour, Amman, Jordan, and U.S. Agency for Interna- U.S. DOL (Department of Labor), Bureau of International Labor Affairs. tional Development, Washington, DC. 2009. “Progress in Implementing Capacity-Building Provisions Jordan, Ministry of Labour. 2008. “Labour Administration and Compliance under the Labor Chapter of the Dominican Republic–Central in Jordan: A Multi-Stakeholder Collaboration.�Ministry of Labour, America–United States Free Trade Agreement. First Biennial Report Amman. Submitted to Congress Pursuant to Section 403(a)(3) of the Dominican Mayer, Frederick W. 1998. Interpreting NAFTA: The Science and Art of Republic–Central America–United States Free Trade Agreement Imple- Political Analysis. New York: Columbia University Press. mentation Act.� January 14, U.S. DOL, Washington, DC. NLC (National Labor Committee). 2006. “Campaigns: Jordan.� National U.S. GAO (General Accountability Office). 2009. “Four Free Trade Agree- Labor Committee, Pittsburgh, PA. http://www.nlcnet.org/campaigns?id ments GAO Reviewed Have Resulted in Commercial Benefits, but Chal- =0017. lenges on Labor and Environment Remain.� Highlights of GAO-09-439 Polaski, Sandra. 2006. “Combining Global and Local Forces: The Case of (July), GAO, Washington, DC. Labor Rights in Cambodia.� World Development 34 (5, May): 919–32. USTR (Office of the U.S. Trade Representative). 2009. “The President’s Trade Rogowsky, Robert A., and Eric Chyn. 2007. “U.S. Trade Law and FTAs: A Policy Agenda.� USTR, Washington, DC. http://www.ustr.gov/assets/ Survey of Labor Requirements.� Journal of International Commerce Document_Library/Reports_Publications/2009/2009_Trade_Policy_ and Economics (July). United States International Trade Commission, Agenda/asset_upload_file810_15401.pdf (accessed March 12, 2009). 21 Human Rights Susan Ariel Aaronson Many of the world’s most important trading economies signed by Canada, Mexico, and the United States, was the have introduced human rights language into their prefer- first PTA to include explicit human rights provisions. ential trade agreements (PTAs). As a result, more than Trade policy makers agreed to include labor rights in a side 70 percent of the world’s governments now participate in agreement. They also included additional chapters and PTAs with human rights requirements. The growing num- language focused on encouraging transparency (access to ber and scope of these trade agreements reflects a new real- information) and public participation. These obligations ity: policy makers understand that economic integration went beyond the provisions of the General Agreement on will not be successful without a stronger focus on improv- Tariffs and Trade (GATT) and its successor organization, the ing governance among trade partners. If human rights pro- World Trade Organization (WTO); scholars call this WTO+. visions are designed carefully, they can work both to Some analysts see these provisions as “legal inflation� improve governance and to empower people to claim their and assert that governments are using trade agreements to rights. Yet policy makers, scholars, and activists still know globalize their social policies or regulatory approaches. very little about the effects of including human rights pro- They argue that trade agreements are not the right place to visions in trade agreements. address human rights issues, and they point out that trade As long as men and women have traded, they have wres- agreements in themselves, even without special provisions, tled with how to advance human rights while expanding may have positive human rights spillovers. Whatever the trade. In some instances, policy makers have used the arguments, the proliferation of human rights provisions incentive of trade expansion; at other times they have used signals the new reality for trade liberalization. Many PTAs trade sanctions—the disincentive of lost trade—to punish go far beyond commercial policy; they are really gover- officials from other countries that have undermined nance agreements that contain thousands of pages of human rights. For example, after the United Kingdom and obligations related to topics ranging from corporate gov- the United States outlawed the slave trade in 1807, the ernance to environmental policy to human rights. Still, United Kingdom signed treaties with Denmark, Portugal, the association of trade and human rights in PTAs is a rel- and Sweden to reinforce its own ban. After the United atively new phenomenon. In a sense it is a shotgun wed- States banned goods manufactured by convict labor, ding; and it is too early to tell whether this marriage will Australia, Canada, and the United Kingdom adopted be effective and enduring. This chapter examines the who, similar measures. These efforts stimulated international what, when, where, how, and why of the trade–human cooperation, and in 1919 the signatories of the Treaty of rights linkage and why we should care about it. Versailles formed the International Labour Organization Who? The demandeurs for the link include both indus- (ILO) to establish fair and humane rules regarding the trial and developing countries, and at least 131 countries treatment of labor (Bidwell 1939). have accepted such links.1 The United States and the European Union (EU) were the What? Human rights are rights and freedoms to which first trade entities to include human rights language in trade all humans are entitled. Our discussion is limited to only agreements. In the 1980s and 1990s, U.S. and EU officials those human rights set forth in the Universal Declaration began to include human rights conditionality clauses in their of Human Rights (UDHR).2 This study finds that only preference programs (Charnovitz 2005, 29, n. 103–05). The some of the human rights contained in the declaration 1993 North American Free Trade Agreement (NAFTA), have been incorporated into trade agreements. 443 444 Susan Ariel Aaronson When? Policy makers first made an explicit link between human rights language in PTAs negotiated by emerging human rights and trade in the U.S. generalized system of economies. Finally, the third section explores some of the preferences (GSP) program in 1984. NAFTA, signed in problems and questions raised by the union of trade and 1993, was the first preferential trade agreement to include human rights. I then offer a conclusion about why govern- specific human rights language. ments are increasingly wedding trade and human rights Where? The provisions may be found in the preamble, and whether this policy union will thrive. in side agreements, or in the body of the agreement. How? Some countries condition the agreement on the Definitions and Background partner’s changing its laws to meet international standards on Human Rights (the U.S. approach); others commit governments “not to reduce� high standards in the interests of attracting invest- This section discusses the international law of human ment or trade. Examples of the latter approach are the rights and the role of human rights in trade agreements, agreement between the European Free Trade Association including the International Trade Organization (ITO), the (EFTA) and the South African Customs Union (SACU), GATT, and the WTO. We then briefly review the findings of and the PTA between the United States and Colombia scholars active in human rights and trade issues. (Bartels 2009a). The “how� can also relate to whether the demandeur and the target government adopt monitoring Human Rights Obligations of Trading Nations or enforcement strategies in concert with the agreement; whether they link the agreement to capacity building States are obligated to act in certain ways in order to pro- designed to build governance expertise and will; and tect, respect, and advance human rights. These obligations whether one signatory can challenge human rights viola- are delineated in the Universal Declaration on Human tions of the trade agreement or suspend it. Rights (UDHR), which was approved by the members of Why? We discuss below the reasons why nations might the United Nations in 1948 and which spells out more than include human rights provisions and why states accept 30 rights that member states are supposed to promote and them. protect. But the declaration does not legally bind member So what? We also discuss the outcomes of efforts to link states (Petersmann 2000). To ensure that human rights trade and human rights. Some governments have changed would be binding obligations, policy makers developed their attitudes and behavior toward particular human two covenants that included all the UDHR rights: the rights. We don’t know if this change in attitudes and behav- International Covenant on Civil and Political Rights ior toward human rights is temporary or permanent. This (ICCPR), and the International Covenant on Economic, chapter does not cover all the human rights provisions in Social and Cultural Rights (ICESCR).4 The ICCPR enu- all PTAs. The discussion is limited to PTAs with explicit merates the rights that a state may not take away from its human rights objectives, language, or policies, no matter citizens, such as freedom of speech and freedom of move- whether the language occurs in the preamble, in the provi- ment. In contrast, the ICESCR generally defines rights sions of the agreement, or in a side agreement.3 We also (often, necessities) that a state should provide for its citi- explore human rights spillovers from provisions related to zens, such as basic education or health care. The signatories access to information (transparency), political participa- of the ICESCR recognize that governments need expertise tion, and due process, according to which foreign and and funds to provide all their citizens with rights such as domestic producers can comment on policies or regula- access to education, jobs, and health care. But it is difficult tions affecting trade. Although property rights are impor- for governments to advance, respect, and realize human tant human rights, we do not focus on them except when rights; it takes considerable governance expertise, funds, trade agreements mention the intersection of property and will. Accordingly, a government is only obligated to rights with other important human rights such as access to provide these cultural, economic, and social rights as far as medicines (Drahos and Mayne 2002; WHO 2006). “it is able.�5 The chapter is organized in three sections. The first sec- The declaration and the covenants have different stand- tion is foundational; it defines human rights, examines the ings in international law. The Universal Declaration of history of the global trading system and the role of human Human Rights is universal in scope; it applies to everyone, rights, and reviews the literature in this area. The second whether or not individual governments have formally section describes how and why the United States, the EU, accepted its principles or ratified the covenants. The EFTA, and Canada became the main demandeurs of provi- covenants, by their nature as multilateral conventions, are sions for human rights governance. Next, we examine legally binding only on those states that have accepted Human Rights 445 them by ratification or accession. They did not go into have few options short of force for changing the behavior force until 1976, when 35 member states of the United of leaders of other countries. Market access can be an Nations ratified them. However, many nations have not important instrument of leverage because it can affect the ratified both covenants.6 In addition, the Universal Decla- economic and political health of targeted countries. Fur- ration does not include all the human rights found in thermore, in recent years policy makers have come to national constitutions, nor does it include many new understand that failure to protect human rights (such as human rights, such as the right to a healthy environment. labor rights) can affect market access conditions for their Since these newer rights are not embodied in the own producers.7 covenants, they are, thus far, not binding on states, and so Although policy makers may respond to public pressure they are not discussed in this chapter. to use trade to advance human rights, most policy makers Of the 38 human rights set forth in the UDHR, some do not make human rights a top priority for trade policy rights are not affected by or are not relevant to trade, but making. In most countries, policy makers develop trade others, such as labor rights, have been explicitly mentioned policies as though they were strictly commercial instru- in trade agreements (see table 21.1). Table 21.2 summarizes ments. They weigh the interests of their producers and some of the human rights embedded in trade agreements, consumers, and although they may consider national secu- as of 2010. Some human rights provisions are in the pre- rity or political concerns, they rarely introduce the interests amble; others are in the body; and still others are expressed of the global community into such deliberations. Although in side agreements. Some provisions are binding on the sig- policy makers are well aware of the human rights conse- natories of the agreement, and others are rhetorical. quences of some of their trade decisions, they have few Under international law, states are supposed to do incentives to ensure that trade policies advance the human everything in their power to respect, promote, and fulfill rights outlined under the UDHR. Moreover, trade policy human rights. But advancing human rights is not easy. As makers are generally not charged with ensuring that trade noted above, many states are unable to meet all of their policies or trade flows do not undermine specific human “international� human rights responsibilities. Moreover, rights at home or abroad. In trade negotiations, govern- few officials win or maintain office on the basis of their ments are charged with pursuing national commercial efforts to promote the human rights of noncitizens. interests, not global interests (Commission on Human However, many people are not comfortable knowing Rights 2004; 3D 2005; Aaronson and Zimmerman 2007). that other human beings lack basic rights in other coun- tries, or live in countries where government officials ITO, GATT, and WTO Provisions on Human Rights undermine human rights. These individuals may demand that policy makers take action to protect human rights in During World War II, the postwar planners devised several other countries. Trade policy is not the only or the best international institutions to govern the global economy. means of extending such protection, but policy makers They envisioned an International Trade Organization (ITO) Table 21.1. Examples of Human Rights Embedded in PTAs: Demandeurs and Position of Provisions in Agreement Access to Right to Indigenous Political Democratic affordable cultural Freedom of minority participation and Labor rights rights medicines participation movement rights due process Privacy Canada Mercosur Costa Rica Canada EFTA/EEA Canada Canada Canada Preamble and Linked Body Preamble and Body and side Body side agreement protocol side agreements agreements Chile United States New Zealand CARICOM New Zealand EU Side agreement Side letters Body Body Mercosur Australia United States Body Body United States Body New Zealand Source: Susan Ariel Aaronson. Note: CARICOM, Caribbean Community; EEA, European Economic Area; EFTA, European Free Trade Association; EU, European Union; Mercosur, Southern Cone Common Market (Mercado Común del Sur); PTA, preferential trade agreement. 446 Susan Ariel Aaronson Table 21.2. The Universal Declaration on Human Rights and Its Two Covenants International Covenant on Civil International Covenant on Economic, Social and Political Rights and Cultural Rights Right to life (Art. 3) Right to marriage and found a family (Art. 16) Right to liberty (Art. 3) Right to social security (Art. 22) Right to security (Art. 3) Right to work, free choice of employment, just and favorable conditions of work, and protection against unemployment (Art. 23.1) Right to the abolition of slavery and slave trade (Art. 4) Right to equal pay for equal work (Art. 23.2) Right to the prevention of torture or cruel, inhuman, Right to just and favorable remuneration (Art. 23.3) or degrading treatment or punishment (Art. 5) Right to recognition before the law (Art. 6) Right to form and join a trade union (Art. 23.4) Right to equality before the law and to equal Right to rest and leisure, including reasonable limitation of protection of the law (Art. 7) working hours and periodic holidays with pay (Art. 24) Right to effective judicial remedy (Art. 8) Right to a sustainable standard of living (including food, clothing, housing, medical care, and necessary social services); right to security in the event of unemployment, sickness, disability, widowhood, old age, or other lack of livelihood in circumstances beyond his control (Art. 25.1) Right to the prevention of arbitrary arrest, detention, Right to special care and assistance for motherhood and or exile (Art. 9) childhood (Art. 25.2) Right to fair and public hearing by a neutral tribunal (Art. 10) Right to education (Art. 26) Right to presumption of innocence (Art. 11.1) Right to cultural participation (Art. 27.1) Right to nonretroactive penal code (Art. 11.2) Right to the protection of intellectual property (Art. 27.2) Right to privacy (Art. 12) Right to freedom of movement and residence in the country (Art. 13.1) Right to leave the country and return (Art. 13.2) Right to seek and enjoy asylum from prosecution (Art. 14) Right to a nationality (Art. 15) Right to freedom of thought, conscience, and religion (Art. 18) Right to freedom of opinion and expression (Art. 19) Right to freedom of peaceful assembly and association (Art. 20) Right to governmental participation, directly or through freely chosen representatives (Art. 21.1) Right of equal access to public services (Art. 21.2) Right to periodic and fair elections (Art. 21.3) Source: Prepared by Philip Van der Celen; in Aaronson and Zimmerman 2007. Note: Italics indicate that the right is included in one or more preferential trade agreements (PTAs). to reduce barriers to trade. The draft treaty for the ITO was GATT agreed to join a new international organization, the the first trade agreement to include explicit human rights World Trade Organization (WTO). The WTO contains language (related to labor rights). The ITO was designed to the GATT agreement, and it has a stronger system of dis- ensure that signatories to the agreement did not “export pute settlement. GATT and WTO signatories must adhere their unemployment� and thereby undermine the ability of to two key principles to reduce trade distortions: the most workers to provide for their families. In addition, the draft favored nation principle and the national treatment princi- ITO allowed signatories to breach its rules through an ple. The most favored nation principle (MFN) requires “exception� for domestic policies “necessary to protect that the best trading conditions extended to one member public morals� or to protect human or plant life and by a nation must be extended automatically to every other health. (It also included a national security exception.) But nation. The national treatment principle provides that the ITO was abandoned after the U.S. Congress failed to once a product is imported, the importing state may not vote on implementing legislation (Wilcox 1949; Diebold subject that product to regulations less favorable than 1952; Charnovitz 1987). those that apply to like products produced domestically. The end of the ITO was not the death knell of efforts to The WTO does not explicitly prohibit countries from pro- link trade and human rights. In 1995 the members of the tecting human rights at home or abroad, but its rules do Human Rights 447 constrain the behavior of governments in providing that conflict (Aaronson and Zimmerman 2007, 43). In addi- when member states seek to promote human rights, at tion, some members bring up human rights during acces- home or abroad, they must not unnecessarily or unduly sions, when new members are asked to make their trade distort trade. It is hard to use trade to promote human and other public policies transparent, accountable, and rights when nations can’t use trade to distinguish among responsive. They have also discussed human rights issues at those nations that may undermine the human rights of trade policy reviews, when member states review the trade their citizens and those that strive to advance these rights. and governance performance of other member states. The GATT and the WTO do not directly address how Finally, members have discussed some human rights issues governments relate to their own citizens, and they say very during recent trade negotiations: examples include food little about human rights.8 But human rights are seeping security, intellectual property rights (IPRs), and public into the workings of the WTO (Aaronson 2007). Some health (Aaronson 2007). WTO members have used the GATT/WTO exceptions to Although the GATT/WTO contains no explicit human advance human rights abroad or to protect human rights rights provisions, it does refer to human rights implicitly. at home. Under Article XX, nations can restrict trade when Some of these provisions relate to economic rights such as necessary to “protect human, animal, or plant life or the right to property, and others, to democratic and politi- health� or to conserve exhaustible natural resources. This cal rights. For example, under GATT/WTO rules, member article also states that governments may restrict imports states give economic actors “an entitlement to substantive relating to the products of prison labor. Although it does rights in domestic law including the right to seek relief; the not refer explicitly to human rights, the public morals right to submit comments to a national agency or the right clause of Article XX is widely seen as allowing WTO mem- to appeal adjudicatory rulings� (Charnovitz 2001). Mem- bers to put in place trade bans in the interest of promoting ber states must also ensure that “Members and other per- human rights (WTO 2001; Howse 2002; Charnovitz 2005). sons affected, or likely to be affected, by governmental Brazil used the Article XX exception to ban imports of measures imposing restraints, requirements and other bur- retreaded tires, which could not easily be disposed of. dens, should have a reasonable opportunity to acquire The national security exception, Article XXI, states that authentic information about such measures and accord- WTO rules should not prevent nations from protecting ingly to protect and adjust their activities or alternatively to their own security. Members are not permitted to take seek modification of such measures.�9 These can be termed trade action to protect another member’s security or to due process, information, and political participation rights protect the citizens of another member. If, however, the (Powell 2005); see box 21.1.10 United Nations Security Council authorizes trade sanc- tions, WTO rules allow countries to use such measures to A Brief Review of the Literature on Trade promote human rights, as when sanctions were instituted and Human Rights against South Africa’s apartheid regime in the 1980s (Aaronson and Zimmerman 2007, 19). In recent years, scholars from many disciplines have exam- Members of the GATT/WTO can use other avenues to ined the relationship between trade and human rights. protect human rights at home and within other member Many economists argue that human rights is not a trade states (table 21.3). In recent years, member states have used issue, but trade can have positive human rights spillovers. temporary waivers of GATT rules to promote human As trade expands, individuals exchange ideas, technologies, rights. For example, after the UN called for a ban in trade processes, and cultural norms and goods. With more in conflict diamonds, WTO member states agreed to tem- trade, people in countries with fewer rights and freedoms porarily waive WTO rules to allow trade in only those become aware of conditions elsewhere, and with such diamonds certified by the Kimberley process to be free of knowledge, they may demand greater rights. Isolated Table 21.3. Examples of Avenues and Actions at the WTO Related to Human Rights, 2005–10 Avenue Human right affected Accessions Labor rights, access to information, due process (Vietnam, Saudi Arabia, Cambodia) Trade policy reviews Members discussed labor rights, women’s rights, access to medicines Disputes Right to health (Brazil tires) Negotiations Access to safe, affordable food Source: Susan Ariel Aaronson. 448 Susan Ariel Aaronson Box 21.1. Transparency, Due Process, and Democracy Spillovers from the WTO From 1948 to 1964, contracting parties to the General Agreement on Tariffs and Trade (GATT) were required to promptly publish laws, regulations, and judicial decisions affecting imports and exports (GATT Article X). In this way, exporting interests could learn about legal developments affecting trade and respond to them. GATT contracting parties gradually strengthened these notification requirements, and members were required to administer trade-related laws, regulations, rulings, and agreements in a uniform, impartial, and reasonable manner. Today, the World Trade Organization (WTO) has strong rules for transparency and due process. It requires governments to make their trade laws and regulations transparent and public and to allow citizens to comment on and challenge these laws and regulations. However, neither the GATT nor the WTO requires that members involve their publics in trade policy making. Moreover, many countries do not have a free press, adequate funds, informational infrastructure, or the political will to effectively involve their citizenry in public policy making. The GATT/WTO may also have some unintended human rights spillovers. Member states must provide the same rules and privileges to domestic and foreign actors. These provisions may prod policymakers to provide access to information and enforce rights to public comment in countries where governance is not transparent and participatory. In repressive states, WTO rules may empower domestic market actors (consumers and taxpayers, as well as producers) who may not have been able to use existing domestic remedies to obtain information, influence policies, or challenge their leaders (Aaronson and Abouharb forthcoming). In WTO countries without a strong democratic tradition, member states may make these changes because they want to signal investors that they can be trusted to enforce property rights, uphold the rule of law, and act in an evenhanded, impartial manner (Dobbin, Simmons, and Garrett 2007; Büthe and Milner 2008; Mansfield and Pevehouse 2008, 273). societies, by contrast, may be more prone to human rights agreement.� Bourgeois, Dawar, and Evenett (2007) argue abuses (van Hees 2004). Thus, many of these economists that the current approach to mainstreaming labor rights in conclude that policy makers need not include human PTAs is ineffective because the provisions commit parties to rights provisions in trade agreements (Bhagwati 1996, 1; enforce domestic labor law only. Horn, Mavroidis, and Sapir Sykes 2003, 2–4). Other analysts disagree; they believe that (2010) conclude that U.S. and EU environmental and labor human rights are trade issues, and they cite history and the standards in PTAs are groundbreaking “means for the two increasing number of human rights provisions in PTAs as hubs to export their own regulatory approaches to their PTA evidence for their perspective. Some legal scholars have partners.� In short, some scholars see the link as ineffective; proposed ways of finding common ground between WTO others as a means of exporting governance. But these schol- trade law and international human rights law (Charnovitz ars did not examine the panoply of trade–human rights 1994; Dunoff 1999; Garcia 1999; Mehra 1999; Petersmann links; they have focused only on labor rights 2001). Some believe that the best strategy is to enhance the Scholars who have examined human rights provisions international human rights system and make it more like in PTAs agree that these provisions are intended to improve the WTO, with stronger dispute settlement and enforce- governance and advance human rights. Petersmann ment mechanisms. Others believe that the WTO should (2006), a legal scholar, has argued that governments use have explicit human rights provisions (Lim 2001). Some their PTAs to achieve extraterritorial political reform. As academics have used case studies to discuss the relationship evidence, he cites the growing number of governments that between trade agreements and specific human rights, such explicitly refer to human rights as an objective or as a fun- as the right to food (Cottier, Pauwelyn, and Bürgi 2005). damental principle of economic integration. Damro Others have suggested bridging mechanisms to ensure bet- (2006) argues that governments include extensive human ter dialogue and coordination between trade and human rights and rule of law provisions in their PTAs because they rights officials (Petersmann 2002). Finally, some scholars recognize they must develop coordinated policies in order have examined how the WTO’s dispute settlement system to address regional threats to security, such as environmen- might address a trade dispute involving human rights (Bal tal damage, illegal migration, drug smuggling, and interna- 2001; Marceau 2001). However, as policy makers began to tional terrorism. refocus their trade liberalization efforts on new PTAs, the In a number of studies, Bartels examines how govern- debate over how best to reach trade and human rights goals ments incorporate human rights into their trade agree- has moved to examining the record of these PTAs. ments (Bartels 2005b, 2008, 2009a, 2009b, and, for an Most of the scholars who have examined human rights analysis of the objectives of the agreements, Bartels 2005a). provisions focus on labor and environmental language— He concludes that provisions linking trade and human what some call “trade and� provisions. Dawar (2008) finds rights are useful because they set up mechanisms for dia- that labor and environmental provisions “constitute an logue, allow civil society in multiple countries to monitor unnecessary, inefficient and inappropriate use of a trade compliance with international norms, and make human Human Rights 449 rights part of the trade relationship. Hafner-Burton domestic reform process. Clearly, to better understand the (2009, 22) compares EU and U.S. approaches toward link- impact of the PTAs, we need both more empirical work and ing human rights and trade in their GSP and free trade country-specific case studies. agreements. She notes that although policy makers may be motivated by “protectionist intent,� the agreements appear Case Studies: PTAs and Human Rights to be having a positive impact on the realization of human rights in many countries. Aaronson and Zimmerman (2007, This section examines how Canada, the EU, EFTA, and the 207) compare how the United States, the EU, South Africa, United States incorporate human rights provision into and Brazil make trade policy and find that governments are their PTAs: table 21.4 summarizes their approaches.11 increasing the scope of human rights, as well as the number Table 21.5 examines these provisions according to specific of agreements with human rights provisions. They con- human rights and shows that industrialized countries are clude that if people are the “wealth of nations,� policy not the only countries to link human rights and trade in makers that weigh human rights as they make trade policy PTAs. We begin our analysis with Canada, which has are more likely to ensure that their citizens thrive at the become an enthusiastic negotiator of PTAs. intersection of trade and human rights. Few scholars have examined the PTA–human rights nexus Canada empirically. Hafner-Burton (2009, 160–64) focuses on physi- cal integrity rights such as freedom from arbitrary imprison- Canada is a trade-dependent nation; trade represents ment and finds that about 82 percent of the countries that more than 70 percent of its gross domestic product have a PTA with the EU improve their human rights protec- (GDP). In recent years the Canadian government has tion, as against 75 percent for countries without a trade embraced PTAs, on the grounds that this strategy will agreement. However, she relies on personal integrity rights ensure trade-related economic growth and international (for example, freedom from arbitrary imprisonment) to competitiveness.12 Canadian policy makers assert that make a generic case about human rights, which is not fully these agreements can help Canada foster a commitment convincing. PTAs may have different effects on different to human rights, freedom, democracy, and the rule of human rights. Finally, some scholars have examined whether law.13 In Canada the executive branch makes trade policy, PTAs serve as an anchor or lock-in mechanism for domestic which is then approved by the parliament. Although reforms, including laws advancing human rights. U.S. Inter- there is no explicit mandate regarding the relationship national Trade Commission (USITC) economist Michael between trade and human rights, the Canadian govern- Ferrantino (2006) examined negotiations on PTAs with the ment has included several types of human rights in its United States. He argues that these agreements may improve recent trade agreements: labor rights, cultural rights, governance but warns that it is difficult to ascertain whether a indigenous rights, and rights to political participation particular reform is stimulated by negotiations or by the and due process.14 Table 21.4. Human Rights in Preferential Trade Agreements: Comparing EFTA, the EU, the United States, and Canada EFTA EU United States Canada Strategy Universal human rights Universal human rights Specific human rights Specific human rights and specific rights Which rights? Labor rights, transparency, Transparency, due process, Transparency, due due process, political political participation, process, political participation, and access to affordable participation, labor privacy rights medicines, and labor rights rights, privacy rights, cultural and indigenous rights How enforced? No enforcement Human rights violations lead In newest agreements, labor Only labor rights to dialogue and possible rights can be disputed under (monetary penalties); suspension, depending on a dispute settlement body use dialogue first nature of violation. affiliated with the agreement; process begins with bilateral dialogue to resolve issues Any challenge? First challenge: Guatemala Source: Susan Ariel Aaronson. Note: EFTA, European Free Trade Agreement; EU, European Union. 450 Susan Ariel Aaronson Table 21.5. Examples of Human Rights Embedded in Preferential Trade Agreements Access to Democratic affordable Right to cultural Freedom of Indigenous Political Labor rights rights medicines participation movement rights participation Privacy Canada Mercosur Costa Rica Canada EFTA Canada Canada Canada Chile United States New Zealand EEA New Zealand United States EU Mercosur CARICOM Australia United States New Zealand Source: Susan Ariel Aaronson. Note: CARICOM, Caribbean Community; EEA, European Economic Area; EFTA, European Free Trade Agreement; EU, European Union; Mercosur, Southern Cone Common Market (Mercado Común del Sur). As of March 2011, Canada has negotiated eight PTAs: labor and environmental provisions cannot be included in the Canadian–U.S. Free Trade Agreement (CUSFTA, now the body of a trade agreement but must be in side agree- part of NAFTA), and agreements with Peru, EFTA, Costa ments, termed labor cooperation agreements.20 According Rica, Chile, Israel, Jordan, Panama, and Colombia, of to Human Development and Skills Resources Canada, which seven are in force. The Canadian government is which negotiates and monitors labor rights internationally, soliciting public comment on negotiations with about Canada has ratified six labor cooperation agreements in its 12 other countries or trading entities. It is engaged in active PTAs with Chile, Costa Rica, Colombia, Jordan. Peru, and negotiations with many of these countries or entities.15 NAFTA. (Interestingly, Canada has also negotiated memo- Canadian officials clearly see these trade agreements as randa of understanding on labor cooperation outside its governance agreements, although they do not call them PTAs.)21 The labor side agreements state that signatories that. The government asserts that “Canadians recognize must ensure that their labor laws comply with ILO stan- that their interests are best served by a stable, rules-based dards and must establish offices to evaluate complaints international system. Countries which respect the rule of related to labor rights. In this way, the side agreements are law tend to respect the rights of their citizens, [and] are clearly governance agreements, since they attempt to build more likely to benefit from development.�16 The Depart- labor rights governance capacity. ment of Foreign Affairs and International Trade (DFAIT) The recent agreements also include a nonderogation also notes that “The UN Charter and customary interna- clause stating that neither party shall waive or otherwise tional law impose on all countries the responsibility to pro- derogate from, or offer to waive or otherwise derogate mote and protect human rights. This is not merely a ques- from, its labor laws in order to encourage trade or invest- tion of values but a mutual obligation of all members of ment (Article 2). Both parties commit not only to core the international community.�17 standards but also to acceptable occupational health and Overview of Canada’s human rights–trade strategy. safety protection and acceptable minimum employment Canada’s approach toward embedding human rights is standards. They also agree to provide migrant workers with both broad and specific. The preambles of recent agree- the same legal protections as nationals with respect to ments with EFTA, Jordan, Peru, and Colombia refer to working conditions. human rights objectives, citing the UDHR, labor rights, These provisions go beyond ILO core labor standards cultural participation, and protection of human rights and because they also focus on both the demand (public) and freedoms.18 These agreements include chapters with lan- supply (policy maker) sides of good labor governance. Sig- guage on labor rights, transparency, and the environment. natories are required to educate and involve their publics The agreements also contain a chapter on exceptions (akin regarding their rights under labor law. Article 4 articulates to those in the WTO) and a provision safeguarding the a right to private action: “Each party shall ensure that a right to regulate and to maintain high standards. The person with a legally-recognized interest . . . has appropriate exceptions chapter notes that nothing in the agreement is access . . . to administrative or tribunal proceedings . . .�22 to apply to cultural sectors and mentions the need to be Article 5 contains procedural guarantees designed to ensure supportive of trade waivers.19 that proceedings “are fair, equitable, and transparent and Labor rights. Because of the division of powers in the respect due process of law.� Canadian policy makers seem areas of environmental and labor regulation under the to agree that by educating foreign workers as to their rights, Canadian constitution, Canadian policy makers believe that these workers are more likely to use these rights. Human Rights 451 Due process provisions. Canada has expanded on the tribute to the conservation and sustainable use of biological WTO’s due process provisions for trade-related policy diversity, subject to national legislation.�26 Because the pro- making. Chapter 19 (transparency) in recent Canadian vision is aspirational, nonbinding, and not disputable, some PTAs requires each party to “ensure that in its administra- Canadian nongovernmental organizations (NGOs) see it as tive proceedings . . . persons of the other Party that are inadequate. directly affected by a proceeding are provided reasonable Response of Canadian NGOs to these human rights provi- notice . . . and reasonable opportunity to present facts and sions. Canadian NGOs working on development, human arguments in support of their positions.�23 The environ- rights, and labor rights remain critical of the country’s mental side agreement also contains due process require- approach. They demanded that Canada perform human ments.24 Here again, Canada uses human rights language rights impact assessments before it initials trade agree- to prod its PTA partners to make trade-related policies in a ments with countries, such as Colombia, where the rule of transparent and accountable manner. law is inadequate.27 They argue that such assessments must Political participation provisions. The Canadian govern- be conducted by international human rights bodies rather ment has incorporated several references to political than by the two governments (CCIC 2009). This debate has participation into its recent trade agreements. This lan- influenced Canada’s parliament, which seems increasingly guage is included in Chapter 19 on transparency, as well as interested in the relationship between trade agreement in the labor and environmental side agreements. The trans- provisions and human rights and in whether Canada’s parency chapter obligates signatories to include regulations approach to dealing with these rights is effective.28 In guaranteeing public participation, public comment, and 2010 Canada and Colombia agreed to perform yearly the ability to challenge relevant regulations. The environ- human rights impact assessments of their PTA.29 Canada mental chapter commits the parties “to promote public became the first nation to require such assessments with awareness of environmental laws and policies by ensuring future trade agreements. However, as of March 2011, that information regarding environmental laws and poli- Canada has yet to do such a human rights impact assessment cies, as well as compliance and enforcement is available to or to provide the public with information as to how it will the public.� These provisions also commit the countries to evaluate human rights at home and abroad. ensure that the public is able to participate in environmen- Canada’s PTAs and human rights: A summary. The tal assessment procedures. In addition, they include a pro- Canadian government has incorporated a wide range of vision allowing any person residing in or established in the human rights into its PTAs. Canada sees its human rights territory of either country to submit a written question to and trade work as complementary. Its approach is both either country obliging the country to make the questions hortatory and pragmatic, making use of language that sets and responses available to the public. Although the lan- forth explicit obligations and delineates objectives that the guage is binding, it is also relatively weak: the signatories signatories will strive or endeavor to meet. are asked to strive to cooperate in these areas, endeavor to Some of this language, such as obligations focused on engage the public, or as the Colombia labor side agreement transparency and political participation, reflects long- says, “encourage[e] education of the public regarding its standing Canadian norms on how to govern. Some of the labor laws� (emphasis added). provisions are designed to encourage some of Canada’s Cultural reservations and exemptions. Canada includes trade partners to comply with international human rights provisions in all its PTAs to ensure that these trade agree- norms such as labor rights. But Canadian policy makers ments do not affect Canada’s ability to maintain the cultural recognize that it is not sufficient for outsiders to demand heritage of the Canadian people or to determine Canadian good governance. The public, both in Canada and in cultural policies. These provisions are contained in the Canada’s trade partners, must be informed about and agreements’ chapters on exceptions. The agreements define involved in the development of rules if these rules are to be Canada’s cultural industries as persons engaged in the pub- perceived as evenhanded and effective. Yet the Canadian lication, distribution, or sale of books, magazines, periodi- public has not been very supportive of Canada’s approach. cals, newspapers, music, films and videos, and so on.25 Many Canadian NGOs see Canada’s PTAs as opaque, Indigenous rights. The Canada–Colombia (as well as ineffective at improving governance, and undemocratic Canada–Peru) side agreement on the environment states, (CCIC 2009). “The Parties also reiterate their commitment, as established Canada does not require all of its trade partners to by the Convention on Biological Diversity, to respect, pre- adopt these human rights obligations. For example, it has serve and maintain traditional knowledge, innovations and not embedded many of these provisions in its PTA with practices of indigenous and local communities that con- EFTA, but, as the next section shows, neither EFTA nor 452 Susan Ariel Aaronson Canada “requires� such obligations in its PTAs. It will be to implement the objectives of the agreement and support interesting to see how Canada negotiates with the EU on partner countries’ efforts to achieve sustainable economic labor and other human rights in preparation for a transat- and social development.34 Some recent PTAs, such as those lantic free trade agreement. Negotiators working on scop- with Canada, Colombia, and the Gulf Cooperation Coun- ing documents for these negotiations agreed that “it would cil (GCC), refer to ILO principles, but contracting parties be appropriate to address sustainable development issues have no obligations in the agreement regarding the ILO. through provisions on the environment and labor rights, Since late 2008, EFTA has been exploring whether to including the core labor standards embodied in the 1998 include environmental and labor and social standards in its ILO Declaration . . . Such provisions could include, inter PTAs. The Committee of Members of Parliament of the alia: the right to regulate while aiming for high levels of EFTA countries has commissioned a study on environmen- protection; effective enforcement of environment and tal policies and labor standards in PTAs but is not looking labor laws; a commitment to refrain from waiving such at human rights per se as an issue to be included in trade laws in a manner that affects trade or investment; a frame- agreements.35 Regarding the addition of labor and envi- work for cooperation; public involvement; and mecha- ronmental provisions, the study notes that although a nisms to monitor and address disputes.�30 growing number of nations have such provisions, they may “breach� the regulatory framework of any of the parties to a PTA. The authors stress that much of the evidence European Free Trade Association (EFTA) regarding the quality of these provisions is qualitative EFTA began in 1960 as a framework for liberalizing trade rather than quantitative, and they conclude that it is diffi- and promoting economic cooperation among several cult to assess the effectiveness of their inclusion.36 EFTA European nations. At that time, some EFTA member states has set up working groups to examine whether or how to were unwilling or unable (for reasons of sovereignty or embed such provisions, but as of the time of writing neutrality) to join the EU. The membership of EFTA has (March 2011), they had not reached a decision. Thus, EFTA changed over time; Austria, Denmark, Finland, Portugal, does not seem to be enthusiastic or to see an urgency about Sweden, and the United Kingdom left to join the EU. Ice- including further provisions in its PTAs. land, Liechtenstein, Norway, and Switzerland are the mem- Human rights in the EEA treaty. Although EFTA has not bers of EFTA as of March 2011. played a leading role in linking trade and human rights, In recent years, EFTA has become an active PTA nego- the European Economic Area (EEA) treaty—EFTA’s main tiator. EFTA has 22 free trade agreements, covering agreement with the EU—does use the language of 31 countries.31 It is now negotiating with Algeria; Bosnia human rights. This agreement, however, is much more and Herzegovina; Hong Kong SAR, China; India; than a trade agreement. It is designed to ensure “four Indonesia; Montenegro; the Russian Federation, Belarus, freedoms�—free movement of goods, services, persons, and Kazakhstan; and Thailand. EFTA’s PTAs do not and capital—throughout the 30 EEA states. The agree- include explicit human rights provisions, although trade ment guarantees equal rights and obligations within the policy makers have included references to human rights in internal market for citizens and economic operators in both the preamble and the investment chapters of several the EEA. The provisions on social security are meant to agreements (Bartels 2009a). The preamble typically men- coordinate the respective national systems of the EEA states tions the desire to create new employment opportunities, and thus to ensure social protection in case of, among improve working conditions and living standards, and other contingencies, sickness, maternity, invalidity, death, promote sustainable development, and it reaffirms the or unemployment. EFTA also has language to safeguard the parties’ “commitment to the principles and objectives set pension rights of persons who exercise their fundamental out in the UN Charter and Universal Declaration.� But the right to move and reside freely within the EEA. agreements do not aim to improve human rights or strive The preamble of the EEA agreement discusses “the con- toward sustainable development; they do not go beyond tribution that a European Economic Area will bring to the WTO exceptions regarding public morals, public health, construction of a Europe based on peace, democracy and and trade in goods made with prison labor.32 For FTAs human rights.� The preamble also notes the importance of with African states, EFTA includes language in its invest- “equal treatment of men and women� and cites the signa- ment chapter noting that “the parties recognize that it is tories’ desire “to ensure economic and social progress and inappropriate to encourage investment by relaxing health, to promote conditions for full employment, an improved safety, or environmental standards.�33 The agreements also standard of living and improved working conditions.�37 note that the EFTA states are to provide technical assistance Part III, Article 28, of the agreement refers to free movement Human Rights 453 of persons. It seeks to secure freedom of movement for This human rights clause is given operative effect through workers through “the abolition of any discrimination a “nonexecution� clause stating that a failure to fulfill an based on nationality between [European Community] obligation under the agreement, including human rights member states and EFTA states.�38 Taken in sum, the EEA obligations, entitles the other party to take appropriate agreement is very much a human rights document. measures, subject to a consultative procedure (Bartels 2009a). EFTA’s PTAs and human rights: A summary. Although As of October 2009, the EU also included human rights EFTA includes some human rights language in its PTAs, it clauses in nine regional trade agreements (association is just beginning to examine whether it should go further, agreements), with Algeria, the Arab Republic of Egypt, by making human rights provisions actionable and/or dis- Israel, Jordan, Lebanon, Morocco, the Palestinian Author- putable. EFTA policy makers might seek inspiration in ity, the Syrian Arab Republic, and Tunisia. Similar human their most important agreement, the treaty that set up the rights clauses were also incorporated into stabilization and EEA. We can best understand this trade agreement as a accession trade agreements with Croatia and the former governance agreement that harmonizes a wide range of Yugoslav Republic of Macedonia; in pending agreements laws and regulations that could distort trade but does so in with Albania, Bosnia and Herzegovina, and Montenegro; ways that enhance human rights. and in regional trade agreements with Chile, Mexico, and South Africa.43 The EU is currently negotiating a PTA with Canada. As noted above, it will be interesting to see how European Union (EU) the two reconcile their unique human rights and trade The EU is the behemoth of world trade; it is the world’s strategies. largest trade bloc.39 It is also the most enthusiastic propo- Economic partnership agreements (EPAs). The nations of nent of the inclusion of human rights provisions in the EU have long-standing trade relationships with their PTAs. With nearly 500 million citizens, the EU possesses former colonies. Not surprisingly, the EU is determined to approximately a quarter of the world’s economic wealth. maintain these relationships, but also to use them to foster Given its size and political influence, its policies move development and economic growth and to improve gover- markets. nance. In 1975, the members of the EU and their 79 former The EU is an active participant in and negotiator of colonies signed a treaty, the Lomé Convention, which set bilateral and regional PTAs. Since 1995, the European out standards for development cooperation. In 2000 the Commission Directorate General for Trade, which makes EU and its Lomé partners adopted the Cotonou Agree- trade policy for the members of the EU, has included social ment.44 The Cotonou Agreement was based on four and labor clauses in all its free trade agreements.40 EU pol- principles: equality of partners; political participation; icy makers have incorporated social and labor clauses into dialogue and mutual obligations, including human rights more than 50 trade agreements involving more than obligations; and differentiation, based on the idea that 120 countries. The EU includes human rights in many of each country is unique. Under the Cotonou Agreement, its agreements with other countries, including its partner- the EU and its development partners agreed to develop ship and cooperation agreements41 and its generalized regional trade agreements (called economic partnership system of preferences (GSP) program with developing agreements) with regionally specific rights and obligations. countries. However, in this chapter we focus only on the The EU is currently negotiating with, or has completed economic partnership agreements, such as that with the negotiations with, six regional groupings of countries: Caribbean Forum of African, Caribbean, and Pacific (ACP) West Africa (Ghana); the Southern African Development States (CARIFORUM) and recent free trade agreements. Community (SADC), with Botswana, Lesotho, Namibia, The EU and its member states have a long history of Mozambique, and Swaziland as members; Pacific (Fiji and using trade agreements to promote human rights. In 1992 Papua New Guinea); Eastern and Southern Africa (ESA), the European Commission included “an essential elements encompassing the Comoros, Madagascar, Mauritius, the clause� in trade agreements with developing countries. The Seychelles, Zambia, and Zimbabwe; and the East African clause states that respect for human rights is an essential Community (EAC), composed of Burundi, Kenya, element of the agreement. The signatories to such agree- Rwanda, Tanzania, and Uganda.45 The EU and the CARI- ments agree that either party may suspend the agreement FORUM countries concluded negotiations on their EPA in without notice if these “essential elements� are violated. In October 2008.46 The EU uses the CARIFORUM EPA agree- 1995 the EU decided to include the human rights clause ment as a model for delineating human rights obligations in all future international trade agreements, whether with in other EPAs. Although several countries have signed developing or with industrial countries (Bartels 2005a).42 these regional agreements, the European Commission is 454 Susan Ariel Aaronson waiting on other countries to approve these agreements, typically covered in the WTO, such as investment, the which are quite controversial both in Europe and in the environment, and social and labor clauses.51 As of March developing world.47 2011, the EU was actively negotiating FTAs with Colombia, The CARIFORUM agreement refers to human rights, Peru, Central America, Canada, India, Malaysia, Singapore, democratic principles, and the rule of law as essential the Southern Cone Common Market (Mercosur, Mercado elements of the Cotonou Agreement and cites “good gov- Común del Sur), and Ukraine, as well as regional and ernance� as a fundamental element.48 The signatories bilateral Euro-Mediterranean agreements.52 The European agree to respect basic labor rights in line with their ILO Commission completed its first FTA, with the Republic of commitments. Article 3 of the agreement states that the Korea, in 2009. In February 2011, the European Parliament parties agree to work cooperatively toward the realiza- approved by agreement.53 tion of “sustainable development centered on the human The Korean agreement illuminates the EU’s new person.� Article 5 continues with that focus on human approach toward linking sustainable development (social beings, noting that the parties will continuously monitor and environmental) clauses to trade. The preamble reaf- the agreement to ensure that benefits for people are max- firms both parties’ commitment to the Universal Declara- imized. Article 32 calls for transparent rules and for pub- tion of Human Rights and to sustainable development. It lic explanation of legislation. Labor issues are delineated in also notes the parties’ desire to “strengthen the development Chapter 5; the EU refers to labor provisions as the “social and enforcement of labor and environmental laws and poli- aspects� of the agreement. In Article 191 the parties reaf- cies, promote basic workers’ rights and sustainable develop- firm their commitment to labor standards and decent ment and implement this agreement in a manner consistent work; in addition, the parties recognize the benefits of fair with these objectives.�54 Like the CARIFORUM agreement, and ethical trade products. Finally, Article 193 is the non- this FTA contains language protecting the right to privacy. derogation clause, which requires the parties not to try to Article 7.43, in the services chapter, states that each party gain competitive advantage by lowering standards or should reaffirm its commitment to protecting the funda- ignoring their laws. Article 195 sets out a process for con- mental rights and freedoms of individuals and should sultation and monitoring and creates a committee of adopt adequate safeguards for the protection of privacy. experts to examine compliance with the agreement. The The agreement with Korea includes a separate chapter committee can be called on to examine concerns among on trade and sustainable development. Article 13.4 com- the members regarding “obstacles that may prevent the mits the two parties to respect, promote, and realize core effective implementation of core labor standards.�49 Chap- labor rights. The parties also reaffirm their commitment to ter 6 of the agreement refers to trade in data. Articles 196 effectively implement the ILO conventions that both states and 197 continue the focus on individuals; here the parties have ratified. Article 13.6 states that the parties “shall strive recognize their “common interest in protecting fundamen- to facilitate and promote trade in goods that contribute to tal rights and freedoms of natural persons, and in particu- sustainable development, including goods that are the lar, their right to privacy, with respect to the processing of subject of schemes such as fair and ethical trade.� Article personal data.�50 13.7 contains the nonderogation clause, in which the par- Although these agreements contain considerable ties agree to effectively enforce environmental and labor human rights language, most of it is aspirational and non- laws and not to weaken, waive, or derogate from those binding. Only the sustainable development chapters of the laws in a manner affecting trade or investment. Finally, EPAs, on the environment and on social issues, employ Article 13.9 commits the parties to “introduce and imple- public scrutiny, expert panels, and consultations as means ment any measures aimed at protecting the environment of resolving disputes (Bartels 2005a, iv–vi). The agree- and labor . . . in a transparent manner with due notice ments have no dispute settlement mechanism. The only and public consultation.� To achieve that goal, the agree- way to hold governments to account is for citizens to mon- ment creates a unique domestic advisory group on sustain- itor human rights violations and press policy makers to able development, as well as a civil society forum on these discuss any such violations bilaterally. In short, the EU issues (Annex 13). Thus, the agreement sets up a citizen relies on negotiations to monitor human rights linked to monitoring process. trade in these EPAs. The new paradigm includes a strategy for government Free trade agreements (FTAs). In 2006 EU member states consultations on social issues, as well as a panel of experts agreed to negotiate what it calls FTAs with several rapidly to examine issues that cannot be settled through govern- growing Asian nations. The EU made it clear that these mental consultations. In contrast with the U.S. model, the trade agreements would include a wide range of issues not EU has designed no formal mechanism for trade disputes Human Rights 455 related to these issues; instead, civil society will monitor In considering a potential agreement, the EU hires commitments (Articles 13.14 and 13.15). consultants to assess the impact of its trade agreements on The new model PTA also has a transparency chapter sustainable development. It seems strange that these assess- (Chapter 12) requiring that information be made available ments do not focus on the bulk of human rights delineated to all interested persons, as well as rules for public com- in the UDHR (Aaronson and Zimmerman 2007, 138–43; ment (Article 12.3) and due process (Article 12.5, on Bartels 2009a). EU policy makers believe that dialogue and administrative proceedings). This new chapter is very sim- capacity building are the best means of changing the ilar to the U.S. and Canadian transparency chapters, but it behavior of other countries.59 EU policies focus on the also includes a nondiscrimination section that provides, supply side of governance but seem to be less focused on in part, “Each party shall apply to interested persons of the empowering citizens in other countries. Moreover, the EU other Party transparency standards no less favourable than rarely cuts trade or adopts trade sanctions toward trade those accorded to its own interested persons, to the inter- agreement partners that may violate human rights. Thus, ested persons of any third country, . . . whichever are the despite the EU’s professed belief in the universality and best.� Finally, Protocol 3 provides for cultural cooperation indivisibility of international human rights, policy makers and recognizes cultural diversity and cultural heritage.55 are sending a message that some rights are more important The new FTA model, as embodied by the FTA with than others and that only some countries can be prodded Korea, contains a wide range of human rights provisions, with trade policy tools to change their behavior.60 some of which are binding on the signatories. But, as with the EPAs, neither policy makers nor citizens can challenge United States another state’s nonperformance. Thus, this model is unlikely to satisfy the many NGOs that are critical of how The United States bears much of the responsibility for the the EU links trade and human rights.56 Moreover, the EU world’s recent renewed focus on PTAs. In 2001 U.S. Trade does not examine the broad impact of its trade policies on Representative Robert Zoellick proposed that the U.S. gov- human rights, but it does hire independent consultants to ernment reorient trade liberalization toward bilateral and carry out sustainability impact assessments. These consult- regional PTAs in the hope of encouraging “competitive trade ants look at income, poverty, and biodiversity; except in the liberalization.� He theorized that if countries saw significant area of gender inequality, they do not attempt to assess how progress in bilateral PTAs, they might accept deeper market the provisions of a free trade agreement might affect access concessions at the WTO. But instead of stimulating a human rights conditions in trading-partner nations.57 renewed commitment to multilateral trade talks, the U.S. Given their concerns about the EU model, some NGOs focus on PTAs prodded other countries to negotiate their have asked EU policy makers not to conclude or ratify trade own PTAs, which in turn has stimulated ever more PTAs.61 or partnership agreements with countries that have ques- U.S. policy makers don’t describe American PTAs as tionable human rights records. They fear that such agree- governance agreements, but the United States does use the ments could strengthen or reward repressive regimes.58 EU lure of its huge market to encourage other countries to policy makers, however, are not eager to cut off trade in the make significant policy changes. The United States has interest of promoting human rights, and they believe it is pushed for governance improvements that protect foreign important to use trade as a tool for shaping relations with investors. Recent research has found that these provisions emerging markets. have human rights spillovers—they seem to empower The EU’s PTAs and human rights: A summary. At first domestic as well as foreign actors who gain benefits from glance, the EU’s approach is supportive of the human increased transparency, greater evenhandedness, and the rights set forth in the UDHR. The EU takes the position due process rules promoted in these agreements. But the that human rights are universal and indivisible and that United States promotes only some human rights in its these rights are key aspects of the rule of law. Although the trade agreements. At the behest of labor unions and their trade agreements include considerable human rights lan- congressional allies, the United States is most concerned guage, much of it is rhetorical, nonbinding, and not dis- about using these agreements to advance labor rights putable. In its new FTAs and EPAs, the EU not only refers among U.S. PTA partners. U.S. trade policy is complicated, to the UDHR; it also emphasizes sustainable development confusing, and often inconsistent, vacillating between (social and environmental issues). This approach creates market-opening strategies and protectionist measures. To new advisory roles for experts and civil society, but neither some degree, this is because no one individual is in charge states nor individuals can use these provisions to challenge of trade policy making; authority is shared between the human rights violations. legislative and executive branches. The 535 members of 456 Susan Ariel Aaronson Congress and the president have divergent views about trade The U.S.–Colombia FTA has been pending since 2006. and about the linkage between trade and human rights Many Democratic members of Congress have signaled that (Smith 2006). Members of Congress “think local� and, in they cannot vote for the agreement because of Colombia’s times of recession, are focused on local economic growth. problems with labor rights. They note the high rate of vio- Without a great understanding of what trade agreements do, lence (murders, arbitrary detentions, and kidnappings of many members have little enthusiasm for negotiating them. trade unionists), as well as weak enforcement of labor laws. As a result, Congress has not provided authority to negotiate Colombia’s labor rights problems are part of a greater new trade agreements at the bilateral or multilateral level problem of impunity, inadequate governance, and corrup- since 2002. Yet, during the eight years of the George tion (Bolle 2009). The pending PTA is not designed to W. Bush administration, the U.S. Trade Representative address these issues per se, yet its fate rests on the public’s (USTR) negotiated trade agreements with Australia, Bahrain, and policy makers’ perception of these problems as mainly CAFTA–DR (Central America Free Trade Agreement— labor rights issues. In this way, the PTA illuminates the Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua— inadequacies of the U.S. focus on labor rights, rather than and the Dominican Republic), Chile, Colombia, Jordan, good governance per se.64 Morocco, Oman, Panama, Peru, and Singapore (U.S. GAO Moreover, the United States seems to be moving from 2009). As of March 2011, Congress had approved all but the using these PTAs as an incentive to improve human rights agreements with Colombia, Korea, and Panama. The to using them to hold nations accountable for specific Obama administration has agreed to move these agree- human rights performance. In July 2009 the Obama ments forward, but it has also asked Korea, Panama, and administration announced that it would make important Colombia to make additional changes in order to win con- changes in trade policy making to advance “the social gressional approval. As of March 2011, the Korean govern- accountability and political transparency of trade policy� ment had agreed to several changes, but the Obama (USTR 2009). On July 16, 2009, U.S. Trade Representative administration continues to work with Panamanian and Ron Kirk made it clear that the U.S. government was less Colombian officials.62 With two wars and a burgeoning interested in delineating labor rights provisions or provid- budget deficit, these agreements do not seem to be a top ing incentives for international labor rights through its priority for either the administration or the Republican- trade agreements than in establishing “a level playing field dominated house. Congress sets the objectives for trade pol- for American workers.� He added that the “USTR will, icy making, but it has not made the advancement of human proactively monitor and identify labor violations and rights through trade agreements a top priority. In the Trade enforce labor provisions . . . When efforts to resolve viola- Promotion Act of 2002 (hereafter, TPA), which grants the tions have been expired, USTR will not hesitate . . . to president “fast-track� authority to negotiate trade agree- invoke formal dispute settlement.�65 However, as of March ments, the words “human rights� never appear.63 Congress 2011, the United States has filed only one such case, against has directed negotiators to focus on the following specific Guatemala, under the CAFTA-DR agreement, for alleged human rights: labor rights, access to information, public violations of obligations on labor rights.66 Nonetheless, the participation, and due process. In recent years, negotiators United States has signaled that labor rights violations have also been directed to balance the rights of U.S. intellec- under FTAs will be investigated, and if violations are not tual property rights (IPR) holders with U.S. obligations in remedied, the United States could bring the issue to a trade the Doha Declaration on Public Health. Almost all recent dispute. This new policy has clearly elevated labor rights. PTAs contain language related to these human rights. Public participation, transparency, and due process. As Negative spillovers of the focus on labor rights rather than noted above, the U.S. Congress requires its trade agreement on broader human rights. Activists and legislators have partners to agree to PTA provisions related to trans- focused most of their attention on provisions addressing parency, due process, and political participation.67 These labor rights (see Elliott, ch. 20 in this volume). This is provisions are embedded in the transparency, anticorrup- partly attributable to the clout of trade unions, but also to tion, and regulatory practices sections of the TPA. The the failure of many U.S. trade partners to enforce the labor Congress declared that the United States aims to “obtain laws on their books. These partners’ difficulties regarding wider and broader application of the principles of trans- labor rights may be a symptom of a larger governance parency� through “increased transparency and opportunity problem, and the focus on labor rights may make it more for the participation of affected parties in the development difficult to use the agreements to promote good gover- of regulations.� The legislation also states that trade nance and human rights. The dilemma is most visible in negotiators should “establish consultative mechanisms the case of Colombia. among parties to trade agreements to promote increased Human Rights 457 transparency in developing guidelines, rules, regulations incentive for democracy in the Middle East and might also and laws.�68 Finally, the act notes that to avoid corruption, cement democracy in Latin American nations such as the United States intends “to obtain high standards and Colombia and El Salvador that had experienced conflict.73 appropriate domestic enforcement mechanisms applicable These officials recognized that democracy could not simply to persons from all countries participating in the applica- be exported; they understood that some countries need ble trade agreement.�69 help in strengthening democratic institutions, processes, The United States has been promoting transparency and accountability.74 They hoped that if the United States and due process rights in multilateral trade agreements helped such countries to become more democratic and since the end of World War II (Aaronson and Zimmerman accountable to their publics, they (and the United States) 2007). U.S. policy makers have long argued that “trans- would gain greater legitimacy, and over time these allies parency is the starting point for ensuring the efficiency, would become more stable. They also believed that trade and ultimately the stability of a rules-based environment agreements with the United States would gain public sup- for goods crossing the border.�70 All U.S. PTAs approved port if citizens in partner countries could comment on since 2002 contain a chapter on transparency, and they various versions of the agreement and in so doing shape thus go beyond the WTO rules—they are “WTO+.� (There key public policies. are also transparency provisions in other chapters of the Under pressure from several members of Congress, PTA.) Although the language in these chapters varies from trade policy makers developed three models for public agreement to agreement, in general, the passages are participation, which were incorporated first in the envi- framed in the language of human rights. They require gov- ronmental chapters of PTAs and later in the labor chap- ernments to publish, in advance, laws, rules, procedures, ters, as well. The first model was for developed democratic and regulations affecting trade, thereby giving “persons of countries such as Australia, and it thus contained minimal the other party that are directly affected by an agency’s public participation provisions in the environmental process . . . a reasonable opportunity to present facts and chapter. The second model was designed for countries arguments in support of their positions prior to any final with relatively weak systems of environmental regulation administrative action.� These agreements also contain a sec- and accountability, or countries relatively new to democ- tion on review and appeal, designed to give the parties a rea- racy; it was used for Bahrain, Chile, Morocco, Oman, and sonable opportunity to support or defend their respective Singapore.75 Under this second model, the bilateral trade positions.71 Such provisions are not intended to promote partners set up an advisory committee, an Environmental human rights as such, but they may have human rights Affairs Council, that would meet regularly and engage the spillovers. public in discussion on the environment.76 Policy makers In contrast, U.S. efforts to foster public participation are also agreed to provide capacity-building assistance to sup- relatively new, dating from 1992. The United States and its port public participation (USTR 2005). partners first experimented with public participation provi- Senator Max Baucus, a Democrat from Montana, sions in the environmental side agreements to NAFTA. As played an important role in developing a third, more part of that agreement, Mexico, Canada, and the United extensive, approach. In 2004 he called on the USTR to put States set up a mechanism, the Citizen Submissions on the public participation provisions directly in the trade Enforcement Matters, to enable members of the public from agreement, to develop benchmarks and “ways to measure any of the three countries to submit a claim when a govern- progress over time,� and to find ways to encourage objec- ment is allegedly not enforcing its environmental laws.72 The tive monitoring and scrutiny by the public (Aaronson and commission investigates the allegation and issues nonbind- Zimmerman 2007, 174–76). The CAFTA–DR agreement ing resolutions. These investigations have occasionally led was the USTR’s first test of the third model. In February governments to change course. For example, the Mexican 2005 the United States and its six partners in CAFTA–DR government has become more responsive to the environ- agreed to establish a mechanism and secretariat allowing mental side effects from trade. Based on public pressure the general public to submit petitions regarding the opera- from citizen submissions, the Mexican government prom- tion of the agreement’s labor or environmental provisions. ised remediation in the case of toxic pollutants abandoned at If members of the public from any party to CAFTA–DR a lead smelter in Tijuana, and Mexico’s president declared believe that any party is not effectively enforcing its labor the Cozumel Coral Reef a protected area (Silvan 2004). or environmental laws, they can make a new submission to The George W. Bush administration (2001–09) decided this subbody, which reports to the Environmental Affairs to enhance public participation provisions. Administration Council. The agreement states that each party should review officials believed that trade agreements could act as an and respond to such communications in accordance with 458 Susan Ariel Aaronson its own domestic procedures.77 CAFTA–DR also authorizes U.S. policy makers are increasingly sensitive to public the tribunal “to accept and consider amicus curiae submis- concerns that there are costs to elevating IPR protection. sions from a person or entity that is not a disputing party.� Drugs and vaccines are increasingly expensive, and the To develop a workable system, the United States agreed to policy does not encourage the development of generic fund the first year of the secretariat’s work.78 The USTR has brands. Moreover, some argue that the policy does not replicated this model in more recent PTAs, such as those with adequately protect indigenous knowledge—knowledge Colombia, Korea, Panama, and Peru.79 The labor chapter of passed down through familiar and cultural ties, but not these PTAs, Article 19.4, notes, “Each party shall ensure that protected under domestic law. In the 2002 TPA, Congress persons with a recognized interest under its law in a particu- required the Bush administration to rethink its approach, lar matter have appropriate access to tribunals for the and “secure fair, equitable, and nondiscriminatory market enforcement of the Party’s labor laws.� In addition, “Each access opportunities for United States persons that rely party shall promote public awareness of its labor laws upon intellectual property protection; and to respect the including by (a) ensuring that information related to Declaration on the [Trade-Related Aspects of Intellectual its labor laws and enforcement and compliance procedures Property Rights (TRIPS)] Agreement and Public Health.�84 is publicly available; and (b) encouraging education of the The executive branch, however, was not given a mandate to public regarding its labor laws.�80 The United States and ensure access to affordable medicines. its trade partners have also established Labor Affairs Beginning in April 2004, all PTAs include a letter enti- Councils with other PTA partners.81 tled “Understandings Regarding Certain Public Health As readers might imagine, the placement of participa- Measures� that is signed by representatives of both govern- tion provisions in trade agreements cannot magically ments. The letter, found in the Bahrain, CAFTA–DR, stimulate democracy. Such an approach may not work in Colombia, Morocco, Oman, and Peru PTAs, says that the countries lacking a tradition of political participation or IPR provisions of the agreement “do not affect a Party’s free speech, and it may appear to violate another country’s ability to take necessary measures to protect public sovereignty or cultural mores. However, these provisions health.�85 But the letter does not make it clear that govern- might push partner governments to allow more public ments can breach IPR obligations in order to ensure that participation and could gradually teach citizens how to their citizens have access to affordable medicines. These engage and challenge policy makers. Since signing a free side letters do not assuage critics, including the World Bank trade agreement with the United States, PTA partners (Mercurio 2006, 234–35). Interestingly, this policy may not Chile, the Dominican Republic, Jordan, Kuwait, Mexico, change dramatically under the Obama administration. and Morocco have established channels through which There is no mention of public health in the President’s organized civil society can comment on trade policies Trade Policy Agenda, and the Trade Agreements Report, (Cherfane 2006: Aaronson and Zimmerman 2007). like others before it, simply stresses that strong intellectual Access to affordable medicines. Congress has long viewed property protection is essential for protecting public health protecting intellectual property rights (IPR) holders as a top (USTR 2011). priority. The United States is by far the most assertive coun- U.S. PTAs and human rights: A summary. The U.S. try in defending intellectual property rights within the con- approach to linking PTAs and human rights is contradic- text of trade policies and agreements, even when trade part- tory. On the one hand, it ignores the internationally ners and human rights activists argue that U.S. policies accepted notion that human rights are universal and indi- undermine the ability of policy makers or citizens to obtain visible, yet on the other hand, the United States works hard access to affordable medicine or to protect indigenous to promote specific human rights and now makes some of knowledge. U.S. assertiveness stems from a long-standing these rights binding. The United States is essentially saying belief among policy makers that the country’s economic to its partner nations: make the rights we value top priori- future is rooted in America’s global economic dominance of ties. In this way, the U.S. strategy toward linking human creative industries such as software, biotechnology, and rights is insensitive toward other cultures, which may have entertainment.82 These intellectual property–based indus- different human rights priorities. This strategy may inspire tries represent the largest single sector of the U.S. economy.83 U.S. trade partners to do more to advance some human To protect that future, U.S. policy makers work with their rights, but it is unlikely to inspire these governments to overseas counterparts to enforce intellectual property rights, devote more resources to human rights in general. seize counterfeit goods, pursue criminal enterprises involved The United States (like Canada) uses its trade agree- in piracy and counterfeiting, and “aggressively engage our ments to improve governance and, in so doing, to empower trading partners to join our efforts� (CEA 2005, 226). citizens to demand their rights. However, the United States Human Rights 459 uses the leverage of trade agreements to induce other cited numerous human rights motivations when it pre- nations to invest in the governance priorities valued by U.S. sented the protocol to the United Nations in 2000. For policy makers instead of in human rights as delineated in instance, when explaining the rationale for a democracy international law. clause, Brazilian policy makers argued that “democracy, development, and respect for human rights and fundamen- tal freedoms are interdependent and mutually reinforcing.�90 South-South Agreements Thus, although the Mercosur trade agreement contains no Although developed countries are the main demandeurs of binding human rights commitments, Brazil and its Merco- human rights provisions in PTAs, a number of emerging sur partners see the protection of human rights and democ- economies include such links in their trade agreements. racy as a rationale for, and a side effect of, the agreement Policy makers have incorporated language dealing with (Anderson and Zimmerman 2007, 106). Nevertheless, Mer- access to affordable medicines, indigenous or minority cosur refused to investigate allegations of human rights rights, due process, cultural rights, and labor rights. These abuses by member state Argentina (Garcia 2003). provisions are usually located in the preamble of the agree- Chile has incorporated labor rights language into sev- ment, with a few in the body of the text. eral of its more recent bilateral PTAs. For example, as noted Some developing countries link trade and human rights by Elliott (ch. 20 in this volume), the trade agreement by building on the exceptions in GATT/WTO Article XX between Chile and China includes two labor and environ- described earlier. For example, Article 8 of the agreement mental side agreements. Article 108 commits both parties between Egypt and Jordan allows measures “for religious, “to carry out cooperation activities in the fields of employ- hygienic, security or environmental reasons as long as they ment and labor policies and social dialogue.� Although are in conformity with the applicable laws and regulations in Chile’s language is aspirational, this language does under- both countries.� (This is the only provision that we have score the importance of labor rights.91 In addition, Chile’s found that associates trade and freedom of religion.) Other agreements with Costa Rica, Colombia, and Korea, as well countries target different human rights. For example, the as the CARICOM–Costa Rica PTA, refer to social protec- Caribbean Community (CARICOM) allows for the free tion (Bartels 2009a). movement of university graduates and those in listed occu- pations. Some countries have more restricted provisions Summary regarding free movement of people. (See Stephenson and Hufbauer, ch. 13 in this volume, on labor mobility.) For These South-South trade and human rights provisions, as example, NAFTA and the Canada–Chile and U.S.–Chile well as those in the U.S., European, and other agreements, PTAs contain chapters on temporary entry of business reflect a changing attitude about what trade agreements persons.86 Article 6 of the treaty establishing the Common should include and what they are about. According to Market of Eastern and Southern Africa (COMESA) commits Bartels, “the idea that these links are valid seems to be members to promote democracy and human rights.87 Several gaining its own dynamic� (Bartels 2009a, 365). Moreover human rights are embedded in the Mercosur agreement although many of these provisions are aspirational and between Argentina, Brazil, Paraguay, and Uruguay.88 If any not binding, such provisions may not remain rhetorical. party to Mercosur fails to protect ILO core labor standards, a Bartels concludes that as the number of agreements con- supranational Commission on Social and Labor Matters can taining preambular references to human rights grows in review allegations at the behest of another member state, number and scope, “this may well lead to operative provi- although it cannot impose trade sanctions or other penalties sions at a later stage,� as happened with EU human rights in the event of such a violation.89 provisions (365). Like the EU, Mercosur is built on the recognition that if the members want to jointly expand trade, they must col- The Effects of the Marriage of PTAs laborate on a wide range of issues. Thus, it is more of a gov- and Human Rights ernance agreement than simply a trade agreement. In 1998 Mercosur members adopted the Ushuaia Protocol on demo- Although the wedding of trade and human rights is new, this cratic commitment, which prohibits the entry of undemoc- marriage is leading to important changes in the way policy ratic states into the common market. Although the protocol makers make trade policy. The sheer number and dispersion text itself makes no explicit mention of human rights, Mer- of PTAs with human rights links seem to be changing some cosur members invoked the protocol as a joint response to a policy makers’ opinions about human rights. These pro- 1996 coup d’état in Paraguay, and the Brazilian delegation visions have thus contributed to the recognition and 460 Susan Ariel Aaronson internationalization of human rights norms. Second, Thus, at their meeting in August 2009, the leaders of the these provisions are changing the behavior of government three NAFTA countries discussed public health, border officials in a wide variety of countries, as described below. controls, and public safety—all issues relating to human Mauritania. The European Commission used consulta- rights.94 tions to warn the new government of Mauritania that the government’s behavior breached that country’s commit- Linkages and Knowledge Gaps: Some ments under the Cotonou Agreement. Soon thereafter, the Problems Presented by Current PTAs new Mauritanian regime pledged to hold free and fair elec- tions and initiated a process to establish an independent The many variants of PTAs raise questions about the link- National Commission for Human Rights. In 2007 the gov- age of trade and human rights and what it means for both ernment held free and fair elections. While one can’t say the international human rights regime and the status of that the PTA’s human rights provisions pushed the military human rights. government to protect human rights, political scientist Hafner-Burton concludes that these provisions gave the 1. Countries Have a Wide Range of Views about Using Trade to EU leverage over the government’s progress toward Advance Human Rights Abroad. reforms (Hafner-Burton 2009, 151–60). The United States, the EU, and Canada are quite comfort- Thailand. In 2003 Thailand entered into PTA negotiations able using their trade agreements to advance human rights with the United States. In 2005 the Thailand National in other countries. While some developing and middle- Human Rights Commission drafted a report on the human income countries have included some human rights provi- rights implications of the proposed trade agreement, express- sions in their PTAs, policy makers in these countries may ing concern about traditional knowledge and intellectual not believe it is appropriate to intervene in the affairs of property rights. The report concluded that Thailand would other nations, even in the most extreme cases of human have to adopt U.S. laws if it agreed to the trade agreement.92 rights abuse. The Thai public began to oppose the agreement. The Thai prime minister pledged to ensure greater public involve- 2. Countries Have a Wide Range of Views as to Whether They ment in the negotiating process to shape the agreement.93 Should Accept Human Rights Provisions in Trade Agreements. However, the negotiations ended after a military coup. Some nations, such as Australia, have actually refused to El Salvador and Guatemala. The USTR asserted that El negotiate trade agreements with human rights provi- Salvador and Guatemala held their first public hearings sions. Yet China has accepted human rights provisions in on trade as they negotiated CAFTA. These governments its PTAs with Chile and New Zealand. How do we continue to engage their citizens in trade. during the explain these differences? Some countries are comfort- negotiations of CAFTA. With such hearings, individuals able using their economic power to promote political and NGOs learn how to influence trade policy. Trade change; some countries are neutral and noninterven- policy makers in these countries may have acted under tionist; and others may be using their acceptance of these U.S. pressure, but they may also have recognized that provisions to signal investors and funders. Perhaps they must involve their publics if they don’t want to China sees adopting these provisions as a way to signal engender significant opposition to such agreements foreign investors that it is evenhanded and is attempting (Aaronson and Zimmerman 2007, 176). As noted earlier, to promote the rule of law. other nations with little tradition of civil society or even business involvement in trade, such as Bahrain, Jordan, 3. Incentives Work Better Than Disincentives to Change and Oman, also created public advisory bodies (Cher- Behavior. fane 2006; Lombardt 2008). Some countries have decided to use disincentives as a Mexico. Since joining NAFTA, Mexican trade policy has means of advancing the human rights embedded in par- become more responsive to public concerns. For example, ticular trade agreements. However, sanctions or fines can the Mexican government revamped its agricultural poli- do little to build demand for human rights or to train gov- cies. It has also begun to work internationally to protect its ernments or factory managers in how to respect human citizens’ labor rights. For example, in September 2009, rights. Isolating a government or punishing it will do little Mexican consulates attempted to educate Mexican guest to increase the targeted country’s commitment to human workers in the United States regarding their labor rights. rights over time. Other countries rely on dialogue to prod NAFTA leaders meet regularly, and human rights and the changes, but dialogue may do little to encourage a country rule of law have become important parts of their discussions. to change its behavior. Still others rely on incentives. We Human Rights 461 need to understand whether these incentives are really Notes effective and, if so, when these incentives should be 1. The United States has included human rights language in all its offered. PTAs (17 in force and 3 pending), except that with Israel, as well as in its generalized system of preferences (GSP) program with 131 countries. The EU has PTAs with 14 countries, and 13 of these agreements include 4. There Is a Lot Scholars Don’t Know about This Relationship. human rights provisions; about 120 countries are subject to human rights Scholars are just beginning to examine the relationship provisions in the EU’s GSP program. between human rights and trade performance.95 Aaronson 2. United Nations High Commissioner for Human Rights website, http://www.unhchr.ch/Huridocda/Huridoca.nsf/0/c462b62cf8a07b13c12 and Zimmerman did a simple correlation and found that 569700046704e?Opendocument. states that protect human rights signal investors that they 3. See the databases at the World Trade Law website, http://www.world are evenhanded and promote the rule of law (Aaronson tradelaw.net/fta/ftadatabase/PTAs.asp; WTO, Regional Trade Agreements and Zimmerman 2007, 193–95). Scholars also don’t know Information Service, http://rtais.wto.org/UI/PublicMaintainRTAHome .aspx; and, for the latest news, Bilaterals.org, http://www.bilaterals.org. if human rights provisions in trade agreements lead to 4. The two covenants and the UDHR together form the International greater trade distortions. Bill of Rights. 5. United Nations High Commissioner for Human Rights, “Fact Sheet,� http://www.unhchr.ch/html/menu6/2/fs2.htm (accessed February 5. These Provisions Have Costs as Well as Benefits. 12, 2006). The world and its people benefit when more governments 6. Ibid. protect, respect, and realize human rights. Yet some human 7. For example, if a government ignores its own labor laws, it is effec- tively allowing its labor-intensive firms to become more cost-competitive rights provisions are expensive for developing countries to with imports (see GATT 1989; Bagwell and Staiger 1998; Brown 2001, implement. These governments often have few resources, 29–31). and yet under many recent trade agreements, they must 8. Some WTO agreements require governments to accord due process choose to protect intellectual property, provide access to rights (such as the right to recognition before the law) to importers as well as exporters. For example, under the WTO’s Safeguards Agreement, when affordable medicines, and/or invest in education. Trade workers or industries petition their government for import relief, the agreements may prod policy makers to make the human responding government must give public notice and hold hearings in which rights priorities of their trade partners their human rights interested parties can respond to a safeguard investigation (Interpretation and Application of Article 1 of the WTO Agreement on Safeguards, priorities. We don’t know if this strategy ultimately http://www.wto.org/english/res_e/booksp_e/analytic_index_e/safeguards increases the demand for and supply of human rights. _02_e.htm#). The Agreement on Safeguards envisages that the interested To gain better understanding of the costs and benefits of parties will play a central role in the investigation and that they will be a primary source of information for the competent authorities. The Agree- the association of trade and human rights, scholars, policy ment on Technical Barriers to Trade requires governments to publish stan- makers, and activists could use qualitative studies, empiri- dards and technical regulations and allow interested parties (whether cal studies, or human rights impact assessments. Scholars foreign or domestic) to become acquainted and respond to the regulation have several global datasets they can use to do empirical (http://www.wto.org/english/res_e/booksp_e/analytic_index_e/tbt_01_e. htm#p). The Customs Valuation Agreement requires governments to research (for example, the Cingranelli-Richards [CIRI] establish in law the right of the importer to appeal a determination of Human Rights Dataset or the now open World Bank customs value. Appeal may first be to a higher level in the customs datasets).96 Human rights impact assessments are rela- administration, but the importer shall have the right in the final instance to appeal to the judiciary (Text of Interpretive Note to Article XI, http:// tively new; they are designed to measure the potential www.wto.org/english/res_e/booksp_e/analytic_index_e/cusval_02_e.htm impact of a trade agreement on internationally accepted #article11A). human rights standards. Trade and human rights policy 9. Paragraph 2(a) General 319, GATT Analytical Index, http://www .wto.org/english/res_e/booksp_e/analytic_index_e/gatt1994_04_e.htm# makers should collaborate with scholars, NGOs, and oth- articleXA. ers to develop a clear and consistent methodology for 10. Powell (2005) notes, as an example, that provisions to implement evaluating such impact (3D 2009; Walker 2009). transparency are embedded in NAFTA Articles 510, 909, 1036, and 1411. 11. See World Trade Law database, http://www.worldtradelaw.net/fta/ ftadatabase/PTAs.asp; WTO Regional Trade Agreements Information Sys- Conclusions tem database, http://rtais.wto.org. 12. See discussion in House of Commons Committee, “Government More countries are marrying trade and human rights. If Response to the Seventh Report of the Standing Committee on Interna- tional Trade,� March 2007, 39th Parliament, 1st Session, 1–3. this marriage is to endure, we need greater understanding 13. “Speaking Notes for Honorable Jean-Pierre Blackburn, Minister as to whether this union is effective and whether it can and of Labor and Minister of the Economic Development Agency of Canada should endure. We can begin by doing a comprehensive at Montreal Council on Foreign Relations, Montreal, Quebec, February 29, 2008,� http://www.hrsdc.gc.ca/eng/corporate/newsroom/speeches/ study of which PTA strategies encourage policy makers to blackburnjp/2008/080229.shtml. do a better job of advancing human rights and if particular 14. Foreign Affairs and International Trade Canada (DFAIT), “Negoti- trade agreements help people realize their human rights ations and Agreements,� http://www.international.gc.ca/trade-agreements- under law. accords-commerciaux/agr-acc/index.aspx. Canada’s four most recent PTAs 462 Susan Ariel Aaronson are with Peru (in force as of August 1, 2009); Jordan (signed June 28, 2009, 32. See, for example, EFTA–Colombia, http://www.efta.int/free-trade/ but not yet approved by Parliament); Colombia (signed November 28, free-trade-agreements/colombia/fta-en.aspx; for Chile, http://www.efta 2008, but not yet approved); and EFTA (in force as of January 28, 2009). .int/free-trade/free-trade-agreements/chile/fta.aspx, Article XXI. 15. DFAIT, “Negotiations and Agreements.� 33. World Trade Law, “Free Trade Agreement between the EFTA 16. United Nations Association in Canada (UNAC), “Canada and States and the SACU States,� http://www.worldtradelaw.net/fta/agree Human Rights,� http://www.unac.org/rights/actguide/canada.html. ments/SACU_EFTA_FTA.pdf. 17. DFAIT, “Canada’s International Human Rights Policy,� 34. Ibid., Article 30. http://www.international.gc.ca/rights-droits/policy-politique.aspx. 35. Committee of Members of Parliament of the EFTA Countries, 18. The Peru, Colombia, and Jordan agreements share the same basic “Environmental Policies and Labor Standards in PTAs,� August 24, 2009, preamble. Brussels. 19. DFAIT, “Canada-Peru Free Trade Agreement,� http://www.inter 36. Committee of Members of Parliament of the EFTA Countries, national.gc.ca/trade-agreements-accords-commerciaux/agr-acc/andean- “Environmental Policies and Labor Standards in PTAs,� Ref. 1090382, andin/can-peru-perou.aspx?lang=en; DFAIT, “Canada-Jordan Free Trade March 18, 2009, Brussels, 3–4. Agreement,� http://www .international.gc.ca/trade-agreements-accords- 37. EFTA, http://www.efta.int/eea/eea-agreement.aspx, 5–6. commerciaux/agr-acc/jordan-jordanie/chapter15-chapitre15.aspx?lang= 38. European Commission, Employment, Social Affairs, and Inclu- eng. sion, “Free Movement of Workers,� http://ec.europa.eu/social/main.jsp? 20. Canada’s approach to labor rights is discussed by Elliott, in chap- catId=458&langId=en; “Equal Treatment,� http://ec.europa.eu/social/ ter 20 of this volume; the present chapter focuses on how the labor rights main.jsp?catId=462&langId=en. side agreement addresses other important human rights. Canadian offi- 39. The European Community was founded as the European Eco- cials state that labor and environmental provisions need to be embedded nomic Community on March 25, 1957; in 1993 it was renamed the in side agreements rather than in chapters because the provinces, which European Union. regulate labor, could challenge the right of the federal government to 40. European Commission, Trade, “FTA Negotiations,� http://trade.ec force them to adhere to a treaty in their areas of jurisdiction. The side .europa.eu/doclib/docs/2006/december/tradoc_118238.pdf; Aaronson and agreements allow for voluntary adherence by the provinces. This Zimmerman (2007, 139). arrangement respects provincial jurisdiction on labor matters but gives 41. European Union, “Partnership and Cooperation Agreements Canada the ability to immediately access the dispute resolution process, (PCAs),� http://europa.eu/legislation_summaries/external_relations/rela regardless of the level of provincial participation in the labor coopera- tions_with_third_countries/eastern_europe_and_central_asia/r17002_en tion agreement. .htm. 21. Human Resources and Skill Development Canada, “International 42. Bartels and other analysts have noted that Australia refused to Labour Affairs,� http://www.hrsdc.gc.ca/eng/lp/ila/index.shtml/. accept a human rights clause in its trade agreement, and thus, the EU and 22. See the 32-page “Agreement on Labour Cooperation� between Australia were unable to negotiate a PTA. Canada and Colombia, http://www.hrsdc.gc.ca/eng/labour/labour_agree 43. European Union, “EC Regional Trade Agreements,� http:// ments/ccalc/Canada-Colombia_LCA.pdf. trade.ec.europa.eu/doclib/docs/2006/december/tradoc_111588.pdf. 23. DFAIT, “Canada-Peru Free Trade Agreement, Chapter 19, Trans- 44. European Commission, “The Cotonou Agreement,� http://ec parency,� http://www.international.gc.ca/trade-agreements-accords-comm .europa.eu/development/geographical/cotonouintro_en.cfm; European erciaux/agr-acc/peru-perou/peru-toc-perou-tdm.aspx?lang=eng. Commission, Trade, “Economic Partnerships,� http://ec.europa.eu/trade/ 24. DFAIT, “Canada-Colombia Free Trade Agreement, Chapter 19, wider-agenda/development/economic-partnerships/. Transparency,� http://www.international.gc.ca/trade-agreements-accords- 45. EU, “Economic Partnerships.� commerciaux/assets/pdfs/EN%2019%20Colombia%20FTA%20-%20Trans 46. European Commission, “An Overview of the Interim Agree- parency.pdf. ments,� http://trade.ec.europa.eu/doclib/docs/2009/january/tradoc_142188 25. Media Awareness Network, “Canada’s Cultural Policies,� .pdf; and, according to http://www.bilaterals.org, as of 2009: http://www.media-awareness.ca/english/issues/cultural_policies/canada_ cultural_policies.cfm; DFAIT, “Culture,� http://www.international.gc.ca/ trade-agreements-accords-commerciaux/fo/index.aspx?lang=en. EU–ACP 26. For Colombia, http://www.ec.gc.ca/caraib-carib/default.asp? subgroup Status of agreement lang=En&n=F24F07DD-1; for Peru, http://www.ec.gc.ca/caraib-carib/ Caribbean • Full EPA initialed in December 2007 and signed in default.asp?lang=En&xml=5CBDDB77-6054-4FD3-A348-125840CEDF29. October 2008 (but not by Haiti) and approved by the 27. Human Rights Impact Resource Center, http://www.humanright- European Parliament (March 2009). simpact.org/news/newsitem/article/roundtable-on-human-rights-impact- Central • Interim EPA initialed (December 2007) and assessments-in-canada-on-16-november-178/?tx_ttnews[backPid]=769& Africa signed by Cameroon only (January 2009). cHash=0a3f7f441d60fbea25cc641a768a1a1c; Canadian Labour Congress, • Seven countries have not initialed anything yet. http://www.canadianlabour.ca/news-room/publications/canada-colombia- West Africa • Interim EPA initialed (December 2007) and signed by free-trade-agreement-round-two-0. Côte d’Ivoire (November 2008) and approved by the 28. Hearing on Canada–South America Trade Relations, 40th Parlia- European Parliament (March 2009). ment, 2nd session, Standing Committee on International Trade, June 16, 2009, http://www2.parl.gc.ca/HousePublications/Publication.aspx?DocId= • Interim EPA initialed by Ghana (December 2007). 4000400&Language=E&Mode=1&Parl=40&Ses=2. • Fourteen countries have not initialed anything yet. 29. International Centre for Trade and Sustainable Development East Africa • Interim EPA initialed by Zimbabwe, the Seychelles, (ICTSD), “Canada–Colombia FTA Gets Human Rights Amendment,� Mauritius, the Comoros, Madagascar, and Zambia http://ictsd.org/i/news/bridgesweekly/73372/. (November–December 2007) and signed by 30. DFAIT, “Canada–European Union Joint Report: Towards a Compre- Zimbabwe, the Seychelles, Mauritius, and Madagascar hensive Economic Agreement,� http://www.international.gc.ca/trade-agree only (August 2009). ments-accords-commerciaux/agr-acc/eu-ue/can-eu-report-can-ue-rapport. • Interim EPA initialed by East African Community aspx?lang=eng. members Burundi, Kenya, Rwanda, Tanzania, and 31. EFTA, “Free Trade Agreements,� http://www.efta.int/free-trade/ Uganda (November 2007). free-trade-agreements.aspx. • Five countries have not initialed anything yet. Human Rights 463 Southern • Interim EPA initialed by Botswana, Lesotho, 64. “The President’s 2011 Trade Policy Agenda,� http://www.ustr.gov/ Africa Namibia, Swaziland, and Mozambique webfm_send/2587. (November–December 2007) and signed by 65. USTR, “Trade Policy: A Level Playing Field for America’s Work- Botswana, Lesotho, Swaziland, and Mozambique only ers,� July 16, 2009, http://www.ustr.gov/about-us/press-office/fact-sheets/ (June 2009). 2009/july/trade-policy-level-playing-field-american-workers. • Angola has not initialed anything yet. 66. USTR, “Guatemala Submission under CAFTA-DR,� http:// Pacific • Interim EPA initialed by Papua New Guinea and Fiji www.ustr.gov/trade-topics/labor/bilateral-and-regional-trade-agreements/ (November 2007) and signed by Papua New Guinea guatemala-submission-under-cafta-dr; USTR, “USTR Kirk Addresses only (July 2009). Workers in Pennsylvania on Obama Administration Enforcement of U.S. • Thirteen countries have not initialed anything yet. Trade Agreements,� http://www.ustr.gov/node/6046, 4; USTR, “The 2010 National Trade Estimate Report: Key Elements,� http://www.ustr .gov/ about-us/press-office/fact-sheets/2010/march/-2010-national-trade-estimate- 47. TradeMark Southern Africa, http://www.trademarksa.org/node/ report-key-elements. 485; http://ictsd.org/i/news/tni/57509/; Forbes.com, “Africa–E.U. Eco- 67. Transparent, accountable governance can foster democracy, capi- nomic Agreement Stalls,� http://www.forbes.com/2010/08/11/eu-africa- talism, and political stability. Thus, by promoting transparency, the rule of epa-business-oxford-analytica.html. law, and political participation, policy makers can promote many human 48. European Commission, Official Journal of the European Com- rights. See UNDP (2002), 2–4. munity, “Economic Partnership Agreement between the CARIFORUM 68. Bipartisan Trade Promotion Authority Act, Section 2102 States, of the One Part, and the European Community and Its Member (b) (5), (8). States, of the Other Part,� http://trade.ec.europa.eu/doclib/docs/2008/ 69. Ibid., Section 2102 (b) 6 (A). february/tradoc_137971.pdf. 70. U.S. communication on Article X of GATT 1994, G/C/W/384, 49. The EPAs can be examined at http://www.bilaterals.org/spip.php? June 7, 2002; Robert B. Zoellick, “Free Trade and the Hemispheric Hope,� rubrique17. Council of the Americas meeting, May 7, 2001, http://ctrc.sice.oas 50. Ibid. Canada has similar provisions. .org/geograph/westernh/zoellick_3.pdf. 51. European Commission, Trade, “European Competitiveness,� 71. For example, Chapter 20 of the Australia–U.S. FTA, http://ec.europa.eu/trade/creating-opportunities/trade-topics/european- http://www.dfat.gov.au/trade/negotiations/us_fta/final-text/chapter_20 competitiveness/index_en.htm; European Commission, “Global Europe: .html; Chapter 17 of the U.S.–Bahrain FTA; Section 19 of the Peru FTA, Competing in the World,� http://ec.europa.eu/trade/creating-oppor www.ustr.gov/Trade_Agreements/Bilateral. tunities/trade-topics/european-competitiveness/global-europe/. 72. Commission for Economic Cooperation of North America, “A 52. European Commission, Trade, “Overview of FTA and Other Guide to Articles 14 and 15,� http://www.cec.org/Page.asp?PageID=122& Trade Negotiations,� updated March 3, 2011, http://trade.ec.europa.eu/ ContentID=1388&SiteNodeID=210. doclib/docs/2006/december/tradoc_118238.pdf. 73. According to President Bush, “open trade . . . spurs the process of 53. European Commission, Trade, “Overview of FTA and Other economic and legal reform. And open trade reinforces the habits of liberty Trade Negotiations.� that sustain democracy over the long term.� Quoted in testimony by 54. European Commission, “Free Trade Agreement between the USTR Robert Zoellick to the Senate Finance Committee, “America’s Trade European Community and Its Member States, of the One Part, and the Policy Agenda,� March 5, 2003. Republic of Korea, of the Other Part,� http://trade.ec.europa.eu/ 74. Robert B. Zoellick, “Global Trade and the Middle East: Reawaken- doclib/docs/2009/october/tradoc_145139.pdf. ing a Vibrant Past,� World Economic Forum, Jordan, June 23, 2003; state- 55. On transparency, European Commission, Trade, http://trade.ec ment by Peter F. Algeier, Acting U.S. Trade Representative, before the .europa.eu/doclib/docs/2009/october/tradoc_145184.pdf; on the cultural Senate Finance Committee, April 13, 2005, 6–8. On Central America, see protocol, http://trade.ec.europa.eu/doclib/docs/2009/october/tradoc_ Robert B. Zoellick, “A Free Trade Boost for Our Hemisphere,� Wall Street 145194.pdf. Journal, November 17, 2003; Zoellick, “The Route From Miami to Eco- 56. Much of the text that follows is from Aaronson and Zimmerman nomic Freedom,� Financial Times, December 9, 2003. (2007, 139–47). 75. Interview with Mark Linscott, Assistant USTR for the Environ- 57. European Commission, Trade, “Sustainability Impact Assessments,� ment, and Jennifer Prescott, Deputy Assistant USTR for the Environment, http://ec.europa.eu/trade/analysis/sustainability-impact-assessments/. August 1, 2006. 58. Human Rights Watch, Christian Solidarity Worldwide, and Inter- 76. See Article 19.3 and 19.4 of the U.S.–Chile FTA, http://www national Crisis Group, “No Trade Agreement for Turkmenistan-Joint Let- .ustr.gov/Trade_Agreements/Bilateral/Chile_FTA/Final_Texts/Section_ ter,� March 20, 2006, http://hrw.org/english/docs/2006/03/20/turkme1 Index.html. 3040.htm (accessed April 24, 2006). 77. Part of the agreement’s work program is to build capacity to pro- 59. EU, “Human Rights and Democracy in the World: Report on EU mote public participation in environmental decision making. The agree- Action July 2008 to December 2009,� 2010, http://www.eeas.europa ment was negotiated by the Department of State; see Environmental .eu/_human_rights/docs/2010_hr_report_en.pdf. Cooperation Agreement, February 1, 2005, http://www.state.gov/g/oes/ 60. European Commission, Communication from the European env/trade/caftacooperation/142688.htm. Also see “U.S., Central America, Commission to the Council and the European Parliament, “The European Dominican Republic Sign Environment Pacts,� U.S. Department of State, Union’s Role in Promoting Human Rights and Democratisation in Third http://www.america.gov/st/washfile-english/2005/February/20050218133 Countries,� COM(2001)252 final, May 8, 2001, Brussels. 452GLnesnoM0.3828546.html. The Environmental Affairs Council met 61. For two views of “competitive liberalization,� see Evenett and for the first time on May 24, 2006. Meier (2008) and Aaronson (2002). 78. On dispute settlement, see CAFTA-DR, Article 10.20. 3; 62. USTR (2011); for Colombia, http://www.ustr.gov/trade-agree- “Communiqué of the Environmental Affairs Council of the Dominican ments/free-trade-agreements/colombia-fta; for Korea, http://www.ustr Republic–Central America–United States Free Trade Agreement,� May 24, .gov/trade-agreements/free-trade-agreements/korus-fta; for Panama, http:// 2006, http://www.worldtradelaw.net/fta/agreements/CAFTADR_RelDoc_ www.ustr.gov/trade-agreements/free-trade-agreements/panama-tpa. CommEnvAff.pdf . 63. Public Law 107-210, August 6, 2002, Section 2102 (a) Trade Nego- 79. Interview with Mark Linscott and Jennifer Prescott, Office of tiating Objectives of the Bipartisan Trade Promotion Authority Act of the USTR, August 2, 2006. For the Peru agreement, see http://www.ustr 2002; USTR 2009. .gov/trade_agreements/bilateral/; for the draft Colombia FTA, see 464 Susan Ariel Aaronson http://www.ustr.gov/Trade_Agreements/Bilateral/Colombia_FTA/Draft ———. 2007. “Seeping in Slowly: How Human Rights Concerns Are Pen- _Text/Section_Index.html. etrating the WTO.� World Trade Review 6 (3): 413–49. 80. Chapter 19, Labor, in the Korea, Panama, Peru, and Colombia Aaronson, Susan Ariel, and Rod Abouharb. Forthcoming. “Strange Bed- FTAs. See, for example, Chapter 19 in the U.S.–Korea FTA, http://www fellows: The WTO and Democratic Rights.� International Studies .ustr.gov/sites/default/files/uploads/agreements/fta/korus/asset_upload_ Quarterly. file934_12718.pdf; Article 20, http://www.ustr.gov/sites/default/files/ Aaronson, Susan Ariel, and Jamie M. Zimmerman. 2007. Trade Imbalance: uploads/agreements/fta/korus/asset_upload_file852_12719.pdf. The Struggle to Weigh Human Rights in Trade Policymaking. New York: 81. USTR (2011, 149–51). “As part of increased engagement in 2010, Cambridge University Press. Labor Affairs Council and labor subcommittee meetings, as provided for Bagwell, Kyle, and Robert W. Staiger. 1998. “The Simple Economics of under the FTAs, were held with several FTA partners, including first-time Labor Standards and the GATT.� NBER Working Paper 6604, National meetings with Peru, Morocco, and Bahrain. (For additional information, Bureau of Economic Research, Cambridge, MA. http://www.nber.org/ see Chapter III.A).� papers/w6604. 82. “House Chairmen Warn USTR against Patent Changes in Doha Bal, Salman. 2001. “International Free Trade Agreements and Human Round,� Inside U.S. Trade (April 14, 2006), 7. Rights: Reinterpreting Article XX of the GATT.� Minnesota Journal of 83. Press release, “Commerce Secretary Carlos Gutierrez Unveils Ini- International Trade 10 (62): 62–108. tiatives to Fight Intellectual Property Theft,� September 21, 2005, Bartels, Lorand. 2005a. “Human Rights and Democracy Clauses in the http://www.commerce.gov/opa/press/Secretary_Gutierrez/2005_Releases EU’s International Agreements.� Prepared for the Subcommittee on /September/09-21-05%20IPR%20initiatives.htm. Keyder (2005) has an Human Rights, Committee on Foreign Affairs, European Parliament, interesting scholarly take on this issue. EP/EXPOL/B/2005/, Brussels. 84. Trade Promotion Authority Act of 2002, Section 2102, (4) (A) (B) (C). ———. 2005b. Human Rights Conditionality in the EU’s International 85. See, for example, “Understandings Regarding Certain Public Agreements. Oxford, U.K.: Oxford University Press. Health Measures,� Colombia, http://www.ustr.gov/Trade_Agreements/ ———. 2008. “The Application of Human Rights Conditionality in the Bilateral/Colombia_FTA/Draft_Text/Section_Index.html (accessed August EU’s Bilateral Trade Agreements and Other Trade Arrangements with 7, 2006). Third Countries.� Report for the European Parliament, Brussels. 86. Basically, free movement of skills entails the right to seek employ- ———. 2009a. “Social Issues: Labour, Environment and Human Rights.� ment in any member state and the elimination of the need for work per- In Bilateral and Regional Trade Agreements: Commentary and Analysis, mits and permits of stay. See CARICOM, http://www.caricom.org/jsp/sin ed. Simon Lester and Bryan Mercurio, 340–66. Cambridge, U.K.: gle_market/skill.jsp?menu=csme. Cambridge University Press. 87. COMESA’s Integration Agenda also rests on the belief that the ———. 2009b. “Trade and Human Rights.� In Oxford Handbook of Inter- consolidation of democracy and the rule of law are the foundation stones national Trade Law, ed. D. Bethlehem et al., 571–96. Oxford, U.K.: for regional socioeconomic development and political stability. Oxford University Press. 88. Mercosur is incorporated into the Latin America Integration Bartels, Lorand, and Federico Ortino. 2006. Regional Trade Agreements Association/Asociación Latinoamericana de Integración (LAIA/ALADI) and the WTO Legal System. Oxford, U.K.: Oxford University Press. legal regime, as Economic Complementarity Agreement 18. These agree- Barton, John H. 2006. The Evolution of the Trade Regime: Politics, Law, and ments are open to accession by any LAIA/ALADI country. Economics of the GATT and the WTO. Princeton, NJ: Princeton Uni- 89. See “Treaty Establishing a Common Market,� March 26, 1991, UN versity Press. Doc. A/46/155 (1991), reprinted in 30 I.L.M. 1041 (1991). Bhagwati, Jagdish. 1996. “Introduction.� In Economic Analysis. Vol. 1 of 90. For the full text of the Brazilian delegation’s speech, see Fair Trade and Harmonization: Prerequisites for Free Trade? ed. Jagdish http://www.un.int/brazil/speech/00d-mercosul-human-rights-2610.htm. Bhagwati and Robert Hudec. Cambridge, MA: MIT Press. 91. Memorandum of Understanding, http://www.sice.oas.org/Trade/ Bidwell, Percy. 1939. The Invisible Tariff: A Study of the Control of Imports CHL_CHN/CHL_CHN_e/Labor_e.asp. into the United States. New York: Council on Foreign Relations. 92. The English version of the draft report is available at http://www Bolle, Mary Jane. 2009. “Proposed Colombia Free Trade Agreement: .measwatch.org/autopage/show_page.php?t=5&s_id=3&d_id=7. Labor Issues.� Report RL34759, Congressional Research Service, 93. MCOT News, “PM Calls for More Public Participation in Library of Congress, Washington, DC. http://88.80.13.160/leak/crs/ FTA Deals,� http://www.bilaterals.org/article.php3?id_article=953&var_ RL34759.pdf. recherche=public+information (accessed August 3, 2006). Meanwhile, the Bourgeois, Jacques, Kamala Dawar, and Simon J. Evenett. 2007. “A Compar- EPA is reviewing its public participation efforts. ative Analysis of Selected Provisions in Free Trade Agreements.� Pre- 94. USTR, “Joint Statement of the 2009 NAFTA Commission Meet- pared for the Directorate General for Trade, European Union, Brussels. ing,� http://www.ustr.gov/about-us/press-office/press-releases/2009/ Brown, Drusilla K. 2001. “Labor Standards: Where Do They Belong on the october/joint-statement-2009-nafta-commission-meeting. International Trade Agenda?� Working Paper, Department of Eco- 95. In a widely cited, albeit out-of-date, study, the Organisation for nomics, Tufts University, Medford, MA. http://ase.tufts.edu/econ/ Economic Co-operation and Development (OECD) examined the rela- papers/200113.pdf. tionship between freedom of association and trade and investment pat- Büthe, Tim, and Helen V. Milner. 2008. “The Politics of Foreign Direct terns and found that, the more successful is trade reform, the greater is Investment into Developing Countries: Increasing FDI through Inter- respect for association rights. The authors concluded that freer trade does national Trade Agreements?� American Journal of Political Science 52 not lead industrialized countries to lower their standards, and higher (4, October): 741–62. standards do not jeopardize trade reforms (OECD 1996, 9, 105). CCIC (Canadian Council for International Co-operation), Americas Pol- 96. http://ciri.binghamton.edu/; http://econ.worldbank.org/WBSITE/ icy Group. 2009. “What Role for Canada in the Americas?� CCIC, EXTERNAL/EXTDEC/EXTRESEARCH/0,,contentMDK:20388241~menuPK: Ottawa. http://www.ccic.ca/_files/en/working_groups/003_apg_2009- 665266~pagePK:64165401~piPK:64165026~theSitePK:469382,00.html. 04-15_statement_what_role_for_cda_in_americas.pdf. CEA (Council of Economic Advisors). 2005. Economic Report of the Presi- dent, 2005. Washington, DC: CEA. Selected Bibliography Charnovitz, Steve. 1987. “The Influence of International Labor Standards Aaronson, Susan Ariel. 2002. “George W. Bush: Protectionist.� Interna- on the World Trading Regime: A Historical Overview.� International tional Economy 16: 54–56. Labour Review 126 (5): 565–84. Human Rights 465 ———. 1994. “The World Trade Organization and Social Issues.� Journal NG13/W/31, GATT, Geneva. http://www.worldtradelaw.net/history/ of World Trade 28 (5, October): 17–33. urdsu/w31.pdf. ———. 1998. “The Moral Exception in Trade Policy.� Virginia Journal of Hafner-Burton, Emilie M. 2009. Forced to Be Good: Why Trade Agreements International Law 38 (4, Summer): 689–745. http://www.world- Boost Human Rights. Ithaca, NY: Cornell University Press. tradelaw.net/articles/charnovitzmoral.pdf. Hamilton, Carl B. 2002. “Globalization and Democracy.� CEPR Discus- ———. 2001. “The WTO and the Rights of the Individual.� Intereconom- sion Paper 3653, Centre for Economic Policy Research, London. ics 36 (2, March): 98–108. SSRN: http://ssrn.com/abstract=282021. Harrison, James. 2009. “Conducting a Human Rights Impact Assessment ———. 2005. “The Labor Dimension of the Emerging Free Trade Area of the Canada–Colombia Free Trade Agreement, Key Issues.� Americas of the Americas.� GWU Law School Public Law and Legal Theory Policy Group, Canadian Council for International Co-operation, Paper 140, George Washington University, Washington, DC. SSRN: Ottawa. http://www.ccic.ca/_files/en/working_groups/003_apg_2009- http://ssrn.com/abstract=705301. 02_hr_assess_of_cfta.pdf. Cherfane, Carol Couchani. 2006. “Trade and Environment Decision- Horn, Henrik, Petros C. Mavroidis, and André Sapir. 2010. “Beyond the Making in the Arab Region.� United National Economic and Social WTO? An Anatomy of EU and US Preferential Trade Agreements.� Commission for Western Asia, Muscat, Oman. http://css.escwa.org.lb/ World Economy 33 (11): 1565–88. 25mar06/14.pdf. Howse, Robert. 2002. “Human Rights in the WTO: Whose Rights, What Commission on Human Rights. 2004. “Economic, Social and Cultural Humanity? Comment on Petersmann.� European Journal of Interna- Rights: On Human Rights Resolution 2003/25.� E/CN.4/2004/10, tional Law 13 (3): 651–59. United Nations, New York. Keyder, Virginia Brown. 2005. “From Information to Property and Back Compa, Lance A., and Stephen F. Diamond, eds. 1996. Human Rights, Again.� Information and Communications Technology Law 14 (3, Labor Rights, and International Trade. Philadelphia, PA: University of October): 299–312. Pennsylvania Press. Lester, Simon, and Bryan Mercurio. 2009. Bilateral and Regional Trade Cottier, Thomas, Joost Pauwelyn, and Elizabeth Bürgi, eds. 2005. Human Agreements: Commentary and Analysis. Cambridge, U.K.: Cambridge Rights and International Trade. Oxford, U.K.: Oxford University Press. University Press. Cox, Phyllida, Peter Newell, Tracy Tuplin, and Diana Tussie. 2006. “Civil Lim, Hoe. 2001. “Trade and Human Rights: What’s at Issue?� E/C Society Participation in Trade Policy-Making in Latin America: 12/2001/WP.2, United Nations Economic and Social Council, New Reflections and Lessons.� IDS Working Paper 267, Institute of Devel- York. http://www.unhchr.ch/tbs/doc.nsf/0/907f88e4d28e4cb9c1256a opment Studies, University of Sussex, Brighton, U.K. http://www.iadb 63003069fd?Opendocument. .org/intal/intalcdi/PE/2009/02463.pdf. Lombardt, Ivin. 2008. “Role and Influence of Civil Society in Malawi’s Damro, Chad. 2006. “The Political Economy of Regional Trade Agreements.� Trade Policy Making Process: The Case of the Malawi Economic Justice In Regional Trade Agreements and the WTO Legal System, ed. Lorand Network.� One World Action, London. http://www.oneworldaction Bartels and Federico Ortino. Oxford, U.K.: Oxford University Press. .org/OneStopCMS/Core/CrawlerResourceServer.aspx?resource=2BAA Dawar, Kamala. 2008. “Assessing Labour and Environmental Regimes in 53AE-4BEF-44CD-9CD0-B5679F7411C4&mode=link&guid=bd9097b Regional Trading Arrangements.� University of Amsterdam Law 13c2f433e891e779111c14bc3. School, Amsterdam, Netherlands. Mansfield, Edward D., and Jon C. Pevehouse. 2008. “Democratization and Diebold, William. 1952. “The End of the I.T.O.� Essays in International the Varieties of International Organizations.� Journal of Conflict Reso- Finance 16: 1–37. lution 52 (2, April): 269–94. Dobbin, Frank, Beth Simmons, and Geoffrey Garrett. 2007. “The Global Marceau, Gabrielle. 2001. “WTO Dispute Settlement and Human Rights.� Diffusion of Public Policies: Social Construction, Coercion, Competi- Prepared for the joint meeting of the American Society of International tion, or Learning?� Annual Review of Sociology 33: 449–72. Law and the World Trade Forum on “International Economic Law and Dommen, Caroline. 2002. “Raising Human Rights Concerns in the World Human Rights.� Trade Organization: Actors, Processes and Possible Strategies.� Human Mehra, Malini, ed. 1999. Human Rights and Economic Globalisation: Rights Quarterly 24 (1, February): 1–50. Directions for the WTO. London: Global Publications Foundation and Drahos, Peter, and Ruth Mayne. 2002. Global Intellectual Property Rights: International NGO Committee on Human Rights in Trade and Knowledge, Access and Development. London: Macmillan. Investment. Dunoff, Jeffrey L. 1999. “Does Globalization Advance Human Rights?� Mercurio, Bryan. 2006. “TRIPS-Plus Provisions in RTAs: Recent Brooklyn Journal of International Law 25 (1). Trends.� In Regional Trade Agreements and the WTO Legal System. Evenett, Simon J., and Michael Meier. 2008. “An Interim Assessment of ed. Lorand Bartels and Federico Ortino. Oxford, U.K.: Oxford Uni- U.S. Trade Policy of Competitive Liberalization.� World Economy versity Press. 31 (1): 31–66, Morales-Moreno, Isidro. 2005. “Mexico’s Agricultural Trade Policies: Ferrantino, Michael. 2006. “Policy Anchors: Do Free Trade Agreements International Commitments and Domestic Pressure.� In Managing the and WTO Accessions Serve as Vehicles for Developing-Country Policy Challenges of WTO Participation: 45 Case Studies, ed. Peter Gallagher, Reform?� U.S. International Trade Commission, Washington, DC. Patrick Low, and Andrew L. Stoler. Cambridge, U.K.: Cambridge SSRN: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=895272. University Press. http://www.wto.org/english/res_e/booksp_e/case Florentine, Robert V., Luis Verdeja, and Christelle Toqueboeuf. 2007. “The studies_e/case28_e.htm. Changing Landscape of Regional Trade Agreements: 2006 Update.� Nickel, James. 2006. “Human Rights.� In Stanford Encyclopedia of Phi- WTO Discussion Paper 12, World Trade Organization, Geneva. losophy: Fall 2006 Edition, Stanford University, Stanford, CA. Garcia, Frank. 1999. “The Global Market and Human Rights: Trading http://plato.stanford.edu/archives/fall2006/entries/rights-human/. Away the Human Rights Principle.� Brooklyn Journal of International OECD (Organisation for Economic Cooperation and Development), Law 25: 51–97. 1996. “Trade Employment and Labor Standards: A Study of Core ———. 2003. “Integrating Trade and Human Rights in the Americas.� Workers’ Rights and International Trade.� OECD, Paris. Research Paper 26, Boston College of Law, Boston, MA. SSRN: Petersmann, Ernest-Ulrich. 2000. “The WTO Constitution and Human http://papers.ssrn.com/sol3/papers.cfm?abstract_id=470201. Rights.� Journal of International Economic Law 3 (1): 19–25. GATT (General Agreement on Tariffs and Trade), Negotiating Group ———. 2001. “Human Rights and International Economic Law in the on Dispute Settlement. 1989. “Non-Violation Complaints under 21st Century: The Need to Clarify Their Interrelationship.� Journal of GATT Article XXIII: 2.� Note by the Secretariat, MTN.GNG/ International Economic Law 4 (1, March): 3–39. 466 Susan Ariel Aaronson ———. 2002. “Time for a United Nations ‘Global Compact’ for Integrat- Paper 02-01, University of Toronto Faculty of Law, Toronto, Canada. ing Human Rights into the Law of Worldwide Organizations: Lessons SSRN: http://ssrn.com/abstract_id=307219. from European Integration.� European Journal of International Law 13 UNDP (United Nations Development Programme). 2002. Human Devel- (3): 621–50. opment Report: Deepening Democracy in a Fragmented World. New ———. 2006. “The WTO and Regional Trade Agreements as Competing York: UNDP. http://hdr.undp.org/reports/global/2002/en/. Fora for Constitutional Reforms: Trade and Human Rights.� In U.S. Congress. 2002. Bipartisan Trade Promotion Authority Act of 2002. Regional Trade Agreements and the WTO Legal System, ed. Lorand Washington, DC: Government Printing Office. http://www.law.cor Bartels and Federico Ortino, 281–312. Oxford, U.K.: Oxford Univer- nell.edu/uscode/html/uscode19/usc_sup_01_19_10_24.html. sity Press. U.S. GAO (General Accountability Office). 2009. “Four Free Trade Agree- Powell, Stephen J. 2005. “Regional Economic Arrangements and the Rule ments GAO Reviewed Have Resulted in Commercial Benefits, but of Law in the Americas: The Human Rights Face of Free Trade Agree- Challenges on Labor and Environment Remain.� Highlights of GAO- ments.� Florida Journal of International Law 17 (1): 59–100. SSRN: 09-439 (July), GAO, Washington, DC. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1115298. USTR (Office of the U.S. Trade Representative). 2005. “Participation, Silvan, Laura. 2004. “10 Years of NAFTA’s Commission on Environmental Empowerment, Partnership: Seeking Sustainable Results through US Cooperation in Mexico: Resolving Environmental Problems and Fos- Trade Capacity Building.� USTR, Washington, DC. tering Citizen Participation.� IRC Strategic Dialogue 1, International ———. 2009. “The President’s Trade Policy Agenda and 2008 Trade Relations Center, Silver City, NM. Agreements Report.� USTR, Washington, DC. http://www.ustr.gov/ Smith, Carolyn C. 2006. “Trade Promotion Authority and Fast-Track about-us/press-office/reports-and-publications/2009/2009-trade- Negotiating Authority for Trade Agreements: Major Votes.� Report policy-agenda-and-2008-annual-report. RS21004, Congressional Research Service, Library of Congress, Wash- ———. 2011. “2011 Trade Policy Agenda and 2010 Annual Report of the ington, DC. http://fpc.state.gov/documents/organization/73937.pdf. President of the United States on the Trade Agreements Program.� Sykes, Alan O’Neil. 2003. “International Trade and Human Rights: An USTR, Washington, DC. http://www.ustr.gov/webfm_send/2597. Economic Perspective.� John M. Olin Law and Economics Working van Hees, Floris. 2004. “Protection v. Protectionism: The Use of Human Paper 188, Law School, University of Chicago, Chicago, IL. SSRN: Rights Arguments in the Debate for and against the Liberalization of http://papers.ssrn.com/sol3/papers.cfm?abstract_id=415802. Trade.� Åbo Akademi University, Turku, Finland. http://web.abo.fi/ Thomas, Melissa. n.d. “The Rule of Law in Western Thought.� World Bank instut/imr/norfa/floris.pdf. Rule of Law and Development website, World Bank, Washington, Walker, Simon. 2009. The Future of Human Rights Impacts Assessments of DC. http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXT- Trade Agreements. Mortsel, Belgium: Intersentia. LAWJUSTINST/0,,contentMDK:20763590~menuPK:1989584~pageP WHO (World Health Organization). 2006. “Public Health, Innovation K:210058~piPK:210062~theSitePK:1974062,00.html. and Intellectual Property.� Commission on Intellectual Property 3D. 2005. “Planting the Rights Seed: A Human Rights Perspective on Agri- Rights, Innovation and Public Health, WHO, Geneva. http://www.who culture Trade and the WTO.� Backgrounder 1, 3/2005, 4, 3D, Geneva. .int/intellectualproperty/report/en/. ———. 2009. “Insights on Human Rights Impact Assessments of Trade Wilcox, C. 1949. A Charter for World Trade. New York: Macmillan. Policies and Agreements.� 3D, Geneva. http://www.3dthree.org/ WTO (World Trade Organization). 2001. “Trade and Labour Standards: A pdf_3D/HRIAsbackgroundinformation.pdf. Difficult Issue for Many WTO Member Governments.� DOHA Minis- Trebilcock, Michael J. 2001. “Trade Policy and Labour Standards: Objec- terial 2001, Briefing Notes, WTO, Geneva. http://www.wto.org/eng tives, Instruments and Institutions.� Law and Economics Research lish/thewto_e/minist_e/min01_e/brief_e/brief16_e.htm. 22 Dispute Settlement Amelia Porges The past 15 or so years have seen the emergence and of 100 percent compliance with the PTA’s obligations. elaboration of ever more complex preferential trade agree- Ensuring compliance through enforcement is essential if ments (PTAs), forming multiple and rapidly proliferating the gains are to materialize. PTAs therefore typically networks. Almost all of these PTAs include a dispute settle- include some mechanism incorporating elements of both ment clause of some sort. Indeed, dispute settlement provi- compliance enforcement and dispute settlement. sions have become a sine qua non for PTA negotiators, even By participating in a PTA with strong dispute settlement though the number of actual government-government dis- provisions, a government signals its level of commitment putes within PTAs is only a fraction of the hundreds of to private and public interests at home and abroad. Each existing agreements. PTA competes with other PTAs for investment, jobs, and Governments enter into PTAs expecting to secure eco- economic growth, in a field that becomes more crowded nomic benefits. In particular, estimates of the welfare bene- every year. Even if no disputes are anticipated, enforcement fits of PTAs normally assume that the parties will faithfully provisions in a PTA reinforce the precommitment of the implement their market access commitments. If private governments, make their promises more credible, and sig- investors doubt that the partners to a PTA will actually nal that the PTA is a solid platform for investment that will keep their commitments, they will not engage in the type create jobs and economic growth. of risk-taking investment that the PTA could otherwise Solid dispute settlement is even more important in generate. If a PTA is to be fully implemented and to yield North-South (or South-South) PTAs with asymmetrical the expected benefits, the agreement should, at a mini- power relations. Recently concluded PTAs in Latin Amer- mum, be equipped with institutions that facilitate infor- ica, Europe, and Asia demonstrate to a striking extent that mation exchange among the parties, help the parties as PTA obligations deepen, become more complex, and monitor implementation, and provide an incentive struc- provide more value, PTA partners seek more certainty than ture that meaningfully supports compliance. purely diplomatic dispute settlement can provide. A dispute settlement arrangement is part of this neces- In theory, the parties to a PTA are the masters of their own sary structure because there will inevitably be disagree- treaty and could design an original dispute settlement mech- ments in a PTA concerning the scope and nature of the anism from the ground up, or have no dispute settlement commitments that the parties have made. The PTA must mechanism at all. In practice, almost all PTAs rely on one of provide an orderly way for its members to settle disputes the three general types of dispute settlement mechanisms: and move on, or else the disputes will poison bilateral rela- diplomatic settlement by negotiation; judgments by standing tions, reduce the PTA’s benefit, and perhaps even lead to tribunals; or the World Trade Organization (WTO) model, in the demise of the agreement. which an ad hoc panel is convened to hear the dispute. Many Dispute settlement mechanisms are also needed to recent PTAs have adopted the third system, based on the ensure that the promises in a PTA are kept. Economic stud- WTO’s Dispute Settlement Understanding (DSU). The WTO ies on PTAs teach that where the parties’ tariffs are low ex model has provided a useful focal point for bargaining; its ante, a PTA between them will only produce gains if it familiarity means that negotiators and stakeholders under- involves deep integration provisions. Those provisions stand how it works and what trade-offs can be made. need to be backed up by enforcement. Every economic pro- This chapter, the last in the volume, discusses the jection of the gains from a PTA is based on the assumption options available for dispute settlement and enforcement 467 468 Amelia Porges provisions in PTAs. Following a brief literature review, we Kwak and Marceau (2006) present an updated cross- outline the three models of dispute settlement in PTAs, regional summary of PTA dispute settlement provisions discuss their scope and exclusions, and compare them and conclude that there is a real possibility of overlaps or from the standpoint of development, infrastructure cost, conflicts of jurisdiction between the WTO and PTA dispute and open regionalism. We then examine how PTAs define settlement mechanisms. They outline in tabular form the the scope of disputes that they will deal with, their han- dispute settlement provisions of the main PTAs in all dling of overlap and forum choice between the PTA and regions, with attention to issues that affect such overlaps or WTO dispute settlement systems, and various procedural conflicts. Morgan (2008) confirms the trend toward greater and institutional issues. Subsequent sections describe legalism in PTA dispute settlement but argues that the alternative dispute resolution procedures and examine absence of effective enforcement mechanisms forces the compliance procedures and enforcement issues and deci- parties to find negotiated solutions. He points out that sions. The final sections explore the reasons for the limited even within the North American Free Trade Agreement use of PTA dispute settlement procedures to date and (NAFTA), Canada opted for a negotiated settlement to its present the conclusions and recommendations that long-running softwood lumber dispute with the United emerge from the study. States, after a series of rulings in NAFTA and the WTO. Literature on PTA Dispute Settlement European Union (EU) PTAs The surge in the number of PTAs is a rather recent phe- Garcia Bercero (2006), a negotiator for the European Com- nomenon. As yet, few cross-regional comparative surveys mission, surveys the development of the Commission’s of PTA dispute settlement have been published, and those thinking on dispute settlement in trade agreements, start- that do exist focus on the dispute settlement mechanisms ing with the traditional diplomatic approach seen in the in particular PTA agreements, regions, or networks rather EU’s association agreements and other agreements before than on the application of the mechanisms. The reason is 2000, and discusses why the Commission’s preferences that the number of cases of government-to-government have shifted toward ad hoc arbitration procedures in the PTA disputes that can be examined is still small. free trade agreements (FTAs) with Mexico and Chile. In studies of earlier EU PTAs, Ramirez Robles (2006) finds that the political model of dispute settlement has been the General Surveys dominant model for EU association agreements, and The most significant recent general survey, by Donaldson Broude (2007) surveys the disputes brought under these and Lester (2009), concentrates on a sample of 20 recently earlier PTAs. Broude argues that the dispute settlement concluded PTAs, primarily in the Asia-Pacific area, and provisions of most EU PTAs contribute to EU regional almost all with dispute settlement systems based on the hegemony by encouraging and perpetuating nonjudicial- WTO model. The authors provide a detailed comparison ized bilateral dispute settlement, where the EU has advan- of the various stages of handling disputes in these PTAs tages. He points out that EU PTA partners do not even use and in the DSU and include a useful discussion of the the WTO system for settlement of their own disputes and institutions that administer dispute settlement. An older do not have the real option of recourse to judicial dispute study (Smith 2000) evaluates legalism in PTA disputes by settlement procedures in their relations with the EU. analyzing a coded dataset of 60 pre-1996 PTAs. Smith finds that legalism improves compliance by increasing the U.S. PTAs costs of opportunism and the probability of detection. He argues that negotiators, in drafting PTAs, weigh the bene- EU association agreements have opted for political settle- fits of improved treaty compliance against the costs of ment of bilateral disputes. Thus, the first PTAs to incorporate limited policy discretion for their own countries. PTA par- formal panel procedures of the type used in the WTO and its ties with high relative economic power accordingly favor predecessor, the General Agreement on Tariffs and Trade less legalistic dispute settlement, and so standing tribunals (GATT), were the Canada–U.S. FTA (CUSFTA) and NAFTA. such as the European Court of Justice (ECJ) exist only in U.S. PTAs still largely follow the CUSFTA/NAFTA model, those PTAs in which asymmetry of power is low. Deeper although post-2001 PTAs have departed from it in some integration also favors legalism because it generates more respects, as discussed below. Annex A gives an account of the economic gains and because the trade barriers involved dispute settlement choices made in negotiating these are more complex. agreements—choices that influenced the design of the WTO Dispute Settlement 469 Dispute Settlement Understanding (DSU). Hart, Dymond, Choi (2004) analyzes the dispute settlement provisions and Robertson (1994); Kreinin (2000); and Cameron and of a selection of East Asian PTAs and makes a number of Tomlin (2002) provide detailed accounts of the negotiations. specific suggestions for PTA dispute settlement procedures. Among the abundant literature on these agreements, Loung- Choi suggests that opening panel hearings in such disputes narath and Stehly (2000) analyze CUSFTA and NAFTA would be undesirable because it would increase pressure on disputes and argue that political pressure has impaired panelists by domestic interest groups. Luo (2005) discusses outcomes for Canada, as the weaker partner. Davey (1996) the “Asian way� of dispute settlement under ASEAN and presents a detailed retrospective on every CUSFTA dis- points out that, following the establishment of the ASEAN pute. He concludes that binational review of trade rem- Dispute Settlement Mechanism (DSM) in 1996, there were edy decisions under CUSFTA Chapter 19 has been gener- many actual disputes among ASEAN members concerning ally reasonable but could be improved, and he makes implementation of the ASEAN Free Trade Area, but none suggestions for minor changes to that end. He also finds were brought to the DSM. Luo suggests the use of an inde- that the general trade dispute procedures of Chapter 18 pendent enforcement body, or access for private parties, to were not an improvement on GATT procedure; in practice, counteract governments’ reluctance to engage in open con- the parties largely preferred GATT for their bilateral dis- flict. Nakagawa (2007) analyzes the use of dispute settle- putes, and he predicts that the NAFTA governments will ment by East Asian governments in the WTO and in PTAs. continue to prefer WTO procedures wherever possible. He argues that most trade disputes between Asian PTA par- Gantz (2006) surveys experience under the three NAFTA ties will be brought to the WTO and that settlement of dis- dispute settlement mechanisms (state-to-state trade dis- putes through negotiated deals will continue to be the trend putes, antidumping and countervailing duty binational in East Asia because of underlying economic factors and the reviews, and investor-state arbitration), reviews U.S. atti- limits on what dispute settlement mechanisms can accom- tudes toward NAFTA dispute settlement, and discusses the plish. Kawai and Wignaraja (2009) present the results of a decline in U.S. government support for NAFTA since 1994. survey of firms in East Asia. They find that most of the He rates Chapter 19 review of antidumping and counter- firms surveyed do not use PTA preferences because of lack vailing duty decisions as a success—although it enjoys little of information, costs related to rules of origin, and low support in the U.S. government, which has not included margins of preference. Obviously, if firms are not using such a provision in any post-NAFTA PTA.1 PTAs, there will be fewer PTA-related disputes. East Asian PTAs Latin American PTAs As Baldwin (2008) points out, East Asia’s regional integra- A wealth of data is available on dispute settlement in tion into “Factory Asia� took place initially not through PTAs in Latin America and the Caribbean.2 Sáez (2007) PTAs but through unilateral cuts in most favored nation finds that countries in the region have been very active in (MFN) tariffs. Baldwin notes the potential insecurity of dispute settlement, that they use the WTO even when such liberalization, which is not backed by any enforceable PTA dispute settlement mechanisms exist, and that in legal obligations. He characterizes the Association of South- Latin American PTAs, disputes on tariff application, east Asian Nations (ASEAN)–China FTA initiative in 2000 drawback, and excise tax discrimination have tended to as the trigger for similar moves by Japan, the Republic of peak during an agreement’s initial period, when tariffs are Korea, and others, resulting in the current “noodle bowl� of being phased out. East Asian PTAs. Wang (2009) describes the Dispute Settle- There is a substantial scholarly literature in Spanish ment Mechanism Agreement of the ASEAN–China FTA as examining the Andean Community institutions, the dis- a “landmark agreement,� since it is China’s first PTA to pro- pute settlement mechanism of the Southern Cone Com- vide for settlement of bilateral or regional disputes through mon Market (Mercosur, Mercado Común del Sur), the formal procedures, even to the extent of authorizing trade Latin American Integration Association/Asociación Lati- retaliation. Snyder (2009) analyzes China’s PTAs and finds noamericana de Integración (LAIA/ALADI), and other that almost all have used WTO dispute settlement as a regional institutions in the light of Latin American template, with the exception of China’s Closer Economic domestic and international legal doctrine. English- Partnership Agreements (CEPAs) with Hong Kong SAR, language sources consulted for this chapter include the China, and Macao SAR, China. Disputes under the CEPA valuable recent empirical studies on the Andean Com- agreements are settled diplomatically, by consultation and munity by Helfer and Alter (2009) and Alter and Helfer agreement between the parties. (2011), as well as a collection edited by Lacarte and 470 Amelia Porges Granados (2004), which covers dispute settlement under Political or Diplomatic Dispute Settlement LAIA/ALADI (Rojas Penso 2004), Mercosur (Opertti Political or diplomatic dispute settlement consists of set- 2004; Whitelaw 2004), the Andean Community (Vigil tling disputes by negotiation and agreement. It gives the Toledo 2004), and various Central American trade agree- parties to a PTA maximum flexibility. The agreement may ments (Echandi 2004). have no dispute settlement provisions at all; it may provide only for consultations; it may provide for consultations African PTAs and refer disputes to a political body for resolution; or it World Bank and International Monetary Fund (IMF) stud- may provide for referral to a third-party adjudicator but ies on African regionalism (Foroutan 1992; Yang and allow a party or parties to block the referral. Gupta 2005), as well as discussions in World Bank (2000), The choice of a political dispute settlement model often Schiff and Winters (2003), Gathii (2010), and Thorp reflects asymmetric power relations between the PTA par- (2010), note the dense web of PTAs in Africa. These studies ties. For instance, China’s Closer Economic Partnership generally find that, although most African PTAs have a Agreements with Hong Kong SAR, China, and with Macao high level of ambition for integration, they have not been SAR, China, state that the parties “shall resolve any prob- effective in eliminating intraregional trade and investment lems arising from the interpretation or implementation of barriers and have struggled with (or succumbed to) eco- the ‘CEPA’ through consultation in the spirit of friendship nomic conflict resulting from an asymmetric distribution and cooperation.� In pre-2000 European Union PTAs and of gains from liberalization. Implementation of PTAs has in the 1969 Southern African Customs Union (SACU), dis- been beset with obstacles, including linguistic differences, putes were settled exclusively through political processes. intra-PTA differences between common law and civil law The use of political dispute settlement may also reflect a systems, and a pervasive shortage of resources for PTA low level of ambition for implementation of intra-PTA lib- institutions, including those involved in dispute settle- eralization, as in the case of partial-scope agreements in ment. As noted by Essien (2006), general information on LAIA/ALADI, in ASEAN trade liberalization in the 1990s, the status of PTA courts in Africa is difficult to find. Gathii and in the Economic Cooperation Organization Trade (2010) presents a broader portrait of African PTAs, arguing Agreement entered into by Central Asian countries. that they must be understood as flexible regimes that The choice of political dispute settlement method may incorporate variable geometry, asymmetric obligations, also reflect a level of deep integration that gives both sides mechanisms for redistributing benefits, and commitments in a PTA real leverage, even without third-party adjudica- that are perhaps not meant to be enforced. If enforcement tion, as in the Australia–New Zealand Closer Economic is not intended, a scarcity of formal trade agreement dis- Relations Trade Agreement (ANZCERTA) and in the rela- putes should be no surprise. tions between Australia and New Zealand under the recently signed ASEAN–Australia–New Zealand agreement. The diplomatic model of dispute settlement was the Dispute Settlement and Enforcement: dominant model in all the EU’s pre-2000 association agree- The Basic Options ments and PTAs. These include the Europe Agreements In theory, the negotiators of a PTA start with a blank slate with Eastern European accession candidate countries; the and can choose any form or type of dispute settlement they Euro-Mediterranean Agreements with the Arab Republic wish. In practice, dispute settlement mechanisms in PTAs of Egypt, Israel, Jordan, Morocco, and Tunisia; and the fall into three broad groups: political or diplomatic dispute Stabilization and Association Agreements with Balkan settlement; systems based on a standing tribunal; and refer- countries (Ramirez Robles 2006). A party to one of these ral to an ad hoc arbitral panel, as in the WTO. In a negoti- agreements may refer to the agreement’s Association ating situation, the choice of system depends, first, on Council any dispute concerning the application or inter- whether the governments wish to have a third-party dis- pretation of the agreement, and the Association Council pute settlement procedure, rather than rely solely on nego- may, by consensus, adopt a binding decision to resolve the tiation to settle disputes. If they opt for a third-party deci- dispute. Broude (2009) describes this concept as “a case of sion maker, they can go down the path of establishing a faux institutionalization—the Association Council is a standing tribunal, or they can follow the currently domi- ministerial-level body, designed to meet but once a year.� nant approach of using a WTO-type ad hoc panel proce- He notes that, in practice, disputes are officially delegated dure. The pros and cons of each approach are discussed to an Association Committee that reports to the Council next. and are handled by diplomatic negotiation. In theory, these Dispute Settlement 471 agreements permit any party to have recourse to arbitra- Standing Tribunals tion if negotiations fail to settle a dispute, but the appoint- ment of arbitrators can be blocked, and there are no dead- The standing tribunals of the European Union represent lines or procedures. The agreements provide no organized the oldest system of this kind. Its example has been highly procedures for ensuring compliance with arbitral awards influential worldwide, but especially in countries with his- except for a “nonexecution clause� that permits a party to torical ties to Europe. take “appropriate measures,� even without going through a The EU system. The European Court of Justice (ECJ) is dispute settlement procedure, if it considers that the other the original PTA standing tribunal. Established in 1952 as party has failed to fulfill an obligation under the agreement. the Court of Justice of the European Coal and Steel Com- Garcia Bercero (2006) states that this diplomatic munity, it later became the Court of Justice of the Euro- approach has been an effective means of settling “low pro- pean Communities and is now known formally as the file trade irritants� but that some disputes linger unre- Court of Justice of the European Union. It now consists of solved for years, if one party is stubborn and the other is the 27-judge Court of Justice itself; a 27-judge court of first unwilling to blow up the relationship by taking retaliatory instance, the General Court; and the Civil Service Tribunal, action. He reports that the arbitration procedure has sel- for EU civil service employment disputes. In 2009 the ECJ dom been used, that the nonexecution clause has been received 562 new cases, completed 588 cases, and had 742 invoked very sparingly, and that most of these invocations cases pending at the end of the year. The 562 new cases have involved EU disputes with the Russian Federation. comprised 302 requests from EU member state courts for The standard dispute settlement procedure in preliminary rulings on issues of EU law, 143 direct actions, LAIA/ALADI is set forth in Resolution 114 of the associa- 105 appeals, and a few other cases (ECJ 2010a). tion’s Committee of Representatives. The resolution, Direct actions include enforcement actions brought by adopted in 1990, provides for consultations between the the European Commission against a member state for parties, after which a member country may request the failure to fulfill an obligation under Article 258 of the Committee of Representatives to propose a nonbinding Treaty on the Functioning of the European Union and solution. The association’s secretary-general has character- (very rarely) cases brought by one member state against ized this system as “virtually useless� and ineffective in another regarding nonfulfillment of EU Treaty obliga- resolving disputes (Rojas Penso 2004). As a result, parties tions (Article 259). In enforcement actions, the Commis- to the partial-scope agreements within LAIA/ALADI have sion sends a letter of formal notice to a member state— adopted specific dispute settlement procedures. Some the most recent official figures show that about 68 agreements simply rely on direct negotiations, and some percent of complaints are settled before this point (Euro- recent ones provide for third-party panels. pean Commission 2009). After giving the member state an The 1992 ASEAN Framework Agreement and the opportunity to reply, the Commission delivers a “reasoned Agreement on the Common Effective Preferential Tariff opinion�; about 84 percent of infringement procedures Scheme for the ASEAN Free Trade Area both provided for based on a complaint are settled before this stage. If the settlement of disputes by agreement between the parties, member state does not comply with the reasoned opinion with referral of nonsettled disputes to a ministerial-level by a deadline set by the Commission, the Commission body. (This scheme has since been replaced by a panel may bring a case before the ECJ; around 94 percent of mechanism, as discussed below.) these infringement procedures are settled before an ECJ To a notable extent, PTAs that started with diplomatic ruling. Thus, only 6 percent of all procedures are resolved or political dispute settlement have moved toward rule- by the court. Judgments under Article 259 or its predeces- oriented third-party dispute settlement modeled on the sor provisions have been extremely rare (fewer than five DSU. Examples include ASEAN, which replaced earlier since 1951). In practice, if an EU government or stake- arrangements with the 2004 ASEAN Protocol on Enhanced holder has a problem with another member state’s com- Dispute Settlement Mechanism; SACU, which imple- pliance with EU Treaty rules, it lobbies the Commission to mented a DSU-type scheme in the 2002 SACU Agreement; negotiate with the noncomplying government and possi- and Mercosur, an agreement formally within the LAIA/ bly bring an action under Article 258. The Commission ALADI framework, which provides an elaborate third- then takes on the resource and reputational costs of nego- party dispute settlement system, including appellate tiation and litigation and is also free to pursue its own review. The FTAs and economic partnership agreements institutional agenda. (EPAs) negotiated by the EU since 2007 have shifted to The ECJ and the European Commission enforcement panel-based third-party dispute settlement. infrastructure represent the maximum in treaty enforcement, 472 Amelia Porges as measured by activity and resources. At the end of 2008, PTAs were negotiated, the GATT dispute settlement mech- the Commission was handling 1,557 complaints and anism was in disuse or did not otherwise provide a positive infringement files. Complaints accounted for two-thirds of model for enforcement of obligations. The project of all cases other than those regarding late implementation of building the European Community and then the European EU directives. In 2008 the Commission opened 2,223 Union provided a stronger model for PTA ambitions. infringement procedures, sent 512 reasoned opinions, and The most judicialized South-South PTA dispute settle- referred 209 cases to the ECJ (European Commission 2009). ment institutions, the Andean Tribunal of Justice (ATJ) The ECJ also provides guidance to the courts in all mem- and the Andean Community General Secretariat, were ber states, through preliminary rulings on EU law requested established in 1979 as part of the Andean Pact. Treaty by national courts. These preliminary references ensure amendments in 1996 created a new Andean Community, uniform application of EU law, and the principles they strengthened the ATJ, gave the General Secretariat a establish affect the entire EU legal order. They have given stronger role, and reduced diplomatic elements in the pro- the ECJ a platform for establishing fundamental principles cedures. The General Secretariat may now initiate a non- of EU law, such as its direct effect in national law, and legal compliance investigation on its own and must initiate an doctrines safeguarding freedom of movement of goods investigation in response to a complaint by a government and services. Private actors can bring domestic court cases or private party. There have been up to 30 such cases per in order to obtain an ECJ preliminary ruling. year.3 The General Secretariat sends a notice to the govern- The ECJ and the judicial structure under it represent a ment concerned, which must respond. The General Secre- very large commitment of resources. Its 27 judges, one for tariat then issues a reasoned opinion, which is published each member state, are assisted by a registrar; 8 Advocates- and with which the government concerned is obligated to General who provide impartial advisory opinions on the comply. If the government does not comply, then the Gen- cases before the Court; and a large staff, including many eral Secretariat, a complaining member state, or a private translators, housed in a new building in Luxembourg. The party whose rights have been affected may bring a non- ECJ also acts as an appellate court for cases brought before compliance case (acción de incumplimiento) against the the 27-judge General Court or the Civil Service Tribunal. noncomplying country to the ATJ, which sits in Quito. The 2011 draft budget for the Court of Justice of the Euro- There have been up to 20 such cases per year.4 If the ATJ pean Union, comprising all these courts, is projected at makes a finding of noncompliance and the losing govern- 345,293,000 euros (about US$450 million), not including ment fails to comply by the deadline set by the ATJ, the the Commission’s enforcement expenses. Of the total, 75 ATJ may initiate a summary procedure (procedimiento- percent goes for personnel and 25 percent for buildings sumario) in response to a request by the General Secre- and other costs (EU 2010). tariat, a member state, or an affected private party and may Other standing tribunals. The obligations of the Euro- authorize sanctions against the noncomplying government pean Free Trade Area (EFTA) and the European Economic by other member states (Vigil Toledo 2004; CAN 2008a). Area (EEA) are enforced by the EFTA Surveillance Author- Alter and Helfer (2011) observe that, compared to the ECJ, ity and the EFTA Court. The most recent figures, for 2008, the ATJ has been deferential to Andean states and unwill- show an actual cost for both of 3,606,035 euros (EFTA ing to push for compliance with Andean law and that the Court 2009). ATJ has “refused to serve as the engine of Andean legal A number of South-South PTAs have patterned integration�—which they characterize as a politically their dispute settlement institutions on the ECJ and the prudent path in the face of Andean states’ “tepid� com- European Commission: they include the Andean Commu- mitments to Andean integration. nity, the Economic Community of West African States The ATJ also can issue rulings in response to references (ECOWAS), the West African Economic and Monetary from national courts; such rulings account for 90 percent Union/Union Économique et Monétaire Ouest-Africaine of ATJ case law. Ninety-six percent of preliminary refer- (WAEMU/UEMOA), the Economic and Monetary Com- ences through 2007 involved intellectual property disputes, munity of Central Africa (CEMAC, Communauté for three reasons: first, private litigants have rarely used Économique et Monétaire de l’Afrique Centrale), the East Andean rules to challenge other policies; second, intellec- African Community (EAC), and the Common Market for tual property agencies in the Andean countries actively Eastern and Southern Africa (COMESA).The reasons for encouraged such references and incorporated them into following the EU model are partly historical and (in the their domestic decision making; third, national courts case of African PTAs) partly attributable to the influence of rarely send novel questions to the ATJ; fourth, there is no the EU as the regional hegemon. At the time many of these infrastructure of scholars and practitioners proselytizing Dispute Settlement 473 for Andean law; and, fifth, these conditions have continued budget of US$60 million, half of which is financed by the same no matter what the level of political support for aid donors (see EAC 2010). the Andean Community project (Helfer and Alter 2009; • The COMESA Court of Justice was established in 1994 Alter and Helfer 2011). The fiscal 2008 budget for the and appointed in 1998, but since moving to Khartoum Andean Tribunal was US$1,170,667.5 in 2006, it has reportedly faced problems with inade- As part of an overall initiative to strengthen Central quate funds and staff, lack of a physical location, and the American regional integration, governments in the region need to adapt to shari’a law. It has been unable to meet established the Central American Court of Justice (CACJ), even twice a year. Its 2001 budget was US$595,538 or Corte de Justicia Centroamericana, in 1994. The CACJ (COMESA 2000, 2009; East African 2006; AICT data).7 has almost never been used for trade disputes. In the late • The SADC Tribunal of 10 part-time members was for- 1990s, the countries established a Central American Trade mally established in 1992 and was inaugurated in 2005. Dispute Resolution Mechanism, modeled on the WTO, It meets, as required, in Windhoek. It has a registrar and which, since 2003, has applied to all countries in the region 16 employees. The SADC budget is largely funded by (Echandi 2004). donor countries. The tribunal’s cases have included The Caribbean Court of Justice was created in 2001 with appeals by SADC employees and three cases in which a dual role: to serve as a final court of appeal in civil and the court condemned Zimbabwe’s land reform program criminal cases for its member states, and to interpret the as racially discriminatory and illegal under the SADC Treaty of Chaguaramas, which established the Caribbean Treaty. Faced with noncompliance, the complainants Community (CARICOM). The extent of regional support have used South African courts to seize property of for this court is unclear, and it lacks the institutional sup- the Zimbabwe government, and the tribunal has asked port of a secretariat to monitor compliance (INTAL 2005). the SADC heads of state to consider a request by one In 2007 the CCJ’s administrative expenses amounted to of the complainants that the SADC expel Zimbabwe US$32.2 million (CCJ 2008). (Nyaungwa 2010; Reuters 2010). A number of African PTAs have also established courts on the ECJ model; these include ECOWAS, WAEMU/ Dispute Settlement by Referral to UEMOA, CEMAC, the EAC, and COMESA. As the follow- Ad Hoc Panel (WTO Model) ing examples illustrate, a regional tribunal created to enforce trade law may also become involved in other issues, The third, currently dominant, model for PTA dispute set- and vice versa: tlement is based on the WTO’s dispute settlement system (originally developed in the GATT). A panel is convened • The ECOWAS Community Court of Justice was created for one dispute (thus, it is “ad hoc�), with terms of refer- in 1991, but its members were only appointed in 2000. ence limited to that dispute. The panel hears the written The court sits in Abuja and has a modest budget for and oral arguments of disputing parties, issues a written handling both enforcement of ECOWAS norms and ruling applying the trade agreement’s law to the dispute, human rights issues. Originally, only states could bring and then disbands. The WTO’s dispute settlement proce- disputes, but since 2005, individuals have been able to dures have shaped expectations of governments and stake- bring cases, including complaints based on human holders regarding credible dispute settlement and enforce- rights instruments. The court has been active recently, ment within trade agreements. As a result, PTA negotiators but mostly on cases with a human rights dimension have converged on the WTO-like model. Dispute settle- (Banjo 2007; Hessbruegge 2011; Daily Independent ment procedures of this type are very widespread and 2011). appear in virtually all new PTAs, such as the following list, • The UEMOA court has a permanent building at its seat which includes some PTAs that have been signed but not in Ouagadougou, and a modest budget.6 CEMAC’s yet ratified: Court of Justice sits in Ndjamena. Each court issues fewer than 10 decisions per year. • Australia: FTAs with Chile, Singapore, and Thailand • The Court of Justice of the East African Community, • ASEAN Enhanced Dispute Settlement Mechanism, which which meets in Arusha, Tanzania, was dormant from covers disputes under at least 46 ASEAN agreements on 1999 until receiving its first case in December 2005. It tariff preferences, tariff nomenclature, investment, serv- hears both disputes between member states and prelim- ices, mutual recognition of standards or certification, inary references by courts in member states. Its fiscal customs, and the like; ASEAN FTAs with China and 2010/11 budget was US$2,841,777, out of a total EAC Korea; the ASEAN–Australia–New Zealand FTA 474 Amelia Porges • Canada: FTAs with Chile, Colombia, Costa Rica, Israel, Sw F 5,691,000, of which personnel costs account for Jordan, Panama, and Peru 69 percent, or Sw F 3,909,500 (WTO 2009, 52). • Chile: FTAs with Central America, China, Colombia, These system costs exclude the costs of participation for EFTA, Japan, Korea, Mexico, Panama, and Peru the parties. Some governments maintain an internal legal • China: FTAs with Costa Rica, New Zealand, Pakistan, staff or otherwise represent themselves in trade disputes. Peru, and Singapore Others hire counsel or expect stakeholders to pay for coun- • EFTA: FTAs with Canada, Colombia, Croatia, Egypt, sel engaged by the government. The standard ceiling fees Israel, Jordan, Korea, Lebanon, the former Yugoslav for legal assistance (without further subsidy) set by the Republic of Macedonia, Morocco, and the Palestinian Advisory Centre for WTO Law provide a lower-bound esti- Authority mate for legal costs in the WTO: Sw F 47,628 for consulta- • EU: FTAs with Chile, Korea, and Mexico and the recent tions and Sw F 143,856 for panel proceedings, or Sw F EPA with the Caribbean Forum of African, Caribbean, 191,484 together (ACWL 2007). and Pacific (ACP) States (CARIFORUM) • India: FTAs with Chile and Mercosur Choosing among the Options • Japan: EPAs with Brunei Darussalam, Chile, Indonesia, Malaysia, Mexico, the Philippines, Singapore, Switzer- In considering dispute settlement, the first choice is land, Thailand, and Vietnam whether to rely exclusively on diplomatic or political dis- • Mercosur: Protocol of Olivos, with Argentina, Brazil, pute settlement or to provide for third-party dispute settle- Paraguay, and Uruguay ment of some sort. The key difference between the two is • Mexico: FTAs with Bolivia, Colombia, EFTA, Israel, that in the former, a complaining party that has tried and Nicaragua, the Northern Triangle countries (El failed to negotiate a mutually satisfactory solution does not Salvador, Guatemala, and Honduras) and Uruguay have the option of obtaining a determination by a neutral • New Zealand: FTAs with Thailand and Singapore third party. In a tribunal or ad hoc panel system, if there is • Singapore: FTAs with the Gulf Cooperation Council no agreement, the complaining party can refer the dispute (GCC), India, Jordan, Korea, Panama, and Peru; to a neutral decision maker who rules on the dispute. Set- Trans-Pacific Strategic Economic Partnership (Trans- tlement by negotiation is possible in any system; indeed, Pacific SEP). the classic WTO model requires that the complaining party • United States: NAFTA; U.S. FTAs with Australia, give the responding party an opportunity to settle the Bahrain, Dominican Republic–Central America Free problem through bilateral consultations, and the prelimi- Trade Agreement (CAFTA–DR), Chile, Colombia, nary stages in a tribunal system may also result in compli- Israel, Jordan, Korea, Morocco, Oman, Panama, Peru, ance or a compromise settlement. and Singapore Negotiators may choose to rely solely on diplomatic or political dispute settlement if the PTA is unambitious or The cost of an ad hoc panel system depends on how represents only a low sunk investment in integration. They many panels exist. If there are none, the cost of dispute set- may also feel that government-government dispute settle- tlement approaches zero, as the disputes that do not reach ment is not necessary if the PTA partners have a transpar- the panel stage will be supervised by existing PTA institu- ent commercial and legal environment with stable trade tions. The out-of-pocket system costs of panels—panelist relations and a high degree of economic integration, as in compensation and expenses; hearing venue; and clerical, ANZCERTA. And there may be no place for third-party translation, and interpretation services—depend on the dispute settlement in a PTA with strongly asymmetric parties’ procedural choices and the number of disputes. relations, such as the CEPAs that China has negotiated The WTO’s annual budget for dispute settlement panels with Hong Kong SAR, China, and with Macao SAR, (excluding the organization’s first three years) has ranged China. Political or diplomatic means of dispute settlement from a high of 1,195,300 Swiss francs (Sw F) in 1999 to a are unlikely, however, to provide a sufficient incentive low of Sw F 655,592 in 2006. For 2011, it was estimated at structure to keep markets open in times of economic Sw F 987,000 for panels and Sw F 50,000 for arbitrations stress. (WTO 2009, 52). The WTO’s costs for panel proceedings Most recent PTAs have opted for third-party dispute are low in relation to the number of disputes handled settlement because of significant factors that distinguish because WTO members whose delegates serve as panelists the new and enhanced PTAs negotiated since 1994 from contribute their services without compensation. The their predecessors of the GATT era, as described by Schiff estimated 2011 budget for the WTO Appellate Body is and Winters (2003). Dispute Settlement 475 First, many of these agreements have embraced deep Defining the Grounds for Complaint integration, going well beyond border measures to cover Breach of obligations. Many PTAs simply state that dispute subjects such as investment, services, and domestic regula- settlement will be available in cases of violation of the tion. Where a partner has relatively low border barriers, PTA obligations or of failure to implement the PTA; some expansion of the scope of the PTA to behind-the-border also allow for the use of the mechanism to settle disputes measures may be necessary if the PTA is to offer substantial regarding interpretation of the PTA. For instance, in economic benefits—as Francois and Manchin (2009) find CUSFTA and later U.S. FTAs, the dispute settlement provi- for a possible PTA between the EU and the Common- sions apply with respect to avoidance or settlement of dis- wealth of Independent States (CIS). The increased level of putes regarding the interpretation or application of the ambition involved in a deep integration agreement agreement, or regarding measures considered to be incon- requires the parties to make a greater resource investment sistent with the obligations of the agreement. Some PTAs in implementing the agreement. Accordingly, negotiators (e.g., the Japan–Chile and ASEAN–Australia–New Zealand typically choose legalized, formal dispute settlement, fol- agreements) draw instead on the formulation in GATT lowing some widely understood model—currently, that of Article XXIII:1(a), which permits invocation of the dispute the WTO. In North-South PTAs with asymmetrical power settlement mechanism when a “benefit accruing . . . relations, binding third-party dispute settlement becomes directly or indirectly under this Agreement is being nulli- even more important (World Bank 2000, 69–70). Recently fied or impaired . . . or the attainment of any objective of concluded PTAs in Latin America, Europe, and Asia this Agreement is being impeded as the result of the failure demonstrate to a striking degree that as PTA obligations of another [party] to carry out its obligations under this deepen, become more complex, and provide more value, Agreement.� PTA partners seek more certainty than can be had through Nonviolation nullification or impairment of concessions purely diplomatic dispute settlement. or obligations. Some PTAs give their parties not just a rem- Second, the dominant paradigm has shifted to integra- edy against violation or noncompliance by another party, tion into the globalized economy through open regionalism. but also a remedy against measures of other parties that are “Open regionalism� implies that PTA parties actively seek PTA-consistent but still take away the benefit of bargained- inclusion in global supply networks under transparent, for PTA market access. These “nonviolation nullification or rule-of-law conditions by using the PTA to secure market impairment� remedies are modeled on GATT Article access rights, and by turning away from import-substituting XXIII:1(b).8 GATT and WTO panels have interpreted Arti- industrialization or administrative protectionism. Open cle XXIII:1(b) as providing for recourse when (a) benefits regionalism can also lead to coexistence of PTA networks, that could reasonably have been expected at the time of a docking (legal connection) of PTA networks, and even mul- negotiated market access concession (b) are nullified or tilateralization of PTA networks. Dispute settlement can impaired by (c) a later (GATT-consistent) government reinforce open regionalism in the first sense by ensuring full measure that upsets the conditions of competition between PTA implementation and reinforcing the PTA’s lock-in domestic and imported products. Under the GATT and effect. Systems based on a regional court can be an impor- now the DSU, remedies in such “nonviolation� cases are tant focus of region building, as seen in the case of the limited to compensatory tariff reduction on other prod- European Court of Justice. However, if the goal is to connect ucts; the WTO cannot require a member to alter measures PTAs into larger networks, ad hoc panel systems are easier to that are WTO-consistent. Almost all GATT nonviolation merge than court-based systems of dispute settlement. disputes concerned subsidies that were GATT-consistent but distorted trade. Structuring a PTA Dispute Settlement PTAs with panel-based dispute settlement show con- System: More Decisions tinued interest in nonviolation remedies as a means of Some threshold decisions are required in order to set the protecting the market access and other benefits that these system’s parameters. The agreement will need to define the agreements provide. However, the PTAs that provide such range of possible complaints, or “causes of action,� as well remedies tend to explicitly identify which benefits are thus as their potential subject scope. Since practically all PTAs protected, typically by citing specific PTA chapters. For overlap with the WTO (and may overlap with other PTAs), instance, all U.S. FTAs explicitly permit disputes regarding PTA negotiators must also decide what will happen in the nonviolation impairment of benefits accruing under event that a party brings claims under the WTO concern- specific identified chapters, usually those concerning ing a measure disputed in the PTA. market access and national treatment for trade in goods 476 Amelia Porges (including rules of origin), cross-border services trade, obligations are often excluded from formal dispute settle- procurement, intellectual property rights (IPRs), and, ment. Such an exclusion makes sure that no dispute set- sometimes, technical barriers to trade (TBTs). Some PTAs tlement panel will ever read “should� as “shall.�10 As a provide nonviolation remedies only for certain chapters— corollary, when obligations regarding a subject area are for example, only for goods, or for goods and services, or limited to soft law, that area is likely to be excluded from for goods, services, and various other categories that may formal dispute settlement.11 include TBTs, sanitary and phytosanitary (SPS) measures, PTAs may also exclude areas from dispute settlement in aviation, procurement, and IPRs.9 A few PTAs make provi- order to ensure policy space for domestic regulation or to sion for disputes regarding nullification or impairment of avoid PTA challenges to determinations by domestic regu- PTA benefits generally (e.g., Japan–Switzerland and lators in particular cases. For instance, some PTAs exclude Canada–Israel). Other PTAs only provide for settlement of from dispute settlement any complaint regarding denial of disputes regarding noncompliance with obligations under rights to temporary entry and stay by business visitors the agreement (the Australia–Singapore, Chile–China, unless the complaint concerns a pattern of practice and the Chile–EU, Japan–Indonesia, and Japan–Mexico PTAs, and nationals involved have exhausted local administrative many others) or explicitly exclude any possibility of non- remedies.12 The objective is clearly to prevent the agree- violation complaints (the ASEAN–Australia–New Zealand ment’s being used for immigration litigation. Some PTAs agreement). exclude TBT or SPS issues, or both, for similar reasons. The Japan–Switzerland FTA limits possible complaints regard- ing the effect of taxation on the PTA while subordinating The Target of a Complaint the PTA to tax treaty obligations, following the example of Measures by a government. It goes without saying that PTA Article XXII:2 of the WTO General Agreement on Trade in dispute settlement deals with government measures—that Services (GATS). In PTAs with competition chapters, the is, existing laws, regulations, or other official actions or fail- relevant chapter is often excluded from dispute settlement ures to act that engage state responsibility, as opposed to because competition authorities see panel review of their actions of private parties. Although there is at least one case decisions as unwelcome and inappropriate. of a dispute (in Mercosur) concerning trade-obstructing PTAs may exclude areas from PTA dispute settlement actions by private parties (see annex B), the actionable when the PTA’s obligations merely reaffirm WTO obliga- measure was a PTA party’s failure to ensure free circulation tions. For instance, if a PTA simply confirms TBT or SPS of goods as required by the PTA. obligations that are derived from the WTO Agreement, it Proposed measures. Some PTAs permit disputes concern- may exclude TBT or SPS issues from PTA dispute settlement. ing proposed measures, such as pending legislation. The first Finally, PTAs may exclude panel interpretation of sub- modern example of such a provision appears in CUSFTA jects that are reserved to specific bodies in one of the par- Chapter 18. Both CUSFTA and NAFTA provide for consul- ties. For instance, all New Zealand PTAs that have formal tations and for requests for a panel decision regarding dispute settlement mechanisms provide that the interpre- whether pending legislation in a partner is consistent with tation of the 1840 Treaty of Waitangi between the British the PTA. Since 2001, U.S. PTAs have permitted consultations, Crown and Maori tribes shall not be subject to dispute set- but not arbitral panel proceedings, on pending legislation. tlement under the PTA. Under New Zealand law, matters As Donaldson and Lester (2009) note, dispute settle- concerning this treaty are reserved to a special tribunal. ment about proposed measures raises significant policy issues. It can be a waste of resources for panels to consider Overlap between the PTA and Another Forum measures that may never be enacted, but early consulta- tions can help limit or prevent trade damage by persuading Overlap in obligations between two legal regimes occurs governments not to enact measures that would violate when the same parties take part in two separate regimes their PTA obligations. and both regimes regulate the matter in dispute at the same Subject matter exclusions. PTA dispute settlement proce- time. Almost all PTAs overlap with the WTO Agreement, as dures generally apply only with respect to the rights and PTAs and the WTO both require national treatment and obligations provided in that PTA. Some subject areas are ban quantitative restrictions in trade. Indeed, many PTAs often excluded from dispute settlement even when they are simply incorporate GATT Articles III and XI by reference. included in a PTA. Some PTA members are also members of another overlap- If a PTA contains “soft-law� obligations that urge but ping PTA, as happens with CAFTA–DR and the Central do not mandate economic or other cooperation, those American Common Market (CACM), SACU and the Dispute Settlement 477 Southern African Development Community (SADC), and the UN Convention on the Law of the Sea or any other ANZCERTA and the ASEAN–Australia–New Zealand FTA treaty concluded by the European Community (MOX (Pauwelyn 2004). Plant Case 2006). When both legal regimes have independent dispute PTA negotiators have also required parties to give prece- settlement regimes, overlap affords opportunities for a dence to PTA rules that provide more policy space than complaining party to choose the most advantageous forum WTO rules (or otherwise are “WTO-minus�) by specifying or to relitigate a case in one forum (usually, the WTO) after the forum for particular disputes. For instance, NAFTA’s an unsatisfactory outcome in another (the PTA). For Article 104 provides that in the event that obligations example, Brazil brought a Mercosur challenge to an Argen- under certain listed environmental agreements are incon- tine antidumping measure on poultry; the Mercosur arbi- sistent with NAFTA obligations, the environmental agree- tral tribunal found that the measure was not regulated by ments prevail, and under Article 2005(3), if a defending Mercosur law and was therefore not inconsistent with party claims that Article 104 would apply, the complaining Argentina’s Mercosur obligations. Brazil then went to the party can only bring the dispute under NAFTA. Article WTO, and the WTO panel found that the antidumping 2005(4) similarly gives the responding party the right to measure violated the WTO Agreement on Antidumping have an SPS or TBT dispute heard only under NAFTA, (Piérola and Horlick 2007). In another case, Brazil pre- which was considered to afford more policy space. The vailed in a Mercosur challenge to an Argentine textiles safe- Chile–Mexico and Canada–Chile FTAs have similar clauses. guard, but Argentina did not remove its quotas, and Brazil NAFTA also gives a third NAFTA party the right to force obtained a settlement only after bringing the dispute to the a preference for PTA dispute settlement. Under NAFTA WTO (Kwak and Marceau 2006). Similarly, Argentina Article 2005(2), before a NAFTA party initiates WTO dis- brought a complaint against Chile’s price band tariffs to pute settlement against another party on grounds that are the WTO after Chile failed to comply with a nonbinding “substantively equivalent� to those available under NAFTA, PTA panel decision (Tussie and Delich 2005). it must notify the third NAFTA party. If the third party For any government, and particularly for developing wishes to litigate regarding the matter under NAFTA, it countries, relitigation poses a resource burden, particularly must inform the complaining party and consult; if the par- if a developing country must mount multiple defenses of ties fail to agree on the forum, the dispute is normally to be the same measure in different forums. Overlap also pres- settled in NAFTA. ents the possibility of conflicting rulings, as in the Preference for the WTO. The other extreme is repre- Argentina poultry case (Pauwelyn 2006). A WTO panel sented by the EU–Chile PTA, which provides that if a does not have jurisdiction to rule on whether a measure dispute concerns a breach of a PTA obligation that is violates a PTA, and vice versa. Domestic law doctrines that equivalent in substance to a WTO obligation, it must be curb duplicate litigation, such as lis alibi pendens and res brought in the WTO. Once a forum is selected, it is to be judicata, are not a solution because the obligations are not used to the exclusion of any other, and all arguments the same.13 Even if GATT Article III is incorporated by ref- regarding forum choice must be resolved within the first erence into a PTA, in the PTA context it is part of PTA law 30 days (Garcia Bercero 2006). and is subject to exceptions and dispute settlement proce- Binding election of forum. Most recent PTAs use some vari- dures that may not be the same as in the WTO (Kwak and ant of the approach adopted first in CUSFTA. That model Marceau 2006). allows disputes arising under both the PTA and the WTO to be settled in either forum at the discretion of the complaining party but provides that, after the “initiation� of dispute settle- Resolving Overlap ment, the procedure employed must be used to the exclusion There are three options for dealing with forum shopping in of any other. (The “initiation� point can be defined as the dispute settlement: give precedence to the PTA proceeding; parties wish; CUSFTA pegs it at the point of referral to a give precedence to the WTO or other proceeding; or allow panel, so that the complaining party can make its choice after the parties to choose but prohibit relitigation. consultations.) The complaining party thus has the option Preference for PTA rules. A PTA may require that all dis- to choose the strongest substantive and procedural rules, putes between PTA parties involving PTA provisions be while duplicative proceedings are excluded. Provisions of settled exclusively within the PTA. The EU Treaty has such this type appear, for instance, in CUSFTA and all later U.S. a requirement, and in the MOX Plant Case, the ECJ inter- FTAs; in Mexican PTAs; and in the China–New Zealand, preted the relevant EU Treaty article as barring Ireland Japan–Indonesia, Japan–Switzerland, Australia–Thailand, from bringing a dispute against the United Kingdom under and SACU–EFTA PTAs. Under Mercosur’s Protocol of 478 Amelia Porges Olivos, which was signed in 2002 in the wake of the poultry complaint. It typically also requires the respondent party to case discussed above, if a dispute may be brought in more consult promptly, and it may obligate the respondent to than one forum, the complaining party may select the forum, bring relevant officials to consultations and to provide suf- or the parties may agree on a forum, but once the dispute is ficient information to facilitate settlement of the dispute. initiated, none of the disputing parties may go to another The consultation clause must define the scope of issues forum. Variations on this formula permit a dispute to be on which the parties are obligated to consult upon request. brought again in another forum if the parties have so Some agreements define this broadly: the consultation agreed (as in the China–ASEAN and ASEAN–Australia– clause in the ASEAN Enhanced DSM includes “any matter New Zealand PTAs), or if substantially separate and dis- affecting the implementation, interpretation or application tinct rights and obligations under different international of the Agreement or any covered agreement.� The consulta- agreements are in dispute (e.g., under the ASEAN–Japan tion clause will also determine whether a consultation and Australia–Chile PTAs). request can include both existing and proposed measures Given a choice between bringing a dispute under the and whether it can include both breaches of the rules and WTO or under a PTA, most complainants now choose the nonviolation nullification or impairment of trade benefits. WTO, which is one reason there are so few PTA disputes. As already noted, the consultation or pre-panel phase of a Before the establishment of the WTO, two Canada–U.S. dispute may also extend to mediation, conciliation or good disputes that could have been brought under GATT offices, or other forms of alternative dispute resolution. Articles XI and III were brought under CUSFTA instead, A consultation clause is important because it gives any because at the time, CUSFTA offered the quickest and most PTA party a right to get another party to have a focused binding dispute settlement mechanism.14 The choice talk about market access barriers in relation to the PTA’s would be different today. rights and obligations. The consultations provide a cost- Complainants prefer the WTO for several reasons: the effective opportunity for the parties to settle their dispute large body of cases (with appellate review), which offers with maximum control over the outcome, by negotiation greater clarity and certainty about WTO obligations, and and agreement. The consultation request is the first visible greater predictability about the likely outcome of a dispute; formal document, but it usually only comes after extensive the fact that the WTO panel process cannot be blocked; the contact between stakeholders and the complaining govern- stabilizing effect of WTO institutions; and the availability ment, or between the governments concerned. of postjudgment compliance obligations and remedies. Consultations are also important because they provide a WTO proceedings may also provide a unique opportunity key opportunity to clarify the facts. Governments normally to mobilize third-party support and exert political pressure do not know the details of a PTA partner’s trade or regula- on the respondent party (Piérola and Horlick 2007). Fur- tory regime; they do not know foreign law; and they do not thermore, WTO panels are composed of neutrals, whereas know what aspects of government regulation will have the PTA panels often include representatives of the parties. most impact on trade flows or the interests of stakeholders. Gantz (2006) suggests that this is a factor in the U.S. prefer- Before the consultations, the government may not have ence for the WTO forum over NAFTA. collected the facts that it needs to determine its PTA rights and prove a violation of PTA law. The government may also not have a sense of the range of options for PTA- Options for Procedures consistent implementation. Consultations provide an In constructing a dispute settlement mechanism, nego- opportunity for a government to gather information to tiators need to consider what procedures to make avail- help evaluate a case before committing to it and to orient able, what financial and infrastructural support will be its litigation strategy and settlement negotiations to maxi- necessary, and what complaints the panels or tribunals mize commercial benefit. For the respondent government, may handle. consultations may offer an opportunity to reduce litigation costs by persuading the other side that certain claims are not worth litigating. Consultations Both functions of consultations—settlement and Whether a diplomatic or political dispute settlement fact-gathering—can also be fulfilled through discussions scheme or an ad hoc panel procedure is employed, a formal in the framework of a specialized PTA committee. Some dispute officially starts with a request for bilateral consulta- PTAs provide that committee consultations can take the tions. The consultation clause may require that the request place of dispute settlement consultations procedurally, as be in writing and that it state the legal grounds for the discussed below. Dispute Settlement 479 Cases of urgency or perishable products. Some PTAs (e.g., the obligation to start the panel appointment process; the the Chile–U.S., Canada–Peru, and ASEAN–Japan agree- Korea–Singapore, China–Singapore, and Australia–Chile ments and the Trans-Pacific SEP) provide accelerated PTAs are in this category. Under the other approach, after consultation timetables for cases of urgency, including consultations are exhausted, the complaining party refers those involving perishable products, following the example the dispute to a political body for conciliation, and after a of Article 4.8 of the DSU. Most agreements do not define specified period of time, or if conciliation otherwise fails, “perishable,� but some do define it as including perishable the complaining party delivers a written request for a agricultural or fish products. The 2009 Canada–Colombia panel to the other party, triggering the panel selection and Canada–Peru PTAs provide accelerated consultations process. This is the approach taken in NAFTA, in later U.S. for “cases of urgency, including those concerning perish- PTAs, and in some later Mexican and Canadian agree- able goods or otherwise involving goods or services that ments, under which the political body is a commission or rapidly lose their trade value, such as seasonal goods or joint committee, nominally at the ministerial level. Where services.� This clause recognizes that if the harvest period there is potential for settlement, the conciliation phase for a crop is limited and the crop is perishable, delay can provides another chance to negotiate, but in practice, the make dispute settlement worthless except as a deterrent to required meeting of the political body can consist of a repetition. brief conference call. Requirement to have consulted. The GATT, the WTO, and all PTAs reviewed for this chapter require that before a Infrastructure and Support for Dispute Settlement party can use the agreement’s common resources for third- party arbitration of a dispute, it must give the responding Institutions. PTA negotiators will need to decide how the party an opportunity to settle the dispute through consul- PTA’s institutions will be used to support formal PTA dis- tations. In the WTO, written consultation requests serve as pute settlement. Dispute settlement requires management evidence that such an opportunity was provided. PTAs of document exchanges and hearings; coordination of any generally require a written consultation request, and pluri- roster; secretarial, translation, and interpretation services; lateral PTAs require that the request be circulated to all provision or rental of a place to hold hearings; research and PTA parties so that they will have an opportunity to partic- drafting assistance to panelists; payment of panelist fees ipate in consultations as third parties. The consultation and expenses; information services; and capacity building. procedures may also afford a minimum consultation Negotiators may decide to have an existing secretariat take period by stipulating that the next step can be taken only on these functions. For instance, the ASEAN Secretariat after a certain number of days have elapsed after the con- provides support to the ASEAN Enhanced Dispute Settle- sultation request. ment Mechanism; the Mercosur Administrative Secretariat supports Mercosur dispute settlement; and each national section of the NAFTA Secretariat provides support for dis- The Formal Phase of a Dispute pute settlement under NAFTA Chapters 19 and 20. This During or after the consultations, the complaining party approach can foster consistency of approach and build may decide that a beneficial settlement by negotiation is not common knowledge. PTAs can also have each panelist possible. At that point, it begins to draw on the collective arrange his or her own support services on a reimbursable resources of the PTA and to seek a formal determination— basis in the event of a dispute; this approach is more eco- by submitting a request to a political body (in PTAs with nomical in the short run but can lead to uneven or legally political dispute settlement, such as the EU association inconsistent results from case to case. agreements), or by submitting legal documents as required Expenses. Where dispute settlement is conducted by the rules in a tribunal system, or by submitting a formal through a political body, each side supports its own diplo- request for establishment of an ad hoc panel. Dispute set- matic efforts. Where dispute settlement is conducted by a tlement rules that use ad hoc panels generally require that standing tribunal, the tribunal and its associated secretariat the request be in writing and that it identify the measure or will have a standing budget process that involves substan- measures at issue and the legal and factual basis for the tial contributions by the parties. The expenses and pay- complaint. ment for an ad hoc panel process typically depend on PTAs based on ad hoc panel arbitration use two whether the process is supported by a secretariat. The par- approaches toward invocation of the panel process. In the ties can set a standard scale for panelists’ fees and expenses, first, delivery of a panel request to the other party or to the eliminating fee competition between them and making PTA’s central institutional body (or both) directly triggers costs more predictable. 480 Amelia Porges A PTA that provides for an ad hoc panel process with- Whatever the language actually used in the dispute, it may out drawing on existing institutional support typically be politically necessary for texts of the decision to be made calls on the disputing parties to split the expenses. For available in all the official languages of the disputing par- instance, where a panel is selected by each party choosing ties, to ensure public acceptance. an arbitrator, with a third or presiding arbitrator chosen by agreement, each party usually pays its own expenses Panel Procedures and the fees and expenses for its arbitrator, and the par- ties divide equally the tribunal expenses (including the Predictability and consistency yield major benefits for any costs of the venue and the interpretation and secretarial litigation process, and PTA dispute settlement is no services for the hearing) and the fees and expenses of the exception. In a PTA that settles disputes through a stand- presiding arbitrator. A few agreements—for example, the ing tribunal, the tribunal has standard rules of procedure Colombia–EFTA and Canada–EFTA PTAs, and the Merco- that establish expectations regarding requirements for sur Protocol of Olivos—permit the arbitrators to appor- submission of pleadings, evidence and arguments, dead- tion costs or expenses among the parties as part of their lines, conduct of hearings, and other issues relevant for an award. In 2006, when the Mercosur Permanent Review orderly and predictable proceeding. In those PTAs that Tribunal rejected as inadmissible an interlocutory appeal settle disputes diplomatically, procedure is less important, by Argentina against the designation of a panel chairman but if a PTA relies on ad hoc panels to settle disputes, the in a case, it assessed all costs and expenses against Argentina negotiators may wish to lay down agreed rules of proce- (Mercosur 2006). dure to ensure that the panels make consistent procedural An existing secretariat, where there is one, may provide decisions. dispute settlement support from its budget. This support Many PTAs require the establishment of model rules of affords extra benefits to those who make more frequent use procedure.15 Typically, the PTA text requires the establish- of dispute settlement (but also provides public goods for ment of model rules and specifies key issues that these other PTA parties). PTAs can also budget and pay for dis- rules must cover, leaving details to be negotiated later. pute settlement separately or case by case. The ASEAN Model rules may deal with issues such as how to com- Enhanced DSM provides for ASEAN member states to ini- mence a proceeding; the number and spacing of submis- tially contribute equally to a separate revolving ASEAN sions to the panel; responsibility for administering dispute DSM Fund, to pay for the expenses of ASEAN panels, the settlement proceedings; who may attend hearings; dead- ASEAN Appellate Body, and related administrative costs of lines and places for filing documents; languages, transla- the secretariat. The parties to a dispute otherwise bear their tion, and interpretation; protection of business confiden- own legal and other expenses and can be assessed for sys- tial information; how the panel makes decisions tem expenses to replenish the DSM Fund. (consensus or vote); whether there can be separate arbitra- Languages. Negotiators may also wish to consider the tors’ opinions; and whether such separate views must be working language for dispute settlement submissions, anonymous. The rules can also provide for transparency. hearings, and decisions. They may need to preserve the Panel requests for information; handling of business confi- option to conduct disputes in their own language or lan- dential information. Some PTA disputes may concern guages, or they may wish to choose a single common lan- product bans, SPS measures, or other measures, and in guage in order to economize on costs, as translation of such cases the parties may base their arguments on asser- submissions and interpretation at hearings can substan- tions of scientific or environmental fact. Examples are the tially increase the expense and time for dispute settlement. disputes on SPS measures and on Argentina’s ban on On the one hand, NAFTA provides meticulously for the use retreaded tires that were adjudicated by Mercosur tribunals of up to three languages in dispute settlement, as do the and the SPS measure on milk imports that went before a Canada–Peru PTA and other Canadian PTAs with Latin CUSFTA panel. American countries. On the other hand, some PTAs opt for A panel faced with evaluating such issues and argu- disputes to be conducted in English (Japan–Switzerland, ments may wish to seek help. The DSU permits a panel, Korea–Singapore, Chile–China, Chile–Korea, ASEAN– without limitation, to seek information from any relevant Australia–New Zealand). The EU–CARIFORUM EPA source, and so do some PTA provisions or model rules of states that the parties are to negotiate a common working procedure. Some PTAs, such as the Trans-Pacific SEP, language but that if they cannot agree, the defending party require that any information obtained must be submitted chooses the language, and each disputing party pays its to the disputing parties for comment. Under some agree- own costs of translating submissions into that language. ments (e.g., the ASEAN–Australia–New-Zealand PTA), the Dispute Settlement 481 panel has to ask the disputants before seeking outside NAFTA panel proceedings have taken well in excess of the information; the parties have a veto over such a request; prescribed time. Defending parties are often motivated to and any information obtained has to be submitted to the delay, and even complaining parties may prefer to take time parties for comment. Some PTAs, such as the China–New to prepare. If the panelists cannot coordinate their timeta- Zealand agreement, further require that if a panel takes bles to schedule a hearing, or do not meet deadlines, there information or technical advice into account in its report, is little that the parties can do. it must also take into account any comments on that infor- mation by the parties. Appeal, Correction, and Remand Claims regarding a breach of PTA rules may involve reliance on proprietary or business confidential factual Where a trade agreement relies on ad hoc panels to settle data. Even information on government programs can be disputes, the PTA parties may seek greater assurance that business confidential in nature; for example, information the decisions of successive panels will be consistent and about Canadian government subsidies to the dairy indus- legally sound. Such assurance may be essential to secure try was excluded from a WTO panel report (WTO 1999). domestic acceptance of panel rulings. Tribunal systems Some WTO disputes (e.g., the dispute about the EU’s such as the ECJ respond to the same concerns with appel- banana import regime and the aircraft subsidies disputes) late review, or provisions for revision or interpretation of have involved information extremely sensitive to the com- past judgments. An appellate body involves incremental panies providing it. If PTAs are to be able to handle such cost to the parties, but it generates public goods in the form disputes, negotiators may wish to consider requiring pan- of enhanced certainty and uniformity in the application of elists and panels to respect the confidentiality of the data. international trade law. Formalized expert groups. NAFTA, in Article 2015, author- The WTO and some PTAs now use two methods to pre- izes panels to request a written report from a scientific review vent and correct panel error. The first method is to give the board on factual issues concerning environmental, health, parties to a dispute an opportunity to comment on a panel’s safety, or other scientific matters raised by a disputing report in draft form before the report is finalized; this inno- party; the Canada–Chile and Chile–Mexico PTAs contain vation in CUSFTA was incorporated into the WTO DSU similar provisions. These provisions have never been used. and has been adopted in many PTA ad hoc panel processes. Parallel provisions in the WTO (DSU Article 13.2 and The second method is appeal. The DSU includes an Appel- Appendix 4) also have never been used; WTO panels in late Body, which the EU required as a quid pro quo for cases involving health or safety issues have consulted indi- accepting binding dispute settlement in the WTO. The EU vidual experts, who did not draw up any group report. has instituted a two-level process, with the General Court as a court of first instance for certain types of disputes, with appeal to the ECJ. Mercosur’s 2003 Protocol of Olivos cre- Duration of Panel Process ated the Tribunal Permanente de Revisión (TPR), an The governments that have negotiated PTAs have shown a appellate tribunal (box 22.1). The Olivos Protocol reacted strong preference for speed in the panel process. For panel- to a number of difficult disputes and responded in part to based dispute settlement procedures that call for initial and smaller states’ concerns about compliance with Mercosur final panel reports, PTAs’ notional deadlines for a panel to law (Whitelaw 2004). The TPR reflects a high level of insti- produce its initial report vary widely: 30 days in tution-building ambition and helps enforce and build not SACU–Mercosur; 60 days in Mercosur and the ASEAN only treaty law but also the decisional law being created by Enhanced DSM; 90 days in NAFTA and the Canada–Peru Mercosur institutions (Opertti 2004). PTA; 120 days in the Chile–U.S. PTA; and 180 days in the The ASEAN Protocol on Enhanced Dispute Settlement U.S.–Korea and Chile–Australia agreements. These PTAs Mechanism of 2004 calls for the ASEAN Economic Minis- typically allow two weeks for comments from the parties ters to establish a seven-person Appellate Body patterned and then ask the panel to finalize its report within 30 to 45 on the WTO Appellate Body. As of early 2011, this had not days after the initial report. yet taken place. Stakeholders favor short deadlines. The consultations and other preliminary phases of a dispute, including panel Labor, Environment, Financial Services, selection, can take considerable time, and business stake- and Other Special Sectors holders are almost never compensated for damage caused them by the breach of trade obligations. In practice, these As noted in earlier chapters of this volume, many PTAs pro- deadlines may be more aspirational than real. So far, vide special dispute settlement procedures for particular 482 Amelia Porges Box 22.1. The Protocol of Olivos The Mercosur Protocol of Olivos established a new five-person Permanent Appeals Tribunal (Tribunal Permanente de Revisión, TPR). It consists of one arbitrator (with an alternate) appointed for a renewable two-year term by each of the four Mercosur members, plus a fifth arbitrator appointed for one three-year term by agreement between the members. The TPR’s permanent headquarters is in Asunción. Disputes between Mercosur members are normally first heard by an ad hoc arbitral tribunal, which provides its award within 60 days (extendable to a total of 90 days). Within 15 days after the award, any party to the dispute may bring a motion for review to the TPR. The TPR may consider only the legal issues dealt with in the dispute and the legal interpretations in the award, not issues of fact. It renders its decision within 30 days. If the dispute involves two members, it is heard by three judges, and if more, by five judges. (Whitelaw 2004). The TPR has heard two appeals. In one, it reversed the panel decision, and in the other, it ruled the request inadmissible. Under the Protocol of Olivos, parties to a dispute may also, by agreement, submit their dispute directly and in a single instance to the TPR; in that case, the TPR acts like a panel and issues a nonappealable award in 60 days, which may be extended by 30 days. The TPR also may rule on appeals by disputants regarding failure to comply with tribunal awards, or regarding the extent of suspension of concessions after a failure to comply, and it may provide opinions to the Mercosur Common Market Group on request. subject matter. Special procedures may refer particular dis- safety, child labor, or minimum wage standards that putes to a political process for settlement, but even in the are covered by mutually recognized labor laws (a failure context of binding third-party dispute settlement, the par- that violates the NAALC), or to effectively enforce ties may wish to bring to bear special expertise, to provide domestic environmental law relating to production of special opportunities to participate, to provide particular traded goods or services (a breach of the NAAEC). The forms of redress, or to limit the possible scope of sanctions panelists must be from a specialized labor or environ- by a prevailing party. The key examples are labor, environ- ment roster. If the panel agrees that such an enforcement ment, and financial services. failure exists, and if no remedial action is taken, the Labor and environment. Chapters 19 and 20 of this vol- panel may ultimately set a monetary assessment, to ume describe the substantive obligations, including dispute be paid by the losing party into a fund to promote settlement, in those PTAs that deal with environment and (depending on the sector) labor law enforcement or labor issues. Such obligations exist in all U.S. PTAs con- improvement of the environment or of environmental cluded since NAFTA, in some Canadian PTAs, in New enforcement in the party complained against. The Zealand’s PTA with Thailand, in the Trans-Pacific SEP, in Canada–Chile PTA contains similar provisions (Bartels the EU–CARIFORUM EPA, and in the chapter on trade 2009). The side agreements also provide for citizen sub- and sustainable development of the EU–Korea agreement. missions claiming that a party is not enforcing its own As Bartels (2009) observes, the treatment of these laws. The NAFTA governments received 37 submissions issues in North American PTAs has changed over time. on labor issues in 1994–2010, of which 24 concerned These dispute settlement provisions can be grouped into conditions in Mexico, 13 pertained to the United States, four categories. and 2 were against Canada (U.S. DOL 2010). The NAFTA Commission on Environmental Cooperation 1. The side agreements to NAFTA—the North American received 77 citizen submissions from 1995 through Agreement on Labor Cooperation (NAALC) and the early 2011; of these, 39 focused on Mexico, 9 on the North American Agreement on Environmental Cooper- United States, 27 on Canada, and 1 on a cross-border ation (NAAEC)—represent one model. They establish Canada–U.S. issue (CEC 2011). The Canada–Costa Rica institutional frameworks for cooperation, require that FTA also provides for citizen submissions, as well as domestic legislation adhere to high standards, and obli- government-government dispute settlement without gate each party to enforce and strive to improve its own fines or trade measures (Bartels 2009). legislation. Panel proceedings require a two-thirds vote 2. Under the Jordan–U.S. agreement, parties must not fail, of the three NAFTA parties. They are to be preceded by through a sustained or recurring course of action or consultations and a report by a committee of experts, inaction, to effectively enforce domestic environmental and they are only available in cases of persistent, trade- or labor laws in a manner affecting trade between the related patterns of failure to effectively enforce worker parties. These labor and environmental obligations are Dispute Settlement 483 enforceable through the same procedures as the rest of Trans-Pacific SEP. In both, the labor obligations are only the agreement, but a side letter provides that they will enforced through consultations. Dispute settlement provi- not be enforced so as to block trade (Bartels 2009). sions in the environment agreements provide for referring 3. Seven other U.S. PTAs negotiated in 2000–07, with a dispute to the interested parties for a report. Those Australia, Bahrain, CAFTA–DR, Chile, Morocco, Oman, affected are obligated to implement the report’s recom- and Singapore, adopt the same standard as in the mendations (Bartels 2009). Jordan agreement. They also provide that parties may Historically, most EU PTAs have treated labor and envi- not weaken, reduce, or waive environmental or labor ronment as “matters for cooperation� (Bartels 2009). How- protections to encourage trade or investment. These ever, the EU’s Global Europe policy, announced in late obligations are enforceable through the same dispute 2006, calls for environment and labor to be part of the settlement procedures as the rest of the agreement, but trade agreement negotiations. For instance, the chapter on no trade retaliation is possible—only fines capped at trade and sustainable development in the EU–Korea FTA US$15 million (adjusted for inflation). Fines are col- calls for high levels of protection in both areas and obli- lected through suspension of concessions, if necessary, gates both parties not to fail to effectively enforce environ- and are spent on labor or environmental initiatives in mental and labor laws “through a sustained or recurring the territory of the party complained against. These course of action or inaction� and not to waive or derogate agreements also create cooperative institutions and pro- from its environmental or labor laws, regulations, or stan- vide for receipt of input from the public. dards in a manner affecting trade or investment between 4. In May 2007 the Bush administration agreed with the the parties. Enforcement, however, consists exclusively of leadership of Congress (then controlled by Democrats) bilateral consultations and an advisory report by a group of on elements that future free trade agreements must have experts. in order to be considered by Congress. These elements In the EU–CARIFORUM EPA, the labor and environ- include increased substantive standards for labor rights ment obligations in Chapters 4 and 5 of Title IV are provisions and an enforcement standard like that in the enforceable through the agreement’s regular dispute set- Jordan agreement. The provisions are subject to the tlement procedures, but only after separate procedures in same dispute settlement mechanisms and penalties as in the labor or environment chapters are pursued for at least other provisions, but dispute settlement must be pre- nine months. These procedures call for bilateral consulta- ceded by consultations in a specialized labor or environ- tions, possibly including advice from relevant international ment council. The Peru–U.S. FTA and pending FTAs environmental bodies or the International Labour Organi- with Colombia, Korea, and Panama all follow this zation (ILO), and may also include a report by a three- pattern (Bolle 2009). member committee of experts. In any ensuing dispute, at least two of the three arbitrators must have specific expert- On July 30, 2010, the United States initiated the first ise on the subject matter and must be drawn from a special formal PTA labor complaint, under the labor chapter of roster. The panel report must recommend how to ensure CAFTA–DR. In a letter signed by U.S. Trade Representa- compliance with the EPA’s trade or environment obliga- tive Ron Kirk and Secretary of Labor Hilda Solis, the tions, and measures taken in case of noncompliance may United States formally requested consultations with not include suspension of EPA trade concessions. Sepa- the Guatemalan government regarding (a) failure by the rately, the agreement’s investment chapter (Article 72) Guatemalan Ministry of Labor to investigate alleged labor obligates the parties to ensure that investors act in accor- law violations or to take enforcement action when the min- dance with core labor standards and do not behave in a istry had identified a violation, and (b) failure by the manner that circumvents international environmental or Guatemalan courts to enforce Labor Court orders in cases labor obligations in agreements to which the EU and the involving labor law violations. The complaint expressed CARIFORUM states are parties. This obligation is subject concerns about the Guatemalan government’s alleged fail- to the agreement’s regular dispute settlement procedures. ure to protect those attempting to exercise labor rights Financial services. As Stephanou (2009) observes, PTA against violence and threats. It followed a submission by parties have been particularly cautious about covering U.S. and Guatemalan unions and informal consultations financial services because of regulatory sensitivities and that had been going on since January 2009 (Kirk and Solis strategic considerations. Countries that follow the GATS 2010; USTR 2010). approach for services in their PTAs have used it for finan- New Zealand has negotiated labor and environment cial services as well, with special adaptations or a sepa- side agreements to the Thailand–New Zealand FTA and the rate chapter expanding on the GATS Annex on Financial 484 Amelia Porges Services. Others follow the NAFTA model, in which invest- power to make binding decisions that matter, the tribunal ment to provide services is treated like any other type of must be composed of people who are perceived as trust- investment and has access to investor-state dispute settle- worthy, impartial, and confidence-inspiring. Negotiators ment. Stephanou (2009) points out that an increasing solve this problem by specifying criteria for decision number of countries prefer a dedicated chapter on finan- makers such as neutrality, geographic distribution, and cial services because it facilitates customizing the applica- professional expertise. They may also try to expedite tion of services disciplines in this area and allows financial panel formation by agreeing in advance on a roster of sector policy makers to control negotiation and implemen- persons who meet the criteria. tation of the obligations. Standing tribunals. In a standing tribunal, judges are Governments have demonstrated in their PTAs that selected on the basis of certain criteria and some form of they want special procedures for settling financial services geographic distribution. The governments bound will need disputes. As in the WTO, PTA provisions often require that to agree on the judges’ terms of service, their payment, and panels for such disputes have specific regulatory or other the funding of the necessary infrastructure. The original financial sector expertise, and they prohibit trade retalia- model, the International Court of Justice (ICJ), consists of tion against financial services in any dispute in any other 15 judges elected to nine-year terms by the United Nations sector. In NAFTA-model PTAs that have a separate finan- (UN) General Assembly and the UN Security Council. The cial services chapter, that chapter limits the scope of Statute of the International Court of Justice requires that investor-state dispute settlement to claims regarding its judges be persons of high moral character who possess expropriation or transfer of payments. National treatment the qualifications required in their countries for appoint- claims, for instance, are only subject to state-state dispute ment to the highest judicial office or who are lawyers of settlement, giving a host state additional flexibility to dis- recognized competence in international law. Each judge criminate in favor of domestic financial services providers participates in all cases. The Statute also requires that in the if its PTA partners agree (Sauvé and Molinuevo 2008). ICJ as a whole “the representation of the main forms of civ- NAFTA-model financial services chapters also refer claims ilization and of the principal legal systems of the world regarding prudential or monetary or exchange-rate meas- should be assured.� If a state party to an ICJ case does not ures for a consensus decision by a financial services com- have a judge of its nationality on the Court, it may appoint mittee composed of financial services regulators; if the an ad hoc judge for that case only. committee agrees, its decision binds the tribunal. The European Court of Justice has one judge for each Other sectors. PTA negotiators can provide specialized of the 27 member states and normally hears cases in pan- dispute settlement for any subject they wish, or they can els of 3, 5, or 13. The judges must have the qualifications require that arbitrators for disputes on technical subjects or competence needed for appointment to the highest have specialized expertise. For instance, the EU–Chile FTA judicial positions in their home countries. Both the ICJ contains a chapter on trade in wine that includes provi- and the ECJ are backed by an elaborate and expensive sions on regulation of labeling and oenological practices. infrastructure of buildings, legal and support staff, The chapter requires the parties to establish a roster of libraries, and translators. The four judges of the Andean arbitrators with oenological expertise for disputes regard- Tribunal (one per member state) also serve full time. ing obligations under the agreement. Other standing tribunals, for ECOWAS, COMESA, and the EAC, are composed of member-state judges who serve part-time (Banjo 2007). Participation in PTA Dispute Settlement The WTO Appellate Body consists of seven part-time and Enforcement members, appointed by consensus for a renewable term of This section reviews participation in PTA dispute settle- four years, who must be “persons of recognized authority, ment, including qualification and selection of decision with demonstrated expertise in law, international trade makers, third-party participation, use of experts, and par- and the subject matter of the covered agreements gener- ticipation by civil society in the dispute settlement process. ally� (DSU Article 17.3). They are geographically balanced through informal agreement. Members have included public international lawyers, trade lawyers, and diplomats Qualifications and Selection of Decision Makers with no formal legal training. A division of three members, Selection of decision makers is a key issue for international determined randomly, considers each case. The Appellate dispute settlement; if governments are to give a third-party Body has a small staff and shares in the infrastructure decision maker (a standing tribunal or an ad hoc panel) provided by the WTO. Dispute Settlement 485 Ad hoc panels. The DSU requires that WTO dispute set- fallback mechanisms. But the existence of a fallback mech- tlement panels be composed of impartial, “well-qualified anism does provide an incentive to reach agreement. governmental and/or non-governmental individuals� and Panel selection can take considerable time. The panel lists types of acceptable experience (WTO litigation; serv- selection process has been a significant obstacle for NAFTA ice as a delegate, or in the WTO Secretariat, or as a senior disputes, according to Gantz (2006), who observes that the trade policy official; teaching or publishing on trade law NAFTA dispute on cross-border trucking services was or policy). Many PTA dispute settlement mechanisms delayed for 15 months by panel formation and that in the modeled on the DSU specify similar characteristics of NAFTA sugar dispute, the U.S. authorities declined for impartiality and expertise. Some, such as the China–New more than four years to appoint panelists. Zealand PTA, also require that the panelists comply with the WTO DSU’s rules of conduct for panelists. NAFTA and Rosters other U.S. FTAs require compliance with a code of conduct to be established by the parties. A roster can speed panel selection by providing a preap- Another issue is citizenship or nationality. In the WTO, proved list of persons who are qualified and willing to citizens of parties or third parties to a dispute cannot serve serve. Since 1907, arbitral institutions, which now include on the panel, unless the parties agree otherwise. The most the International Centre for Settlement of Investment Dis- common model in PTAs is for each party in a bilateral dis- putes (ICSID), the Permanent Court of Arbitration, and pute to select an arbitrator, who can be (and usually is) a many others, have maintained rosters of arbitrators to national of that party. The parties to the dispute then select facilitate dispute resolution (Schreuer 2001). In the WTO, a neutral chair by agreement, with a fallback to selection by DSU Article 8.4 provides for the maintenance of a roster of lot or by an appointing authority, as discussed below. governmental and nongovernmental individuals qualified to serve on panels (WTO 1996b, 724). For special obliga- tions in agreements (on labor, environment, financial Ad Hoc Panels and the Blockage Problem services, or other designated topics), PTAs may require Ad hoc panels, as bodies created to settle a particular dis- specialized rosters to ensure that the decision makers in a pute, present particular moral hazard problems that the dispute have the necessary expertise. text of a PTA can and should anticipate and prevent. The A roster of panelists can be open or closed. The WTO central problem arises when a defending party refuses to has an open roster; disputing parties may choose any pan- cooperate with the process by declining to name its arbitra- elists they wish. The WTO’s DSB accepts practically all tors or to cooperate with the panel selection process. names nominated by members. At the other extreme is a The WTO’s DSU procedures have successfully overcome closed roster, like that provided for in the EU–Chile FTA. this problem. The WTO Secretariat takes the initiative to As required by that agreement, the EU–Chile Association nominate panelists to the parties. If a panel has not been Committee has established a list of 15 persons. Of these, completed by agreement within 20 days from the decision five are Chilean, five are from the EU, and five, of neither by the Dispute Settlement Body (DSB) to establish the nationality, are identified as possible chairpersons (Euro- panel, then, if either party so requests, the WTO Director- pean Commission 2007). The FTA provides that within General can and will select the missing panelists, in consul- three days of any panel request, the chair of the Association tation with the WTO’s political leadership. Committee must select by lot two arbitrators, one each Some PTAs deal with the failure-to-appoint situation by from the EU and Chilean sublists, as well as a chair from designating an appointing authority, such as the WTO the list of individuals identified as potential chairs. Panel Director-General, the Secretary-General of the Permanent selection takes days instead of months. (The initial list was Court of Arbitration in the Hague, or a regional secre- not finalized until two years after the FTA entered into tariat.16 Other PTAs provide for the missing arbitrator or force, however.) Similarly, panelists for ad hoc tribunals for neutral to be selected by lot; this is the method chosen by Mercosur disputes are selected from sections of a Mercosur the Japan–Mexico, Japan–Chile, and Japan–Thailand PTAs roster; if a disputant fails to appoint its arbitrator, or if and by NAFTA and other U.S. FTAs. If negotiators choose there is no agreement on the presiding arbitrator, the arbi- selection by lot, the selection should be conducted by a trator in question is appointed by Mercosur’s Administra- body that is not controlled by the defending party. tive Secretariat from the roster. In practice, governments may wish a PTA panel, and its The PTA rules can also favor panel selection from the decision, to have the legitimacy that flows from consent. roster—for instance, by barring any veto of panelists selected They may therefore be reluctant to actually use these from the list—without excluding nonroster candidates. If 486 Amelia Porges a roster is used, it is advisable that it remain valid until it is GATT (WTO 1996a) and the WTO. Rules for multiparty replaced. NAFTA provides for each party to select panelists disputes also exist in, for example, NAFTA, CAFTA–DR, from the roster who are not its citizens (“cross-selection�) the Trans-Pacific SEP, and PTAs involving EFTA. but this innovation has not been picked up in any other NAFTA permits a third party to join as a complaining PTA, including later U.S. FTAs. party as of right, if it delivers a timely notice to the disput- Some other PTAs, such as the ASEAN Enhanced ing parties and the NAFTA Secretariat. The third party can DSM, provide for indicative lists modeled on that of the then participate in the consultations, but if it fails to join WTO. Many do not use a roster for panel selection at all, the dispute, it is “normally� precluded from initiating or however, permitting governments to select the candi- continuing a NAFTA or WTO dispute on the same matter, dates that they prefer when the dispute arises. in the absence of a significant change in circumstances. The NAFTA Commission must consolidate two or more pend- ing NAFTA disputes regarding the same measure and may Panelist Ethics consolidate other cases that are appropriately considered PTA provisions on the qualifications of panelists often jointly. Similar provisions exist in the CAFTA–DR and include qualitative requirements designed to ensure impar- Chile–Central America PTAs. The Trans-Pacific SEP tiality. In December 1996 the WTO’s DSB adopted a code Agreement requires consolidation of disputes on the same of conduct for panelists (WTO 1996b) that is also applied measure, and the Colombia–EFTA, China–ASEAN, and by the WTO Appellate Body. This code requires that pan- ASEAN–Australia–New Zealand PTAs express preference elists be independent and impartial, avoid direct and for consolidation or use of common arbitrators. Where indirect conflicts of interest, and respect the confidentiality there are multiple complaining parties, the panel selection of WTO proceedings. The disclosures required by this procedure may be adjusted accordingly, as is done in code have become a routine aspect of panel disputes in NAFTA and the CAFTA–DR, Chile–Central America, and the WTO. Mercosur agreements. Influenced by the WTO example and by the require- Most plurilateral PTAs permit another PTA member to ment for a code of conduct in NAFTA, a number of PTAs participate in dispute settlement consultations as of right; call for panelists to adhere to a code of conduct to be some make such participation conditional on consent by agreed by the parties. These PTAs include the Singapore– the disputing parties. Many also permit another PTA mem- GCC, Mexico–Northern Triangle, Chile–Colombia, Chile– ber to attend panel hearings, to make written and oral sub- Australia, Canada–Peru, and Chile–EU agreements; NAFTA; missions to the panel, and to receive some or all written and all later U.S. PTAs. In addition, following the example submissions of the disputing parties (as in WTO panel pro- of Article 18.1 of the DSU, PTA model rules of procedure ceedings).17 Broader participation is not cost-free, of or codes of conduct may prohibit ex parte contact with course. It involves costs for the third party or extra com- panelists, as do the ASEAN Enhanced DSM and the Singa- plainant and for the defending party, and it decreases the pore–GCC PTA. likelihood of early settlement (Busch and Reinhardt 2003). Plurilateral PTAs and Multiparty Disputes, Joinder and Transparency and Civil Society Participation Consolidation, and Other Third-Party Participation One of the parties to a PTA negotiation may seek to have When a dispute arises in a plurilateral PTA, another party the PTA dispute settlement provisions incorporate ele- to the PTA may wish to join the complaint or merely to ments of procedural transparency. WTO practice has been observe and submit views. The reasons for wanting to do so evolving since 1995 in the direction of increased trans- may include commercial competition in the markets con- parency in dispute settlement, but views are still divided cerned, a desire to prevent discriminatory settlements, an within the WTO as to whether to mandate such trans- interest in the interpretation of common PTA rules such as parency. PTA practice varies widely, but it too has evolved rules on market access or rules of origin, and concern for substantially in recent years. Some U.S. PTAs now mandate endeavors such as regional public goods. Third-party par- considerably more dispute settlement transparency than ticipation can also be a useful way for developing countries the WTO. to build capacity in dispute settlement. Similar motives The WTO’s dispute settlement rules largely codify the have built a rich practice of joinder and consolidation of informal, diplomatic, nontransparent practice under the disputes, multiparty cases (as in the EU banana import GATT, which treated dispute settlement as a private negoti- regime dispute), and third-party participation, under the ating process between the parties. The DSU provided that Dispute Settlement 487 disputing parties could make their submissions and non- Complaints by stakeholders against governments. Articles confidential summaries public but made no other changes. 39–44 of the Mercosur Protocol of Olivos permit individu- Since 1998, the United States has advocated that all panel als or juridical persons to bring complaints against any submissions be available to the public; that panel hearings Mercosur party for applying legal or administrative meas- be open to public observation, except where there is a need ures that have a restrictive, discriminatory, or unfairly to protect confidential business information; and that pan- competitive effect, in violation of the Treaty of Asunción or els accept amicus curiae submissions from the public or the associated agreements and legal instruments. Such from nongovernmental organizations (NGOs). These pro- complaints must be filed with the national section of the posals have gained at least partial support from the EU, Common Market Group of the country in which the Canada, and others, to help strengthen public support for claimant resides or has its headquarters. If the national sec- trade liberalization. The United States has made all of its tion supports the complaint, it must negotiate directly for dispute settlement submissions since 1995 public; other 15 days to resolve the matter with the defending country’s members have made some of their submissions public; national section. If the complaint is not resolved, it must be amicus briefs have been submitted and even considered; brought to the Common Market Group. Unless the Com- and in a number of WTO disputes, at the request of the mon Market Group rejects the complaint summarily by parties, the panel or Appellate Body hearing has been open consensus, it must convene a group of three neutral for public observation via closed-circuit television. experts, who are to report in 30 days. If the panel unani- Not all governments agree. As described by Mercurio mously agrees with the complaint, any Mercosur party can and LaForgia (2005), some WTO delegations have argued demand that corrective measures be adopted, or that the that openness would undermine the character of the WTO dispute measure be annulled, and can then proceed to as a forum for confidential discussion between govern- Mercosur state-state dispute settlement if the defending ments, that it would burden members’ ability to participate state does not comply within 15 days. If the panel’s report effectively in disputes, and that allowing observers at hear- is not unanimous, the claim terminates. Any Mercosur ings would lead to trials by media. Although almost all party can bring the same complaint. WTO members strongly criticized the Appellate Body for Amicus curiae submissions. Civil society groups have accepting amicus submissions in the asbestos dispute in argued that panels should consider not just the arguments 2000, later panels and the Appellate Body have continued to presented by governments but also facts and arguments accept such submissions. Many developing countries con- presented by others, even those who may be at odds with tinue to oppose amicus submissions as a resource burden. the governments participating in dispute settlement. If dis- Direct participation by stakeholders and civil society. No pute settlement provisions are neutral regarding amicus government opposes all participation by civil society in submissions, presumably a panel could consider them. dispute settlement. On the contrary, governments usually Some PTAs go further and explicitly favor consideration welcome stakeholder input and guidance on facts, com- of amicus submissions. U.S. FTAs after NAFTA have mercial data, and negotiating priorities. WTO studies required that panels consider requests from nongovern- (Tussie and Delich 2005; Xuto 2005) illustrate the critical mental entities in the parties’ territories to provide written role played by developing-country stakeholders in dis- views that may assist the panel in evaluating the submis- putes. Governments also have put in place organized struc- sions and arguments of the parties. A number of other tures for receiving stakeholder complaints.18 Moreover, recent PTAs explicitly authorize the panel to accept and PTA provisions on labor and environment have provided consider amicus submissions under certain conditions. for receipt of public submissions on these issues; the first For instance, the Trans-Pacific SEP and the EU–Chile, PTA labor dispute ever, initiated by the United States EU–Korea, Chile–Panama, Canada–Colombia, and against Guatemala in July 2010, followed an April 2008 Canada–Peru PTAs all permit amicus submissions, but public submission by the AFL-CIO labor group in the only if timely (submitted within 10 days of panel composi- United States and by Guatemalan unions, under the labor tion), concise (not over 15 typed pages, including annexes), chapter of the CAFTA–DR PTA (USTR 2010). and directly relevant to the factual and legal issues before Practically all governments involved in trade disputes the panel. The submission must also describe the submit- now expect stakeholders to hire or pay for legal counsel to ter, its activities, its source of funding, and the nature of its assist the government. A few PTAs (Canada–Israel, interest in the proceeding. The Canada–Peru PTA adds EU–Chile, Canada–Chile, Chile–Central America) explic- more requirements: in deciding whether to permit an itly guarantee the right of parties to be assisted by counsel NGO to make a submission, a panel must consider, among and impose conditions regarding the counsel’s conduct. other matters, whether there is a public interest in the 488 Amelia Porges proceeding, whether the NGO has a substantial interest in PTA institutions can provide a cost-effective channel for the proceeding, and whether the written submission would gathering information. Governments, and particularly add a perspective, particular knowledge, or insight that is those of resource-poor developing countries, normally do different from that of the parties. (A footnote clarifies that not know the details of a PTA partner’s trade or regulatory an interest in jurisprudence, in the interpretation of the regime and may not have access to foreign legislation. Insti- PTA, or in the subject matter of the dispute is not enough tutional contacts provide a way to obtain those facts. Coop- to establish an NGO’s substantial interest.) An NGO sub- eration within PTA institutions, and the regular contact that mission may not introduce new issues to the dispute, must it implies, offer a means of building mutual confidence at stay within the dispute’s terms of reference, and must avoid the personal level, resolving routine trade irritants and disrupting the proceeding. It must preserve the equality of minor issues, and negotiating further trade liberalization. the disputing parties, who must have an opportunity to Asia-Pacific Economic Cooperation (APEC), for instance, respond. has provided useful settings for dialogue and regulatory Transparency in the dispute settlement process. All U.S. harmonization on customs and trade facilitation, chemical PTAs after NAFTA have required that any dispute settle- regulation, SPS measures, and automotive standards. ment submission (including written submissions, texts of oral statements, and other documents) must be made Specialized Institutions public within 10 days. Other PTAs authorize, but do not require, their parties to make their own submissions PTA institutions can be particularly helpful for regional public; these include the Canada–Colombia, Canada–Peru, integration in complex issue areas or regulated industries. ASEAN–Australia–New Zealand, and Japan–ASEAN PTAs may establish specialized committees composed of agreements. regulators from each country. Contacts between regulators All U.S. PTAs after NAFTA have required that hearings build trust in each other’s judgments, facilitating market in disputes are to be open to the public; the recent access for regulated products or services. Some agreements Canada–Colombia, Canada–Peru, and EU–Korea PTAs provide an incentive to consult about issues in these tech- also require open hearings. The 2009 Japan–Switzerland nical committees by providing that such discussions will PTA provides for open hearings unless either disputing satisfy the usual requirement for consultations before party objects. In some other PTAs (the Trans-Pacific SEP recourse to a dispute settlement panel. The CAFTA–DR and the EU–Chile agreement), hearings are closed unless agreement, for instance, establishes committees on agricul- the disputing parties agree otherwise. tural trade (Article 3.19), market access for goods, rules of origin, and customs issues (Article 3.30), SPS measures (Article 6.3), TBTs (Article 7.8), and financial services Alternative Enforcement and Dispute (Article 12.16). Each committee is able to consider routine Settlement Methods market access issues. Consultations that have taken place in Formal disputes take place only in exceptional cases; the the TBT Committee may substitute for the first step in the universe of large and small trade disputes in a PTA is much dispute settlement process. The Chile–EU FTA establishes larger. Statistics on dispute settlement, which record formal special committees on rules of origin (Article 81); stan- disputes, systematically understate the total number of dis- dards, technical regulations, and conformity assessment putes and the proportion of disputes that is settled. (Article 88); and financial services (Article 128), as well as Most PTAs set up an institution or institutions to main- special consultation processes on SPS requirements, trade tain the PTA, to address practical problems of PTA imple- in wine, and trade in spirits; these can substitute for the mentation and market access, and to provide a framework consultation stage of dispute settlement proceedings for further negotiations. These institutions may include (Articles 89, 129, Annex 5). joint committees or councils at the ministerial or senior official level to oversee the operation of the PTA. As Collective Enforcement Procedures Donaldson and Lester (2009) note, most such committees meet annually (and more often, as required), and a few The adversary process used in panel-based dispute settle- meet biennially. They may be ad hoc or standing com- ment procedures, and triggered only if one PTA party mittees at the working level, or even joint public-private brings a formal complaint against another, is not the only working groups. Maintaining these institutions requires a possible structure for enforcement of PTA obligations. In commitment of resources and personnel but can be eco- the EU, for instance, the predominant method of enforce- nomical and effective in resolving problems. ment is through action by the supranational European Dispute Settlement 489 Commission, which actively monitors member states’ “parties to a dispute.� The implication is that a dispute actions and can bring enforcement actions before the EU’s exists, but Article 5 has never been used during a dispute. courts even in the absence of any stakeholder or member The only known WTO mediation, in 2002 on European state complaint. The secretary-general of the Andean Community tariff preferences for canned tuna, was used in Community has similar enforcement powers, as do the sec- place of dispute settlement proceedings. retaries-general of COMESA and the EAC. Enforcement of Good offices may include using shuttle diplomacy, this type is typically coupled with a standing tribunal restoring contact between the parties, inviting them to empowered to adjudicate cases and impose sanctions. meetings, or offering suggestions for settlement. In public Where PTA members have pooled their sovereignty, as in international law, good offices can also include supervision the EU, such enforcement is strong, but where the mem- of plebiscites or armistices. The UN Secretary-General, the bers are less ambitious, it is not (Alter and Helfer 2011). Swiss government, and the Holy See have provided good As another example, ASEAN formally agreed in 2004 to offices in international conflicts. When an organization is establish an ASEAN Compliance Body (ACB), modeled on called on to provide good offices, a senior official is often the WTO Textiles Monitoring Body. The optional compli- appointed to take charge of the issue. The Director-General ance procedures (Yoshimatsu 2006) provide for group peer of the GATT, through a designated representative, pro- review of measures on a 90-day timeline. The ACB’s find- vided good offices in a number of instances, notably in the ings, drawn up by countries not party to the dispute, would GATT dispute between Latin American countries and the not be binding but could serve as inputs for formal dispute EU regarding banana trade (WTO 1996a, 765–67). Gathii settlement. However, the ACB does not appear to be opera- (2010) reports a recent dispute in COMESA over the Kenya tional at present. Sugar Board’s nontransparent auctioning of import True multilateral sanctioning mechanisms (such as the licenses for sugar, which affected import trade from UN Security Council) are quite rare, but collective persua- COMESA members. In September 2009 COMESA sent to sion mechanisms exist, including the specialized institu- Kenya a Sugar Sector Safeguard Assessment Mission, which tions described in the preceding subsection. The Montreal recommended elimination of auctioning as a nontariff Protocol on the Protection of the Ozone Layer has such a barrier but apparently did not characterize the auctioning mechanism in its Implementation Committee of 10 treaty as a treaty violation. The intervention successfully per- parties, which regularly examines the treaty parties’ com- suaded Kenya to eliminate the auctioning. pliance reports, can receive other parties’ submissions, and In mediation, the neutral party actively proposes can investigate and provide reports and recommendations options for settlement. The success of mediation depends to the Meeting of the Parties. But accounts indicate that it very much on the mediator and on the willingness of the is ineffective at stopping even intentional and continuing parties to make concessions (Merrills 1991, 29, 39–41). violations (Yang 2007). There are few widely reported cases of such mediation within a PTA. An example involving the EU is the 2002 WTO tuna mediation. Its success is attributable to a num- Alternative Dispute Resolution: Good Offices, ber of factors: the complainants’ ability to use leverage (a Mediation, and Conciliation threat to block an EU waiver and indirectly prevent the Good offices, mediation, and conciliation all are tradi- launch of the Doha Round) as a means of obtaining the tional, widely recognized means of dispute settlement in EU’s political commitment to the process; skilled media- public international law, as well as means of alternative dis- tion by a veteran, neutral dealmaker who suggested a prac- pute resolution (ADR) used in commercial situations. In tical solution; EU goodwill in promptly implementing a each case, the parties cooperate voluntarily with a neutral, solution that increased the complainants’ market access; uninvolved third party. The third party, which, in a public and the fact that the problem was framed not in terms of international law context, could be an individual, an inter- legal rights but as a question of impairment of interests national organization, or a state, offers to assist the parties (Porges 2003). in settling their differences by negotiation and agreement, In conciliation, the parties set up a permanent or ad hoc without arbitrating the merits of the legal claims con- commission, or refer a dispute to a conciliator that is to cerned. The three types of ADR overlap but are distin- impartially examine the dispute and suggest an acceptable guished by the extent and nature of the neutral party’s settlement. Historically, conciliation has worked best when involvement. the main issues in a dispute are legal but the parties wish to Article 5 of the WTO DSU provides for the possibility of settle (Merrills 1991, 77). For instance, conciliation was used good offices, mediation, or conciliation, if agreed to by the to wind up the defunct original East African Community in 490 Amelia Porges the 1980s. In response to a request by Kenya, Tanzania, and compliance, through negotiations or an arbitration process, Uganda, a Swiss diplomat, Victor Umbricht, acted as con- with a 15-month benchmark for the compliance deadline. ciliator to locate and value the EAC’s assets and liabilities, During the period before the deadline, the complying party propose a formula for dividing them among the member must report on its actions, and after the deadline elapses, if states, and assist in an agreement winding up the organiza- compliance has not occurred, the Dispute Settlement Body tion (Merrills 1991, 65–72). will authorize the complainant to suspend concessions or A number of WTO-model dispute settlement systems in other obligations under the WTO, in an amount equivalent PTAs provide for these alternative dispute resolution to the nullification or impairment (trade damage) caused mechanisms—because of a preference for consensual set- by the breach of the rules. If there is a dispute regarding tlements, because negotiators view ADR as potentially compliance, it must be settled under the DSU by recourse fruitful and as free or low cost, or as a reaction against the (whenever possible) to the original panel; a dispute regard- time the WTO takes to determine rights and obtain com- ing the amount of the suspension of concessions must be pliance.19 For instance, Annex 14-A of the EU–Korea FTA arbitrated (whenever possible) by the original panel. sets out a mediation procedure for nontariff measures PTAs’ approaches toward compliance vary. PTAs based affecting industrial products. The parties can appoint a sin- on diplomatic or political settlement of disputes rely on gle mediator who is a subject matter expert and who would settlement by agreement, not sanctions, and a failure to use an informal, confidential process to help the parties comply with an agreement settling a case simply means the clarify the problem and its trade effects and reach a mutu- start of another negotiation. In PTAs with standing tri- ally agreed solution. The process is agnostic regarding rule bunals, the tribunal remains available to enforce its judg- violations. The mediation process, and the positions taken ments on a continuing basis, and a strong tribunal may in the process, are confidential and would not be admissi- have the power to impose sanctions on a noncomplying ble in dispute settlement. The EU–CARIFORUM EPA sim- government. When the European Commission considers ilarly provides for nonbinding, confidential mediation by that an EU member state has not complied with a judg- agreement between the parties to a dispute and for keeping ment by the ECJ, under Article 260 of the Treaty on the a roster of mediators. Functioning of the EU, the Commission may issue a rea- As the COMESA example discussed above illustrates, soned opinion and then bring the case before the ECJ, ADR can serve as a useful channel through which the other specifying an appropriate fine; the Court may impose a members of a PTA can focus efforts in a practical, time- fine on the member state as a lump sum or on a continuing effective way on removing a PTA member’s illegal trade basis during the period of noncompliance. In the Andean barriers; the benefits of providing for ADR in a PTA are Community, Article 27 of the 1996 treaty establishing the speculative but are very likely to exceed any cost. Formal Andean Tribunal requires a member state to comply within ADR mechanisms may also offer a pathway for less formal 90 days with any tribunal judgment of noncompliance diplomatic settlement. Further empirical research would be with its Andean Community obligations, and authorizes useful to determine the extent to which the parties to existing the tribunal to determine suspension of benefits against a PTAs use ADR in practice, the disputes they use it for, the noncomplying state by the claimant state or any other solutions produced by ADR, and factors leading to the success member; the tribunal may decree other measures if such a or failure of ADR. The use of ADR may well be underreported suspension would be ineffective or would make the situa- in the existing literature because it may not be characterized tion worse. Under Article 30 of the same treaty, private per- as dispute settlement and because the internal records of sons may use the tribunal’s verdict as sufficient basis for ADR proceedings are usually not publicly available. claiming compensation in domestic courts for damage caused to them by the noncompliance. Because member states have often responded with changes that did not Implementation, Compliance, and Sanctions bring them into compliance, the tribunal often mandates in PTA Dispute Settlement that a member state comply and refrain from employing When an ad hoc panel, arbitral tribunal, or other third any measure that is contrary to the tribunal’s judgment party produces its decision, the process of rights determi- (SIEL 2010). nation may be complete, but trade agreement enforcement A PTA ad hoc panel is convened for the limited purpose is far from done. The WTO DSU procedures recognize this of deciding a dispute, or making recommendations and rul- by outlining an organized compliance process. ings to the parties. For this reason, it generally is not given WTO compliance procedures begin by requiring the the power to enforce its own decisions; either the PTA itself party found in breach of the agreement to set a deadline for or the parties to the PTA authorize enforcement-related Dispute Settlement 491 actions such as suspension of concessions. U.S. FTAs cur- is followed in, for example, the Chile–Japan, EU–Korea, rently take the following approach, exemplified by the India–Korea, Japan–Indonesia, Japan–Switzerland, and Peru–U.S. FTA. After the panel report, the parties are to China–New Zealand agreements. In another variation, the agree on the resolution of the dispute. If they cannot agree original panel determines the level of the suspension on elimination of the noncompliance, they must negotiate when it rules on compliance (e.g., the ASEAN–India on compensation. If they cannot agree on compensation, Framework Agreement on Dispute Settlement and the or if the complaining party considers that the defending Australia–Singapore and China–Singapore PTAs). party has failed to carry out a settlement agreement, the Some recent PTAs also provide explicitly for arbitration complaining party may notify the defending party that it by the original panel if the defending party believes it has will suspend concessions, and at what level, and may sus- complied with the panel decision after concessions were sus- pend concessions 30 days later. The defending party can pended; examples are the China–New Zealand, EU–Korea, then ask the panel to reconvene to rule on whether it has Korea–U.S., and Peru–U.S. agreements. The WTO DSU actually complied or whether the suspension is manifestly does not make explicit provision for this situation. excessive. The same approach to compliance and suspen- sion appears in a number of other PTAs—for instance, the PTA Disputes: The Experience Canada–Colombia, Canada–Peru, Canada–Israel, and Thailand–Australia agreements. We now examine disputes brought using PTA dispute set- Since 2001, U.S. FTAs have provided that concessions tlement procedures and the written decisions that have may not be suspended if the defending party provides resulted from them. A survey of PTA dispute settlement timely notice that it will pay an annual monetary assess- experience presents a mixed picture, complicated by a ment to the complaining party. The assessment amount is shortage of organized information and by a definitional set by agreement between the parties, with a fallback to 50 question: at what point does a bilateral trade irritant ripen percent of the trade damage determined by the panel or by into the status of being a dispute? It is difficult to know the the complaining party. When the circumstances warrant, full extent of dispute settlement activity under any PTA, the PTA’s supervisory body can direct that the assessment particularly those featuring settlement of disputes by nego- be paid into a fund to be spent on initiatives to facilitate tiation or by ad hoc panels. Comparison with the WTO is trade between the parties. This provision responds to criti- useful and demonstrates that to some extent, disputes are cism that trade retaliation damages the economies of both being brought not in PTAs, but in the WTO. parties and harms innocent exporters in the other party, For PTAs with diplomatic or political dispute settlement, without achieving compliance. the true level of dispute activity is unknown and perhaps Some countries are unable to accept the U.S. approach to unknowable. Disputes do exist: for instance, the long- compliance and suspension of concessions because that running, legally focused dispute in the EC–Israel Associa- approach affords very little flexibility regarding the period tion Agreement regarding the status of products made in for compliance and relies in the first instance on the com- the West Bank, East Jerusalem, the Golan Heights, and the plaining party’s unilateral determination regarding compli- Gaza Strip (see Harpaz 2004; Broude 2007). Because these ance. These countries have adopted a different approach agreements treat dispute settlement as a diplomatic issue, patterned on the DSU and on proposals tabled for DSU they do not systematically require that panel decisions (if reform. The ASEAN–Australia–New Zealand agreement, any) or dispute outcomes be public. As Broude (2007) for instance, provides for a DSU-like process for determin- observes, the EU has no official list of disputes under its ing the postpanel compliance period by negotiation or by association agreements, and (at least at the time of that arbitration by the panel. Any disagreement about compli- article) there is no group within the European Commis- ance must be resolved by the original panel, reconvened for sion whose job is PTA dispute settlement. this purpose, which hands down a quick ruling. If the The true level of formal disputes in these agreements defending party states that it will not comply, or if the panel may also be low. In the EU’s association agreements, either has found a failure to comply, the defending party must party can block a panel proceeding. Moreover, the lack of negotiate on compensation, and if there is no agreement, detail in the agreements means that the disputing parties the complaining party has the right to suspend concessions must agree on procedures before they can convene a panel, equivalent to the nullification or impairment. The defend- further increasing the burden on a would-be disputant. A ing party can ask the panel to reconvene again if it believes it dispute may simmer for years as a diplomatic issue without has complied or that the level of the suspension is excessive going through any formal process. Garcia Bercero (2006) (in some agreements, “manifestly excessive�). This pattern mentions a long-standing dispute between Turkey and the 492 Amelia Porges EU over nonadoption of the EU acquis on pharmaceuticals, fits are still being phased in. It is natural for a PTA to post- and a dispute between the EU and Ukraine was blocked pone implementation for difficult sectors to the latest because of disagreement on how to split the costs of the point possible (as for sugar and Mexican corn in NAFTA, arbitration. In a PTA with no third-party dispute settlement for instance). If implementation is postponed, so are dis- procedures at all—for instance, the Closer Economic putes about failure to implement. The implication is that Partnership Agreement between China and Hong Kong an upsurge can be expected in the future. Second, PTA SAR, China—no dispute can even be considered unless institutions, and the repeated contacts they involve, pro- both parties recognize it exists. vide opportunities to avoid or proactively resolve disputes, There are also selection effects at work; as Broude diminishing the amount of trade conflict. Third, where a (2007) observes, the EU’s partners in its association agree- market access dispute can be brought either in the WTO or ments have not even litigated much in the WTO. Thus, in a PTA, the WTO may be a more attractive forum for PTAs with diplomatic or political dispute settlement complainants for several reasons: the WTO’s familiar insti- arrangements may attract governments that do not place a tutions and unblockable dispute settlement; the desire to high value on enforcement. And in the EU context, some be able to mobilize greater pressure against illegal denial of issues that might have been litigated in bilateral dispute market access by suspending MFN tariffs and other WTO settlement, such as the application to PTA members of obligations (particularly where the PTA’s margin of prefer- rules of origin or border trade measures, have instead been ence is low); the larger pool of neutral panelists in the litigated in the EU courts (see, e.g., ECJ 2010b). WTO; the broader issue scope of the WTO compared with Tribunal-based PTA dispute settlement has an extremely some PTAs; the possibility of forming alliances; access to broad range. The ECJ and its related supranational court technical assistance such as the Advisory Centre for WTO system, benefiting from abundant resources at the EU level Law; and the price tag. (The system cost of WTO dispute and linking to the domestic court systems in EU member settlement is included in a member’s annual assessment, states, have handled thousands of cases since 1952. By con- but in most PTAs the parties pay the panelists or pay for the trast, some African PTA tribunals, struggling with lack of cost of the tribunal.) infrastructure or resources, have little or no reported activ- The list of disputes in annex B, however, shows that some ity. The Andean Tribunal falls somewhere in the middle; real issues can only be dealt with through PTA dispute set- 85 tribunal cases against members for noncompliance were tlement. Among them is denial of rights that are only cre- initiated during the period 1987–2006, although much of ated by the PTA agreement, such as preferential market the Andean Tribunal’s work focuses on relatively narrow access, or application of preferential rules of origin. Because intellectual property issues (Helfer and Alter 2009). The Mercosur creates a right of free circulation, Uruguay República Bolivariana de Venezuela’s 2006 withdrawal brought and won a dispute in Mercosur against Argentina’s from the Andean Community eliminated a significant num- toleration of blockades on international bridges. Moreover, ber of pending disputes about Venezuelan trade measures.20 where the MFN tariff rate is high (35 percent for all the Panel-based dispute settlement has also been quite vari- Mercosur countries), the PTA, and the enforcement of PTA able, but there have been relatively few panel decisions or rights, may be essential for obtaining real market access, as arbitral awards. There have been 25 known decisions in is shown by the high number of Mercosur panel proceed- PTA formal proceedings, relating to 16 disputes, as listed in ings listed in annex B. The Mercosur partners have fully annex B. By comparison, in the WTO (1995 through litigated 12 disputes in the Mercosur forum, but they have March 3, 2011) there have been 423 WTO complaints; 136 only litigated three disputes against each other in the WTO, panel reports; 78 Appellate Body reports; 28 panel reports and two of those concerned antidumping measures not and 18 Appellate Body reports in compliance proceedings; covered by Mercosur. and 45 arbitration awards of various types.21 As in the WTO, the disputes formally raised under PTAs may sub- Conclusions stantially exceed the number of panel reports. Information available on dispute settlement under Chapter 20 of Since the beginning of the world economic crisis in 2008, NAFTA indicates that whereas there have been only 3 protectionist measures have increased. PTA negotiation has reports of Chapter 20 panels from 1994 through March 1, increased as well, both as an economic life raft and in reac- 2011, at least 11 disputes in this period were settled or were tion to the lack of progress in multilateral trade liberaliza- abandoned after formal consultations. tion. Reports on Mercosur, for instance, indicate an Why are PTA dispute levels so low? The first, and pri- increase in pure border protection measures such as mary, reason is that so many PTAs are very new, and bene- nonautomatic import licensing by Argentina, sectoral Dispute Settlement 493 private voluntary restraints on exports to Argentina pre- to, subtract from, or vary those procedures. There is no cipitated by import licensing, and up-valuation of reason not to borrow from other PTAs’ creative and imports using reference prices, which have affected constructive ideas for addressing dispute settlement. almost 11 percent of Argentina’s imports from Brazil and 22 percent of its imports from China (INTAL 2010). Annex A. Dispute Settlement And Mercosur is not the only jurisdiction involved: else- in CUSFTA and NAFTA where there are also tariff increases, valuation issues, new preshipment inspection requirements, buy-local or buy- Since the early 1990s, almost all dispute settlement pro- national requirements, export restrictions on strategic cedures in preferential trade agreements have been based materials, and other restrictions on trade in goods and on referral of disputes for decision by ad hoc panels. This services (Evenett 2010). All these measures have been put “NAFTA model� started with the Canada–U.S. Free in place with no great apparent upsurge in PTA dispute Trade Agreement (CUSFTA) in 1988 and underwent settlement. minor revisions in NAFTA. The choices made in 1988 Periods of sustained unemployment are a difficult time and 1993 remain influential today, and so it is useful to to push back against protection, but a rollback will be examine some key decisions and the rationales described needed when recovery occurs—and in order for recovery by negotiators. to occur. Both WTO and PTA dispute settlements and The CUSFTA negotiations took place against a back- institutions will have a part to play, and it will be a consid- ground of increasing trade conflict, particularly over U.S. erable challenge. Only time will tell whether these institu- trade remedies. According to Canada’s dispute settlement tions will do their job in helping move governments away negotiator, Canada’s overarching goal for the negotiation from crisis protectionism. was to obtain secure access to the U.S. market, which Some practical conclusions of use to negotiators emerge meant obtaining agreement to binding dispute settlement from the discussion in this chapter. (von Finckenstein 2000). Canada sought a permanent tri- bunal that could bind both parties and issue remedial • Every PTA has to have a way of settling disputes, and a orders. The United States preferred ad hoc panels; it PTA that promotes growth needs enforcement provi- wanted to avoid creating a new bureaucracy or a tribunal sions. Economic projections of the gains from a PTA are that might see itself as an independent player in bilateral based on the assumption of 100 percent compliance relations and might be perceived as telling the U.S. govern- with the PTA’s obligations. Ensuring compliance ment what to do, and it wished to preserve a right to retali- through enforcement is essential if the projected gains ate for noncompliance (Hart, Dymond, and Robertson are to materialize. Even if no disputes are anticipated, 1994, 302). The FTA eventually included two dispute set- enforcement provisions in a PTA reinforce the precom- tlement procedures: binational review (Chapter 19) for mitment of the governments, make their promises more antidumping and countervailing duty decisions, and credible, and signal that the PTA is a solid platform for government-government procedures for other disputes investment that will create jobs and economic growth. (Chapter 18). Chapter 18 was designed to address both • PTAs create public goods, in the form of economic sides’ objections about GATT dispute settlement proce- growth, transparency, and stability in the trading regime dures: the GATT process was too lengthy and could be and an environment for trading goods and services delayed; it involved panelists from other countries who based on the rule of law. They also promote open region- might not have the required expertise; and it provided no alism by strengthening institutions to make trade liberal- certainty regarding adoption or implementation of panel ization more transparent, less exclusionary to traders reports (von Finckenstein 2000). outside the PTA, and more accessible to firms investing Chapter 18 provided a specific timetable, a standing in the PTA area. roster of panelist candidates, and procedures designed to • The best time to reach agreement on fair rules to settle prevent most ways of blocking dispute settlement. It estab- disputes is during the PTA negotiation, and in advance lished a ministerial-level commission to oversee the func- of any known dispute. tioning of the agreement and to administer disputes. In a • Even if the parties start with low ambitions, experience provision sought by Canada, disputes could address not demonstrates that stronger, more ambitious rules can just actual measures but also proposed measures such as evolve later. pending legislation. The dispute process included consulta- • Most PTA dispute settlement procedures are now based tions, referral to the commission, and then referral to a on those in the WTO, but negotiators remain free to add panel for arbitration. Arbitration would be binding only if 494 Amelia Porges the parties agreed (or if the disputes concerned safeguards). • The provision on choice of forum remains, and before a In CUSFTA disputes, as it turned out, the parties never NAFTA party initiates a WTO dispute against another chose nonbinding arbitration (von Finckenstein 2000). party on grounds substantially equivalent to those avail- As for panel composition, CUSFTA called for five- able to it under NAFTA, it must consult with the third person panels composed of two citizens from each side, party. If the three cannot agree, the dispute must nor- chosen wherever possible from a standing roster. If a party mally be settled under NAFTA. failed to appoint its panelists, they would be selected from the roster by lot. The fifth panelist would be selected jointly A few changes responded to environmental concerns. and, in the absence of agreement, would be chosen by the NAFTA’s Article 103 gives precedence to obligations under four already selected panelists, or by lot. In practice, the five environmental treaties. In a WTO dispute between two sides initially decided by lot (by coin flip) which party NAFTA parties, if the responding party claims that its would choose the panel chair, and they then discussed can- measures are subject to Article 103, the complaining party didates (Winham 1993). The panel then provided an initial can only bring a NAFTA dispute. Similarly, if a dispute report and a final report responding to any objections by between NAFTA parties concerns sanitary and phytosani- the parties. tary (SPS) measures or environmental standards, the If, after receipt of a final panel report, the commission responding party can insist that the dispute be pursued was not able to reach agreement on a settlement within 30 only under NAFTA. A panel or a party can request a scien- days, and a party considered that its fundamental rights or tific review board on factual issues concerning environ- benefits under CUSFTA would be impaired by implemen- mental, health, safety, or other scientific matters. All of tation or maintenance of the measure at issue, that party these changes responded to criticism that GATT panels and would be free to suspend CUSFTA benefits of equivalent GATT rules were hostile to environmental regulation. effect. At the time, the GATT did not automatically author- ize suspension of concessions. Because CUSFTA incorporated by reference the GATT Annex B. PTA Dispute Settlement: provisions on national treatment and quantitative trade Ad Hoc Panel Decisions restrictions, the negotiators also included a choice of In comparison with experience in the GATT and the WTO, forum provision. Article 1801(2) provided that disputes ad hoc panel proceedings under preferential trade agree- arising under both CUSFTA and the GATT could be settled ments (PTAs) have yielded relatively few completed deci- in either forum, at the discretion of the complaining party, sions. The known panel decisions are listed and briefly but that once an election of forum had been made by initi- described here. The sources for this annex are Davey ation of dispute settlement, the procedure initiated would (1996); Reich (1996; Grebler (2003); Tussie and Delich be used to the exclusion of any other. This provision gave (2005); Gantz (2006); Barral (2007); SIEL (2010); Merco- the complaining party the option to choose the strongest sur arbitral awards on the Organization of American States substantive and procedural rules, while duplicative pro- (OAS) website at http://www.sice.oas.org/Dispute/mercosur/ ceedings were ruled out. ind_s.asp; and newspaper reports. NAFTA’s Chapter 20 continues the framework of CUSFTA Chapter 18, with a few changes. Canada-U.S. FTA (1988–93), Chapter 18 • Because NAFTA is trilateral, Chapter 20 gives interven- tion rights to a NAFTA party not involved in a dispute, 1. Canada’s landing requirement for Pacific Coast salmon including rights to attend hearings and to receive and and herring, final report, October 16, 1989. At issue was make submissions; the procedures also allow for two- a Canadian landing requirement that replaced GATT- complainant cases. inconsistent export restrictions on certain fish. • Chapter 20 makes provision for a consensus roster of 30 2. Lobsters from Canada, final report, May 21, 1990. The individuals experienced in law, international trade, and case involved a U.S. ban on interstate transport or sale of dispute settlement. Each party selects two panelists who whole live lobsters smaller than a minimum size. are citizens of the other disputing country. The selection 3. Article 304 and the definition of direct cost of processing or cannot be blocked if a panelist is on the agreed roster. direct cost of assembling, final report, June 8, 1992. The • Because panels under Chapter 18 had visibly split on complaint concerned a U.S. rule that did not allow cer- national lines, views in Chapter 20 panel reports are tain nonmortgage interest payments to count toward anonymous. meeting CUSFTA rules of origin. Dispute Settlement 495 4. The interpretation of and Canada’s compliance with Arti- (United States v. Canada, 1994); import restrictions on cle 701.3 with respect to durum wheat sales, final report, sugar (Canada v. United States, 1995); restrictions on February 8, 1993. The United States claimed that sales of small-package delivery (United States v. Mexico, 1995); durum wheat by the Canadian Wheat Board for export restrictions on tomato imports (Mexico v. United States, to the United States violated a CUSFTA ban on a gov- 1996); the Helms-Burton Act (Mexico and Canada v. United ernment entity’s selling products for export at a loss States, 1996); Mexican rebalancing for U.S. safeguards on (that is, at a price below the acquisition price of the broom-corn brooms (United States v. Mexico, 1996); goods, plus storage, handling, and other costs incurred restrictions on sugar imports (Mexico v. United States, with respect to those goods). 1998); farm products blockade (Canada v. United States, 5. Puerto Rico regulations on the import, distribution and 1998), bus service (Mexico v. United States, 1998); sport sale of ultra-high-temperature (UHT) processed milk fishing laws (United States v. Canada, 1999); and restric- from Quebec, final report, June 3, 1993. The case con- tions on potatoes (Canada v. United States, 2001). cerned milk standards in Puerto Rico. Mercosur (Since 1998) Israel–U.S. FTA 1. Application by Brazil of restrictive measures to trade Machine tools from Israel, settled informally after the panel with Argentina, award, April 28, 1999. The panel found report. The dispute concerned the U.S. decision to count that Brazilian import-licensing requirements on imports of machine tools assembled by Sharnoa Ltd., in imports from Argentina breached the Mercosur treaty Israel, from parts from Taiwan, China, against the U.S. and recommended compliance by December 31, 1999. import quota for machine tools from Taiwan, China. 2. Subsidies on production and export by Brazil of pork to Argentina, award, September 27, 1999. The panel rejected claims by Argentina regarding a system for NAFTA (Since 1994) corn stocking and Brazil’s advances on exchange con- 1. Tariffs applied by Canada to certain U.S.-origin agricul- tracts. It found that use of the PROEX export-financ- tural products, final report, December 2, 1996. The dis- ing program by Brazil was only acceptable for capital pute was brought by the United States against Canada’s goods. maintenance of tariff-rate quotas on certain dairy and 3. Application by Argentina of safeguard measures on tex- poultry products after full tariff elimination. Examining tiles from Brazil, award, March 10, 2000. The panel the relationship between the CUSFTA chapter on agri- found that Argentina’s application of safeguards to cultural trade, the Uruguay Round tariffication of agri- textiles was incompatible with the Mercosur legal cultural import quotas, and NAFTA, the panel found no regime and ordered revocation of the safeguard meas- breach by Canada. ure within 15 days. Brazil brought a complaint about 2. U.S. safeguard action taken on broom-corn brooms from the same textile safeguard to the WTO Textiles Moni- Mexico, final report, January 30, 1998. The dispute toring Body and requested a WTO panel (WT/DS190), brought by Mexico concerned the application of global which was established on March 20, 2000. The parties safeguard action on broom-corn brooms to imports of notified a settlement to the WTO in June 2000. such brooms from Mexico. The panel found the meas- 4. Application of antidumping measures on imports of ure in breach. whole chickens from Brazil, award, May 21, 2001. The 3. Cross-border trucking services, final report, February 6, panel found that Mercosur law did not regulate the 2001. In a dispute brought by Mexico, the panel found a application of antidumping measures and rejected the U.S. breach of the NAFTA commitment to permit opera- claim by Brazil. Brazil then took the same dispute to tion of Mexican trucking firms in four U.S. border states. the WTO (WT/DS241) and prevailed there. After U.S. congressional action terminating the pilot 5. Market access restrictions in Argentina on bicycles program for Mexican trucking, Mexico, on March 19, imported from Uruguay, award, September 23, 2001. 2009, announced suspension of NAFTA concessions Argentina treated Uruguayan bicycles made by one (i.e., an increase in tariffs to most favored nation, or MFN, company as non-Mercosur in origin and therefore levels) on imports of 90 products from the United States. subject to the common external tariff. The panel ruled that this measure violated Argentina’s Mercosur obli- Other matters settled or abandoned after consultations gations and ordered its revocation and the restoration but before panel proceedings concerned uranium exports of market access. 496 Amelia Porges 6. Brazilian import ban on remolded tires from Uruguay, April 26, 2008. The tribunal found that Argentina had award, January 9, 2001. The panel found that Brazil’s not brought itself into compliance and that until it did, ban on imports of remolded tires was incompatible Uruguay had the right to maintain compensatory with the Mercosur standstill on new trade restrictions. measures. Brazil later defended related measures in the WTO, 15. Failure by the Argentine state to adopt appropriate raising environmental defenses not mentioned in the measures to prevent and/or cease impediments to free Mercosur proceedings. circulation caused by blockages in Argentine territory of 7. Barriers to entrance of Argentine phytosanitary products access roads to the international bridges General San into the Brazilian market, award, April 9, 2002. The Martín and General Artigas, which connect Argentina panel found that Brazil had failed to implement in its and Uruguay, award, September 6, 2006. Uruguay domestic law five Mercosur Common Market Group challenged Argentina’s failure to act against environ- resolutions designed to create a streamlined phytosan- mental groups that blocked international bridges itary system for evaluating and registering food. It between Uruguay and Argentina from December found that Brazil was obligated to implement these 2005 to May 2006 to protest the construction of pulp measures within a reasonable period of time and that mills in Uruguay. Uruguay argued that the blockage six years was not reasonable; the panel ordered enact- injured imports, tourism, and transport, in violation ment within 120 days. of Mercosur guarantees of free circulation of goods, 8. Application of Uruguay’s specific internal taxes on the services, and factors of production via the elimina- sale of cigarettes, award, May 21, 2002. The panel found tion of quantitative restrictions and measures of that Uruguay’s method of calculating taxes on equivalent effect. The panel largely agreed. The imported cigarettes discriminated against Paraguayan underlying dispute concerned Argentina’s objections cigarettes, denying national treatment. It ordered to the construction of pulp mills in Uruguay, which Uruguay to cease discrimination within six months. Argentina separately appealed to the International 9. Uruguayan subsidies for processing of wool, award. April Court of Justice. 3, 2003. The panel found that Uruguayan export subsi- 16. Interlocutory appeal by Argentina objecting to selection dies for processed wool products exported to Merco- of the panel chairman in the free circulation dispute, sur were inconsistent with Mercosur law and had to be award, July 6, 2006. The tribunal rejected Argentina’s eliminated within 15 days. appeal as inadmissible under Mercosur rules and 10. Discriminatory and restrictive measures by Brazil on assessed all costs and expenses of the proceeding trade in tobacco and tobacco products, award, August 5, against Argentina. 2005. Uruguay brought a complaint concerning a Brazilian decree raising tariffs on tobacco and tobacco Other products to 150 percent. Brazil repealed the decree during the proceedings. Chilean price bands (application of price-band tariffs 11. Argentine ban on imports of remolded tires, award, on imports of vegetable oils). (a) Bolivia brought a dispute October 25, 2005. In the first case under the Protocol against Chile in 2000 under Chapter XIII of the of Olivos, the panel found that the Argentine ban was Bolivia–Chile (LAIA/ALADI) Economic Complementa- consistent with Mercosur law. tion Agreement 22, which provides that a dispute settle- 12. Appeal by Uruguay of award on Argentine ban on ment panel decision is fully binding on the parties. Chile imports of remolded tires, award, December 20, 2005. then reimbursed the safeguard duties collected. (b) The appellate tribunal reversed the award, finding that Argentina brought a complaint in 2000 under the the Argentine measure was incompatible with Merco- Administrative Commission of the Mercosur–Chile sur laws. (LAIA/ALADI) Economic Complementation Agreement 13. Request for ruling regarding excess in compensatory 35, which provides that a panel decision is nonbinding. measures in the dispute between Uruguay and Argentina After Argentina prevailed but Chile failed to comply, in on the prohibition of imports of remolded tires from October 2000 Argentina brought a dispute in the WTO Uruguay, award, June 8, 2007. The appellate tribunal (WT/DS207, Chile–Price Band System and Safeguard found that the Uruguayan compensatory measure was Measures Relating to Certain Agricultural Products), and proportional and lawful. prevailed in 2002. Argentina later prevailed in WTO 14. Review of tribunal decision regarding Argentine compli- compliance proceedings in 2007 but has not suspended ance with the tribunal award on remolded tires, award, concessions. Dispute Settlement 497 Notes 12. This is the case for the Japan–Mexico, Japan–Chile, Australia– Chile, ASEAN–Australia–New Zealand, Chile–U.S., and Singapore–U.S. The views expressed herein do not represent those of any present or past agreements. client or employer. 13. Res judicata is the legal doctrine that once a case has been deter- 1. CUSFTA and NAFTA panel decisions are available from many mined, neither party can bring the same claims regarding the same sub- sources, including the NAFTA Secretariat website, http://www.nafta-sec- ject matter against the other in another court. Lis alibi pendens is the legal alena.org/. Panel decisions of CUSFTA, NAFTA, and Mercosur, as well as doctrine that proceedings regarding the same facts cannot be commenced the texts of PTAs involving Western Hemisphere countries, are available at in a second court if the lis (i.e., action) is already pendens (pending) in the website of the Organization of American States (OAS) Foreign Trade another court. Information System, http://www.sice.oas.org/. 14. The two disputes brought under CUSFTA rather than the GATT 2. During the period 1996–2000, the OAS Trade Unit prepared an were those on Canada’s landing requirement for Pacific Coast salmon and inventory of dispute settlement in the Western Hemisphere (FTAA 2000), herring (final report October 16, 1989) and on lobsters from Canada as an input for the Free Trade Area of the Americas (FTAA) Negotiating (final report May 21, 1990). Group on Dispute Settlement. The FTAA negotiations have been stalled 15. PTAs that require model rules of procedure include NAFTA and since 2003, but the inventory remains a useful snapshot of these provi- all later U.S. PTAs; the Mercosur Protocol of Olivos; the Singapore–GCC, sions as they stood in 2000. Considerable information on dispute settle- Chile–EU, Chile–Australia, Chile–Colombia, Chile–Central America, ment and institutions in Latin America and the Caribbean is available at ASEAN–China, and Korea–Singapore PTAs; the ASEAN Enhanced the websites of Mercosur, the Andean Community (CAN), the Central DSM; and the Trans–Pacific SEP. American Court of Justice, the Caribbean Court of Justice, and regional 16. Appointing authorities named in PTAs include the WTO institutions, including the OAS, the Inter-American Development Bank Director-General (e.g., the China–New Zealand, ASEAN–Japan, Japan– (IADB), the secretariat of the Latin American Integration Association/ Malaysia, Japan–Vietnam, and New Zealand–Singapore PTAs and the Asociación Latinoamericana de Integración (LAIA/ALADI), and the Trans-Pacific SEP), the Secretary-General of the Permanent Court of United Nations Economic Commission for Latin America and the Arbitration in the Hague (e.g., the Japan–Switzerland and Canada–EFTA Caribbean/Comisión Económica para América Latina (ECLAC/CEPAL). PTAs), or a regional secretariat (the Mercosur Administrative Secretariat The last-named organization maintains a database on trade agreement for disputes under the Mercosur Protocol of Olivos, and the ASEAN dispute settlement, “Base de Datos Integrada de Controversias Comer- Secretary-General for the ASEAN Enhanced DSM). ciales de América Latina y el Caribe/Integrated Database of Trade Dis- 17. Participation by another PTA member is permitted by NAFTA, putes for Latin America and the Caribbean,� http://badicc.eclac.cl/. the ASEAN Enhanced DSM, the Trans-Pacific SEP, and the ASEAN– 3. CAN, “Procesos del Tribunal de Justicia: Acciones de Incumplim- Australia–New Zealand, Canada–EFTA, Colombia–EFTA, Japan–ASEAN, iento� (updated list of all Andean Tribunal noncompliance proceedings), China–ASEAN, Chile–Central America, and CAFTA–DR agreements. http://www.comunidadandina.org/canprocedimientosinternet/ListaEx 18. Examples of provisions for stakeholder input are the EU Trade pedientes11.aspx?CodProc=7&TipoProc=%27S%27. Barriers Regulation, U.S. Section 301, and China’s analogous legislation. 4. Ibid. 19. Examples of PTAs with provisions for alternative dispute resolu- 5. Andean Community Decision 680 of 2007 (authorizing ATJ tion include China–New Zealand (Article 187), Thailand–Australia budget for fiscal year 2008), http://www.comunidadandina.org/norma (Article 1803), Singapore–Australia (Article 16.3), the Trans-Pacific SEP tiva/dec/D680.htm. (Article 15.5), NAFTA (Article 2007) and later U.S. FTAs, and the 6. Investir en Zone Franc (IZF) website, http://www.izf.net/pages/ ASEAN Enhanced DSM, which authorizes good offices, mediation, or institutions-afrique---zone-uemoa-et-cemac/2199. The site contains conciliation by the ASEAN secretary-general (Article 4). information on trade and investment in the Central and West African 20. CAN, “Procesos del Tribunal de Justicia: Acciones de franc zones, including the legislation and jurisprudence of WAEMU/ Incumplimiento.� UEMOA and CEMAC. Information on legal systems in francophone 21. “Facts and Figures on WTO Dispute Settlement,� http://www Africa is also found at http://www.archive.org. .worldtradelaw.net. 7. African International Courts and Tribunals (AICT), http://www .aict-ctia.org/. 8. The term “nonviolation nullification or impairment� comes from the pioneering work by Robert E. Hudec, The GATT Legal System and Bibliography World Trade Diplomacy (Hudec 1975). ACWL (Advisory Centre on WTO Law). 2007. “Decision 2007/7 Adopted 9. Nonviolation remedies apply as follows in the PTAs studied: by the Management Board on 19 November 2007: Billing Policy and Chile–Japan, only for the chapters on trade in goods; India–Korea, Time Budget.� ACWL/MB/D/2007/7, ACWL, Geneva. http://www Korea–Singapore, Panama–Singapore, Canada–Costa Rica, Canada–Chile, .acwl.ch/e/documents/ACWL-MB-D-2007-7.pdf. Canada–Colombia, and Chile–Peru, goods and services; Panama– Alter, Karen J., and Laurence R. Helfer. 2011. “Legal Integration in the Andes: Taiwan, China, goods, services, and TBTs; El Salvador–Honduras– Law-Making by the Andean Tribunal of Justice.� SSRN: http://ssrn.com/ Taiwan, China, goods, services, SPS measures, and TBTs; Chile–Central abstract=1757772. Forthcoming in European Law Journal 17 (5). America, goods, services, TBTs, and aviation; Chile–Colombia and the Apaza, Pamela. 2006. “The Interaction between the Dispute Settlement Trans–Pacific SEP, goods, services, TBTs, and procurement; Nicaragua– Mechanisms of the American Regional Trade Agreements and the Taiwan, China, goods, services, TBTs, SPS measures, and IPRs; and WTO.� Master’s thesis, World Trade Institute, Bern. Chile–Australia, goods, services, TBTs, procurement, and IPRs. APEC (Asia-Pacific Economic Cooperation). 2004. “Best Practices for 10. Examples include the general cooperation chapters in the following RTAs/FTAs in APEC.� 2004/AMM/003, 16th APEC Ministerial Meet- PTAs: China–New Zealand, Japan–Mexico, Japan–Malaysia, Japan– ing, Santiago, Chile, November 17–18. Philippines, Japan–ASEAN, ASEAN–Australia–New Zealand, Australia– ———. 2008. “Managing the Noodle Bowl: The Fragility of East Asian Thailand, and Korea–Singapore. Regionalism.� Singapore Economic Review 53 (3): 449–78. 11. Examples include intellectual property cooperation provisions Banjo, Adewale. 2007. “The ECOWAS Court and the Politics of Access to in the Japan–Philippines and Japan–Thailand PTAs and the competi- Justice in West Africa.� Africa Development 32 (1): 69–87. tion policy cooperation obligations of the Australia–Chile, Australia– Barral, Welber. 2007. “The Brazilian Experience in Dispute Settlement.� Thailand, ASEAN–Australia–New Zealand, Australia–U.S., Peru–U.S., and ECLAC Project Documents Collection, UN Economic Commission SACU–EFTA PTAs. for Latin America and the Caribbean, Santiago. 498 Amelia Porges Bartels, Lorand. 2009. “Social Issues: Labour, Environment and Human ———. 2009. “COMESA Court of Justice.� COMESA, Lusaka. http:// Rights.� In Bilateral and Regional Trade Agreements: Commentary and about.comesa.int/lang-en/institutions/court-of-justice. Analysis, ed. Simon Lester and Brian Mercurio. Cambridge, U.K.: Daily Independent (Lagos). 2011. “Ecowas Court Rules on Oil Pollution: Cambridge University Press. Says FG, NNPC Can Be Sued.� January 19. Blackwood, Robin. 2008. “Report on International Organizations: Mer- Damro, Chad. 2006. “The Political Economy of Regional Trade Agree- cado Comúndel Sur/Southern Common Market.� Report on Mercosur ments.� In Regional Trade Agreements and the WTO Legal System, ed. PTR Decision of April 2008 on Argentina Remolded Tires Dispute, Lorand Bartels and Federico Ortino. New York: Oxford University American Society of International Law, Washington, DC. http://www Press. .asil.org/rio/mercosur.html. Davey, William. 1996. Pine & Swine: Canada–United States Trade Dispute Bolle, Mary Jane. 2009. “Overview of Labor Enforcement Issues in Free Settlement. The FTA Experience and NAFTA Prospects. Ottawa: Centre Trade Agreements.� Report Rs22823, Congressional Research Service, for Trade Policy and Law. Library of Congress, Washington, DC. Donaldson, Victoria, and Simon Lester. 2009. “Dispute Settlement.� In Bolle, Mary Jane, and M. Angeles Villarreal. 2009. “Peru Trade Promotion Bilateral and Regional Trade Agreements: Commentary and Analysis, Agreement: Labor Issues.� Report Rs22521, Congressional Research ed. Simon Lester and Brian Mercurio. Cambridge, U.K.: Cambridge Service, Library of Congress, Washington, DC. University Press. Bown, Chad, and Bernard Hoekman. 2007. “Developing Countries and EAC (East African Community). 2009. “East African Court of Justice: Enforcement of Trade Agreements: Why Dispute Settlement Is Not Information, Documentation, Rulings, Judgments and Advisory Enough.� CEPR Discussion Paper 6459. Centre for Economic Policy Opinions.� EAC, Arusha, Tanzania. http://www.eac.int/eacj.html. Research, London. http://www.cepr.org/pubs/dps/DP6459.asp. ———. 2010. “EALA Passes Budget Estimates for FY 2010/2011.� Press Broude, Tomer. 2007. “From Pax Mercatoria to Pax Europea: How Trade release, May 3, EAC, Arusha, Tanzania. http://www.eac.int/compo Dispute Procedures Serve the EC’s Regional Hegemony.� In Economics nent/content/article/1-latest-news/449-press-release-eala-passes-eac- of European Union Law, ed. Paul B. Stephan, 319–34. Cheltenham, budget-estimates-of-us-59963040-for-financial-year-20102011.html. U.K.: Edward Elgar. Also available as bepress Legal Series Working East African (Nairobi). 2006. “COMESA Court Faces Legal, Funding Paper 650, http://law.bepress.com/expresso/eps/650. Woes.� March 29. ———. 2009. “Regional Economic Integration in the Middle East and Echandi, Roberto. 2004. “Settlement of Trade Disputes between States: The North Africa: A Primer.� Research Paper 12–09, Hebrew University of Experience of Central American Countries.� In Inter-Governmental Jerusalem. SSRN: http://ssrn.com/abstract=1333292. Trade Dispute Settlement: Multilateral and Regional Approaches, ed. Busch, Marc, and Eric Reinhardt. 2003. “The Evolution of Dispute Settle- Julio Lacarte and Jaime Granados. London: Cameron May. ment.� In Trade Policy Research 2003, ed. John M. Curtis and Dan ECJ (European Court of Justice). 2010a. Annual Report 2009 of the Court Ciuriak. Ottawa: Department of Foreign Affairs and International of Justice of the European Communities. Luxembourg: ECJ. Trade. http://curia.europa.eu/jcms/jcms/Jo2_7000/. Cameron, Maxwell, and Brian Tomlin. 2002. The Making of NAFTA: How ———. 2010b. Brita GmbH v. Hauptzollamt Hamburg-Hafen, Case the Deal Was Done. Ithaca, NY: Cornell University Press. C-386/08, OJ C 100/4, April 17, Luxembourg. CAN (Comunidad Andina/Andean Community). 2008a. “Manual de EFTA (European Free Trade Association) Court. 2009. “Financial State- Procedimientos del Sistema Andino de Solución de Controversias.� ments for 2008.� EFTA Court, Luxembourg. http://www.eftacourt.int/ Secretaría General de la Comunidad Andina, Lima. http://www.comu index.php/court/mission/financial_statement/. nidadandina.org/public/libro_Manual.pdf. Essien, Victor. 2006. “Regional Trade Agreements in Africa: A Historical ———. 2008b. “Decision 680: Presupuesto del Tribunal de Justicia de la and Bibliographic Account of ECOWAS and CEMAC.� Hauser Global Comunidad Andina para el año 2008.� CAN Commission, Lima. Law School Program, New York University, New York. http://www http://www.comunidadandina.org/normativa/dec/dectema_inst_pres .nyulawglobal.org/globalex/CEMAC_ECOWAS.htm. .htm. EU (European Union). 2010. “Budget on Line.� EU, Brussels. http://eur- Cárdenas, Emilio J., and Guillermo Tempesta. 2001. “Arbitral Awards lex.europa.eu/budget/www/index-en.htm: Draft 2011 Budget, Section under Mercosur’s Dispute Settlement Mechanism.� Journal of Interna- 4 on Court of Justice, at http://eur-lex.europa.eu/budget/data/DB_ tional Economic Law 4 (2, March): 337–66. 2011/EN/SEC04.pdf. CCJ (Caribbean Court of Justice). 2008. “Annual Report 2007–2008. European Commission. 2007. “Proposal for a Council Decision on a Appendix: Financial Statements of the Caribbean Court of Justice.� Community Position within the EU–Chile Association Committee on CCJ, Port of Spain. http://www.caribbeancourtofjustice.org/annualre the Establishment of a List of Arbitrators as Required by Article 185 port0708.html. Paragraph 2 of the Agreement Establishing an Association between CEC (Commission on Environmental Cooperation). 2011. “Registry of the European Community and its Member States, of the One Part, and Citizen Submissions.� NAFTA Commission on Environmental the Republic of Chile, of the Other Part.� COM(2007)47, February 7, Cooperation, Montreal, Quebec. http://www.cec.org. Accessed OJ C 138/9, June 22, 2007, Brussels. February 7, 2011. ———. 2009. “26th Annual Report on the Monitoring and Application of CEPAL/ECLAC (Comisión Económica para América Latina/Economic Community Law.� COM(2009) 675, European Commission, Brussels. Commission for Latin America and the Caribbean). 2006. “Reducir la http://ec.europa.eu/community_law/infringements/infringements_a Incertidumbre Juridica, Un Desafio de la Integración Subregional en nnual_report_26_en.htm. America Latina y el Caribe.� Boletin FAL (Facilitatión del Comercio y Evenett, Simon J., ed. 2010. “Will Stabilisation Limit Protectionism? The el Transporte en America Latina y el Caribe)� 235 (March 2006). San- 4th GTA Report. A Focus on the Gulf Region.� Global Trade Alert, tiago: CEPAL/ECLAC. Centre for Economic Policy Research, London. Choi, Won Mog. 2004. “Making a Better Dispute Settlement Mechanism Fiorentino, Roberto V., Luis Verdeja, and Christelle Toqueboeuf. 2006. for Regional Trade Agreements: Lessons of Integration Efforts in East “The Changing Landscape of Regional Trade Agreements: 2006 Asia.� In WTO and East Asia: New Perspectives, ed. Mitsuo Matsushita Update.� Discussion Paper 12, World Trade Organization, Geneva. and Dukgeun Ahn. London: Cameron May. Foroutan, Faezeh. 1992. “Regional Integration in Sub-Saharan Africa: ———. 2007. “Legal Analysis of Korea: ASEAN Regional Trade Integra- Experience and Prospects.� Policy Research Working Paper 992, World tion.� Journal of World Trade 41 (3): 581–603. Bank, Washington, DC. COMESA (Common Market for Eastern and Southern Africa). 2000. “Offi- Francois, Joseph, and Miriam Manchin. 2009. “Economic Impact of a cial Gazette of COMESA.� COMESA, Lusaka. http://about.comesa.int/ Potential Free Trade Agreement (FTA) between the European Union attachments/062_2000%20Gazette%20Vol.%206%20No5.pdf. and the Commonwealth of Independent States.� CASE (Center for Dispute Settlement 499 Social and Economic Research), Warsaw. SSRN: http://ssrn.com/ Hicks, Raymond, and Soo Yeon Kim. 2009. “Credible Commitment abstract=1393697. through PTAs and Their Effects on Trade: A Study of Asia’s Reciprocal FTAA (Free Trade Area of the Americas). 2000. “Inventory of Dispute Set- Trade Agreements.� Prepared for the Annual Meeting of the American tlement Mechanisms, Procedures and Legal Texts Established in Exist- Political Science Association, September. SSRN: http://ssrn.com/ ing Trade and Integration Agreements, Treaties and Arrangements in abstract=1449204. the Hemisphere and in the WTO.� FTAA ngds/w/08/Rev.4, February 7. Horn, Henrik, Petros C. Mavroidis, and André Sapir. 2010. “Beyond the http://www.ftaa-alca.org/ngroups/NGDS/publications/dsw8r4_e.asp. WTO? An Anatomy of EU and US Preferential Trade Agreements.� Gama Sá Cabral, Jeanine, and Gabriella Giovanna Lucarelli de Salvio. World Economy 33 (11): 1565–88. 2008. “Considerations on the Mercosur Dispute Settlement System House Committee on Ways and Means. 2007. “Congressional-Adminis- and the Impact of Its Decisions on the WTO Dispute Settlement Sys- tration Agreement on a New Trade Policy for America.� U.S. House of tem.� Journal of World Trade 42 (6): 1013–40. Representatives, Washington, DC. http://waysandmeans.house.gov/ Gantz, David. 2006. “The United States and NAFTA Dispute Settlement: News.asp?FormMode=release&ID=513. Ambivalence, Frustration and Occasional Defiance.� Arizona Legal Hudec, Robert E. 1975. The GATT Legal System and World Trade Diplo- Studies Discussion Paper 06-16, University of Arizona, Tucson, AZ. macy. New York: Praeger. SSRN: http://ssrn.com/abstract=918542. ICJ (International Court of Justice). 1950. Interpretation of Peace Treaties Garcia Bercero, Ignacio. 2006. “Dispute Settlement in European Union with Bulgaria, Hungary and Rumania, Advisory Opinion, First Phase Free Trade Agreements: Lessons Learned?� In Regional Trade Agree- [1950] ICJ Rep., 65; Second Phase [1950] ICJ Rep. 221. Luxembourg, ICJ. ments and the WTO Legal System, ed. Lorand Bartels and Federico Inama, Stefano. 2005. “The Association of South East Asian Nations– Ortino. New York: Oxford University Press. People’s Republic of China Free Trade Area: Negotiating beyond Gathii, James Thuo. 2010. “African Regional Trade Agreements as Flexi- Eternity with Little Trade Liberalization?� Journal of World Trade 39 ble Legal Regimes.� North Carolina Journal of International Law and (3): 559–79. Commercial Regulation 35 (3, Spring). SSRN: http://ssrn.com/ INTAL (Institute for the Integration of Latin America and the abstract=1622274. Caribbean). 2005. “CARICOM Report No. 2.� INTAL, Inter-American Gómez, R., Hernando Jose, and Javier Gamboa.2007. “Cambios institu- Development Bank, Buenos Aires. cionales para la negociación del TLC de Colombia con los EEUU y _____. 2006. “MERCOSUR: The Permanent Review Tribunal Issued Its derivados de su implementación.� Colección de documentos de First Award.� INTAL Monthly Newsletter 114 (January). proyecto, LC/W.146, Comisión Económica para América Latina/ ———. 2009. “MERCOSUR Report No. 13, 2007 (Second Semes- Economic Commission for Latin America and the Caribbean ter)–2008 (First Semester).� INTAL, Inter-American Development (CEPAL/ECLAC), Santiago. Bank, Buenos Aires. Gonzalez, Anabel. 2006. “La solución de controversias en los acuerdos ———. 2010. “MERCOSUR Report No. 14, 2008 (Second Semester)– regionales de America Latina con paises desarrollados.� Serie Com- 2009 (First Semester).� INTAL, Inter-American Development Bank, ercio Internacional 68, Division de Comercio Internacional y Inte- Buenos Aires. gracion, LC/L.2525-P, Comisión Económica para América Karli, Mehmet. 2008. “Assessing the Development Friendliness of Dispute Latina/Economic Commission for Latin America and the Caribbean Settlement Mechanisms in the Economic Partnership Agreements and (CEPAL/ECLAC), Santiago. an Analytical and Comparative Guide to the Dispute Settlement Pro- Goode, Walter. 2005. Negotiating Free-Trade Agreements: A Guide. Can- visions in the EU’s FTAs.� European Studies Centre, Occasional Paper, berra: Department of Foreign Affairs and Trade, Australia. University of Oxford, Oxford, U.K. SSRN: http://ssrn.com/ http://www.dfat.gov.au/publications/negotiating_ftas/index.html. abstract=1270111. Granados, Jaime, and Ann Casanova. 2004. “Treatment of Asymmetries Kawai, Masahiro, and Ganeshan Wignaraja. 2009. “The Asian Noodle in Size and Development in Trade Agreements: Some Suggestions Bowl: Is It Serious for Business?� ADBI Working Paper 136, Asian for Regional North-South Dispute Settlement Negotiations.� In Development Bank Institute, Manila. Inter-Governmental Trade Dispute Settlement: Multilateral and Keck, Alexander, and Roberta Piermartini. 2005. “The Economic Impact Regional Approaches, ed. Julio Lacarte and Jaime Granados. London: of EPAs in SADC Countries.� Staff Working Paper ERSD-2005-04, Cameron May. World Trade Organization, Geneva. Grebler, Eduard. 2003. “Dispute Settlement: Regional Approaches, 6.2 Keck, Alexander, and Simon Schropp. 2007. “Indisputably Essential: The Mercosur.� Module for the United Nations Conference on Trade and Economics of Dispute Settlement Institutions in Trade Agreements.� Development (UNCTAD) Course on Dispute Settlement in Interna- Staff Working Paper ERSD-2007-02, World Trade Organization, tional Trade, Investment, and Intellectual Property, UNCTAD/EDM/ Geneva. Misc.232/Add.28, United Nations, New York and Geneva. Kirk, Ron, and Hilda Solis. 2010. Consultation request addressed to Min- Guira, Jorge. 2000. “Mercosur: The Emergence of a Working System of Dis- ister of Economy and Minister of Labor and Social Protection of pute Resolution.� NAFTA Law and Business Review Americas 6: 255–75. Guatemala, July 30, USTR, Washington, DC. Harpaz, Guy. 2004. “The Dispute over the Treatment of Products Kreinin, Mordechai, ed. 2000. Building a Partnership: The Canada–United Exported to the European Union from the Golan Heights, East States Free Trade Agreement. Lansing, MI: Michigan State University. Jerusalem, the West Bank and the Gaza Strip: The Limits of Power and Kwak, Kyung, and Gabrielle Marceau. 2006. “Overlaps and Conflicts of the Limits of the Law.� Journal of World Trade 38 (6): 1049–58. Jurisdiction between the WTO and Regional Trade Agreements.� In Hart, Michael, Bill Dymond, and Colin Robertson. 1994. Decision at Mid- Regional Trade Agreements and the WTO Legal System, ed. Lorand Bar- night: Inside the Canada–US Free Trade Negotiations. Vancouver, tels and Federico Ortino. New York: Oxford University Press. Canada: UBC Press. Lacarte, Julio, and Jaime Granados, eds. 2004. Inter-Governmental Trade Helfer, Laurence R., and Karen J. Alter. 2009. “The Andean Tribunal of Jus- Dispute Settlement: Multilateral and Regional Approaches. London: tice and Its Interlocutors: Understanding Preliminary Reference Pat- Cameron May. terns in the Andean Community.� Vanderbilt University Law School Loungnarath, Vilaysoun, and Céline Stehly. 2000. “The General Dispute Law and Economics Research Paper 09-01, Vanderbilt University Law Settlement Mechanism in the North American Free Trade Agreement School, Nashville, TN. SSRN: http://ssrn.com/abstract=1334733. and the World Trade Organization System: Is North American Hessbruegge, Jan Arno. 2011. “ECOWAS Court Judgment in Habré v. Regionalism Really Preferable to Multilateralism?� Journal of World Senegal Complicates Prosecution in the Name of Africa.� American Trade 34 (1): 39–71. Society of International Law, Insights (February 3). http://www.asil Luo, Yan. 2005. “Dispute Settlement in the Proposed East Asian Free .org/insights110203.cfm. Trade Agreement: What We Can Learn from the EU and the NAFTA.� 500 Amelia Porges Presented at ILA British Branch Spring 2005 Conference, Edinburgh, Rojas Penso, Juan. 2004. “New Dispute Settlement Perspectives in the Latin Scotland, May. http://www.hss.ed.ac.uk/ila/Day2.htm. American Integration Association (ALADI).� In Inter-Governmental Mercosur (Southern Cone Common Market, Mercado Común del Sur), Trade Dispute Settlement: Multilateral and Regional Approaches, ed. Tribunal Permanente de Revisión. 2006. “Laudo del Tribunal Perma- Julio Lacarte and Jaime Granados. London: Cameron May. nente de Revisión Constituido en Plenario para Entender en el Sáez, Sebastian. 2007. “The Countries of Latin America and the Caribbean Recurso de Revisión Presentado por la República Argentina contra la and Trade Disputes: An Analysis.� Boletin FAL (Facilitation del Com- Decisión del Tribunal Arbitral AD HOC de Fecha 21 de junio de 2006, ercio y el Transporte en America Latina y el Caribe) 249 (May). Also Constituido para Entender en la Controversia Promovida por la published in Spanish by Comisión Económica para América República Oriental del Uruguay contra la República Argentina sobre: Latina/Economic Commission for Latin America and the Caribbean ‘Impedimientos a la Libre Circulación derivado de los Cortes en Terri- (CEPAL/ECLAC) as “América Latina y el Caribe en las Controversias torio Argentino de Vías de Acceso a los Puentes Internacionales Gral. Comerciales: Un Análisis Aplicando BADICC,� CEPAL/ECLAC, San Martín y Gral.Artigas.’� Award of July 6, 2006. http://www.sice.oas Santiago. .org/Dispute/mercosur/Laudo002_006_p.pdf. Sauvé, Pierre, and Martin Molinuevo. 2008. “Does Architecture Matter? Mercurio, Bryan, and Rebecca LaForgia. 2005. “Expanding Democracy: The Treatment of Trade and Investment in Financial Services in Why Australia Should Negotiate for Open and Transparent Dispute Selected Preferential Trade Agreements.� NCCR Trade Regulation Settlement in Its Free Trade Agreements.� Melbourne Journal of Inter- Working Paper 2008/9, Swiss National Centre of Competence in national Law 6: 485–514. Research, Bern. http://www.nccr-trade.org/nccr-publications/8.html. Merrills, J. G. 1991. International Dispute Settlement. 2nd ed. Cambridge, Schiff, Maurice, and L. Alan Winters. 2003. Regional Integration and U.K.: Grotius Publications. Development. New York: Oxford University Press. Morgan, David. 2008. “Dispute Settlement under PTAs: Political or Schreuer, Christoph. 2001. The ICSID Convention: A Commentary. Cam- Legal?� In Challenges to Multilateral Trade: The Impact of Bilateral, bridge, U.K.: Cambridge University Press. Preferential and Regional Agreements, ed. Ross Buckley, Vai Io Lai, and SIEL (Society of International Economic Law). 2010. “Mercosur Award Laurence Boulle. Alphen aan den Rijn, Netherlands: Kluwer Law No. 1/2005.� SIEL Translation Project. http://www.sielnet.org/siel- International. Also available as University of Melbourne Legal Studies translationproject. Research Paper 341, SSRN: http://ssrn.com/abstract=1203022. Smith, James McCall. 2000. “The Politics of Dispute Settlement Design: MOX Plant Case. 2006. Commission v. Ireland, Case No. C-459/03, Judg- Explaining Legalism in Regional Trade Pacts.� International Organiza- ment in OJ C 165, July 15. http://eur-lex.europa.eu/LexUriServ/ tion 54: 137–80. LexUriServ.do?uri=OJ:C:2006:165:0002:0003:EN:PDF. Snyder, Francis. 2009. “China, Regional Trade Agreements and WTO Nakagawa, Junji. 2007. “No More Negotiated Deals? Settlement of Trade Law.� Journal of World Trade 43 (1): 1–57. and Investment Disputes in East Asia.� Journal of International Eco- Stephanou, Constantinos. 2009. “Including Financial Services in Preferen- nomic Law 10 (4): 837–67. tial Trade Agreements: Lessons of International Experience for China.� Nyaungwa, Nyasha. 2010. “Zimbabwean Government Could Lose More Policy Research Working Paper 4898, World Bank, Washington, DC. Assets in SADC.� Namibia Economist (April 23). http://allafrica.com/ Sucharitkul, Sompong. 2003. “Dispute Settlement: Regional Approaches, stories/201004230689.html. 6.3 ASEAN.� Module for the United Nations Conference on Trade and Opertti, Didier. 2004. “The New MERCOSUR Dispute Settlement Mecha- Development (UNCTAD) Course on Dispute Settlement in Interna- nism.� In Inter-Governmental Trade Dispute Settlement: Multilateral tional Trade, Investment, and Intellectual Property, UNCTAD/ and Regional Approaches, ed. Julio Lacarte and Jaime Granados. Lon- EDM/Misc.232/Add.29, United Nations, New York and Geneva. don: Cameron May. Szepesi, Stephen. 2004. “Comparing EU Free Trade Agreements: Dispute Pauwelyn, Joost. 2004. “Going Global, Regional, or Both? Dispute Settle- Settlement.� In Brief 6G (July). Maastricht, Netherlands: European ment in the Southern African Development Community (SADC) and Centre for Development Policy Management. http://www.ecdpm.org/ Overlaps with the WTO and Other Jurisdictions.� Minnesota Journal of inbrief6g. World Trade 13: 231–304. Thorp, Teresa. 2010. “Challenges of Incorporating the Economic Acquis ———. 2006. “Adding Sweeteners to Softwood Lumber: The WTO- Communautaire of the East African Community in a New Common NAFTA ‘Spaghetti Bowl’ Is Cooking.� Journal of International Economic Market.� Society of International Economic Law (SIEL), Second Bien- Law 9: 1–10. nial Global Conference, University of Barcelona, July 8–10. SSRN: Peña, Felix. 2004. “Overview of New Trends in Dispute Settlement Mech- http://ssrn.com/abstract=1632439. anisms under Preferential Trade Agreements.� In Inter-Governmental Tussie, Diana, and Valentina Delich. 2005. “Dispute Settlement between Trade Dispute Settlement: Multilateral and Regional Approaches, ed. Developing Countries: Argentina and Chilean Price Bands.� In Man- Julio Lacarte and Jaime Granados. London: Cameron May. aging the Challenges of WTO Participation: 45 Case Studies, ed. Peter Piérola, Fernando, and Gary Horlick. 2007. “WTO Dispute Settlement and Gallagher, Patrick Low, and Andrew L. Stoler. Cambridge, U.K.: Dispute Settlement in the ‘North-South’ Agreements of the Americas: Cambridge University Press. Considerations for Choice of Forum.� Journal of World Trade 41 (5): Unzurrunzaga Valle, Maria Ester. 2005. “Trade Disputes Settlement in the 885–908. European Union Agreements with Mexico and Chile.� Inter-American Porges, Amelia. 2003. “Settling WTO Disputes: What Do Litigation Mod- Development Bank (IADB), Washington, DC. http://www.iadb.org/ els Tell Us?� Ohio State Journal on Dispute Resolution 19 (1): 141–84. intal/aplicaciones/uploads/ponencias/Foro_ELSNIT_2005_02_Unzur Ramirez Robles, Edna. 2006. “Political and Quasi-Adjudicative Dispute runzaga.pdf. Settlement Models in European Union Free Trade Agreements: Is the U.S. DOL (U.S. Department of Labor). 2007. “Status of Submissions Quasi-Adjudicative Model a Trend or Is It Just Another Model?� Staff under the North American Agreement on Labor Cooperation Working Paper ERSD-2006-09, WTO Economic Research and Statis- (NAALC).� List updated to September 2007, U.S. DOL, Washington, tics Division, Geneva. http://www.wto.org/english/res_e/reser_e/ersd DC. http://www.dol.gov/ilab/programs/nao/status.htm. 200609_e.pdf. ———. 2010. “Status of Submissions under the NAALC.� Division of Reich, Arie. 1996. “From Diplomacy to Law: The Juridicization of Interna- Trade Agreement Administration and Technical Cooperation, U.S. tional Trade Relations.� Northwestern Journal of International Law & Department of Labor, Washington, DC. http://www.dol.gov/ilab/ Business 17 (2/3): 775–849. programs/nao/status.htm. Reuters. 2010. “Tribunal Wants SADC to Suspend Harare over Land USTR (Office of the U.S. Trade Representative). 2010. “USTR Kirk Case.� July 17. http://www.polity.org.za/article/tribunal-wants-sadc- Announces Labor Rights Trade Enforcement Case Against to-suspend-harare-over-land-case-2010-07-17. Guatemala.� Press release, July 30, USTR, Washington, DC. Dispute Settlement 501 Victor, David G. 1998. “The Operation and Effectiveness of the Montreal Woolcock, Stephen. 2007. “European Union Policy towards Free Trade Protocol’s Non-Compliance Procedure.� In The Implementation and Agreements.� ECIPE Working Paper 03/2007, European Centre for Effectiveness of International Environmental Commitments, ed. David International Political Economy, Brussels. G. Victor, Kal Raustiala, and Eugene B. Skolnikoff. Cambridge, MA: World Bank. 2000. Trade Blocs. World Bank Policy Research Report. New MIT Press. York: Oxford University Press. Vigil Toledo, Ricardo. 2004. “Dispute Settlement in Andean Community WTO (World Trade Organization). 1996a. Analytical Index/Guide to Law.� In Inter-Governmental Trade Dispute Settlement: Multilateral GATT Law and Practice. 6th ed. Geneva: World Trade Organization. and Regional Approaches, ed. Julio Lacarte and Jaime Granados. Lon- ———. 1996b. “Rules of Conduct for the Understanding on Rules and don: Cameron May. Procedures Governing the Settlement of Disputes.� WT/DSB/RC/1, Vinuesa, Raul Emilio. 2005. “Enforcement of Mercosur Arbitration WTO, Geneva. Awards within the Domestic Legal Orders of Member States.� Texas ———. 1999. “Canada–Measures Affecting the Importation of Milk and International Law Journal 40: 425–42. the Exportation of Dairy Products.� Panel Report, WTO, Geneva. von Finckenstein, Konrad W. 2000. “Dispute Settlement under the Free ———. 2009. “2010–2011 Budget Proposals.� WT/BFA/W/189, WTO, Trade Agreement.� In Building a Partnership: The Canada–United Geneva. States Free Trade Agreement, ed. Mordechai Kreinin. Lansing, MI: Xuto, Nilaratna. 2005. “Thailand: Conciliating a Dispute on Tuna Exports Michigan State University. to the EC.� In Managing the Challenges of WTO Participation: 45 Case Wang, Jiangyu. 2009. “ASEAN–China Free Trade Agreement.� In Bilateral Studies, ed. Peter Gallagher, Patrick Low, and Andrew Stoler. Cam- and Regional Trade Agreements: Commentary and Analysis, ed. Simon bridge, U.K.: Cambridge University Press. Lester and Brian Mercurio. Cambridge, U.K.: Cambridge University Yang, Tseming. 2007. “International Treaty Enforcement as a Public Press. Good: Institutional Deterrent Sanctions in International Environ- Whitelaw, James. 2004. “The MERCOSUR Dispute Settlement System.� In mental Agreements.� Michigan Journal of International Law 27: Inter-Governmental Trade Dispute Settlement: Multilateral and 1131–84. Regional Approaches, ed. Julio Lacarte and Jaime Granados. London: Yang, Yongzheng, and Sanjeev Gupta. 2005. “Regional Trade Arrange- Cameron May. ments in Africa: Past Performance and the Way Forward.� IMF Work- Winham, Gilbert. 1993. “Dispute Settlement in NAFTA and the FTA.� In ing Paper WP/05/36, International Monetary Fund, Washington, DC. Assessing NAFTA: A Trinational Analysis, ed. Steven Globerman and Yoshimatsu, Hidetaka. 2006. “Collective Action Problems and Regional Michael Walker. Vancouver, Canada: Fraser Institute. http://oldfraser Integration in ASEAN.� Contemporary Southeast Asia: A Journal of .lexi.net/publications/books/assess_nafta/dispute.html. International and Strategic Affairs 28: 115–40. INDEX Notes are indicated with n following the page number. A tariff cuts, 151–52 Absorption principle, 168 trade flows and preferences, 148–49 Abuja Treaty of 1991, 113 U.S. agreements, 155–57 ACCC. See Australian Competition and Consumer Commission World Trade Organization agreement, 146 ACP. See African, Caribbean, and Pacific states AIA. See ASEAN Investment Agreement Ad hoc panels, 485, 494–96 Aid for trade, 263 ADR. See Alternative dispute resolution Alternative dispute resolution, 489 Africa American National Standards Institute, 199 African Common Market, 113 AMU. See Arab Maghreb Union African Economic Community, 113 Andean Community African Growth and Opportunity Act, 149, 173–74 services trade, 270n Common Market for Eastern and Southern Africa, 24, 155, Andean Tribunal of Justice, 472 169, 328, 361 ANSI. See American National Standards Institute dispute settlement, 470 Antidumping rules, 182–86, 194 European Union-South Africa Trade and Development ANZCERTA. See Australia-New Zealand Closer Cooperation Agreement, 222–24 Economic Relations Trade Agreement preferential trade agreement participation, 56–57 ANZGPA. See Australia and New Zealand Government Southern African Customs Union, 223, 330, 358 Procurement Agreement Southern African Development Community, 155, 174–75, 223, 330 APEC. See Asia-Pacific Economic Cooperation West African Economic and Monetary Union, 154–55 Arab Maghreb Union, 56 African, Caribbean, and Pacific states, 52 ASEAN. See Association of Southeast Asian Nations African Common Market, 113 ASEAN Free Trade Agreement, 343 African Economic Community, 113 ASEAN Investment Agreement, 320 African Growth and Opportunity Act, 149, 173–74 ASEAN Trade in Goods Agreement, 228–29 AFTA. See ASEAN Free Trade Agreement Asia. See also Association of Southeast Asian Nations Agency for International Development, 436, 439 Asia-Pacific Economic Cooperation, 23, 205–206, 328, 343, 372 Agglomeration forces, 78–80 dispute settlement, 469 Agreement on Agriculture, 145–46 preferential trade agreement participation, 50–56 Agreement on Rules of Origin, 163 services trade, 262 Agreement on Safeguards, 461n Asia-Pacific Economic Cooperation, 23, 205–206, 328, 343, 372 Agreement on the Application of Sanitary and Phytosanitary Association of Southeast Asian Nations Measures, 210–13, 218 certification and accreditation services, 109 Agreement on the Technical Barriers to Trade, 210–13, 218 liberalization strategy, 84 Agricultural trade preferential trade agreements, 42, 55–56 bilateral preferential trade agreements, 153–75 Trade in Goods Agreement, 228–29 domestic subsidies, 152–53 At-the-border policies, 18–19 economic integration, 146–47 ATIGA. See ASEAN Trade in Goods Agreement economics of, 144–47 ATJ. See Andean Tribunal of Justice European Union agreements, 154–55 AUSFTA. See Australia-United States Free Trade Agreement evaluating trade preferences, 149–51 Australia export subsidies, 152–53 Australia and New Zealand Government Procurement Agreement, 373 institutions, 153 Australia-New Zealand Closer Economic Relations multilateral commitments, 144–46 Trade Agreement, 152, 357 preferential commitments, 144–46 Australia-United States Free Trade Agreement, 219–20 public goods, 153 Australian Competition and Consumer Commission, 359 regional preferential trade agreements, 151–53 labor agreements with developing countries, 297–99, 303 regional trade flows, 148 labor mobility in preferential trade agreements, 283–84 safeguards, 152 Australia and New Zealand Government Procurement Agreement, 373 503 504 Index Australia-New Zealand Closer Economic Relations Central American Free Trade Agreement, 105, 150, 156–57, 230, 328, 438 Trade Agreement, 152, 357 Central Asia Australia-United States Free Trade Agreement, 219–20 preferential trade agreement participation, 50–54 Australian Competition and Consumer Commission, 359 Central Corridor Transit Transport Facilitation Agency, 345n Auto Pact, 123, 132 Central European Free Trade Agreement, 39 Automobile tariffs, 83–84 Centre d’Études Prospectives et d’Informations Internationales, 149 CEPAs. See Closer Economic Partnership Agreements B CEPTs. See Common effective preference tariffs Barbados CET. See Common external tariff services trade, 266 CEZ. See Common Economic Zone Basel Committee, 98 CGE. See Computable general equilibrium Baucus, Max, 457 Chile Behind-the-border policies, 18–19 Chile-China Free Trade Agreement, 229 Better Work, 429 Chile-United States Free Trade Agreement, 225–26 Bilateral cumulation, 166 European Union-Chile Free Trade Agreement, 221–22, 328 Bilateral investment free trade agreements, 156 impact of, 324 patent-registration linkage, 394 intellectual property rights and, 401–402 trade agreements, 398–99 liberalization of, 313–15 worker rights provisions, 434–35 Bilateral investment treaties, 108n, 155, 290 China Bilateral labor agreements, 287–90 Chile-China Free Trade Agreement, 229 Bilateral safeguards, 182–84, 187 China-New Zealand Free Trade Agreement, 226–27 Bilateral trade agreements Closer Economic Partnership Agreements, 469 agricultural trade provisions, 153–75 Clean Development Mechanism, 411 differentiated from plurilateral agreements, 43–44 Clinton, Bill, 431 disadvantages of, 123 Closer Economic Partnership Agreements, 469 investment protection, 317 Cluster economics, 80 Bipartisan Agreement, 397 Cobden-Chevalier Treaties, 127 BITs. See Bilateral investment treaties Codex Alimentarius Commission, 201, 213 BLAs. See Bilateral labor agreements COMESA. See Common Market for Eastern and Southern Africa Break-even curves, 74–76 Committee for Standardization, 199, 201 Bush, George H. W., 431, 433 Common Economic Zone, 53–54 Bush, George W., 430, 457 Common effective preference tariffs, 84 Business visitors, 277 Common external tariff collection of duties, 117 C design and level of, 114–16 CACJ. See Central American Court of Justice economic implications of, 113–14 CACM. See Central American Common Market ownership of collected duties, 116–17 CAFTA. See Central American Free Trade Agreement Common Market for Eastern and Southern Africa, 24, 155, 169, 328, 361 Canada Community Competition Commission, 358 Canada-Costa Rica Trade Agreement, 328 Competition curves, 74–75 Canada-U.S. Free Trade Agreement, 136, 153, 493–95 Competition models, 89–92 human rights provisions, 449–52, 462n Competition policy labor agreements with developing countries, 294–95, 301 benefits of various types of agreement, 350–51 labor mobility in preferential trade agreements, 281–82 consumer policy, 359–60 worker rights provisions, 433–34 consumer protection, 352 Canada-U.S. Free Trade Agreement, 136, 153, 493–95 cost-benefit analyses, 353–54 Capacity building agreement, 230 development and, 347–49 Caribbean Basin Initiative, 157 economics of, 349–55 Caribbean Forum of African, Caribbean, and Pacific States effects on third parties, 355 cultural cooperation and aid for trade, 263 enforcement mechanisms, 351 EU Economic Partnership Agreement, 328 hard versus soft law, 360 services trade, 266–69 implementation costs, 352–53 tourism liberalization, 260 implementation issues, 361–63 Caribbean region implications for open regionalism, 354–55 Caribbean Basin Economic Recovery Act, 157 international cooperation, 356 Caribbean Basin Initiative, 157 models of regional competition regimes, 357 Caribbean Basin Trade Partnership Act, 157 positive spillovers, 351–52 Caribbean Community Regional Negotiating Mechanism, 154 regional policy in practice, 355–63 Caribbean Court of Justice, 473 scope, 358–59 preferential trade agreement participation, 54–55 Competitive liberalization, 133 CARIFORUM. See Caribbean Forum of African, Caribbean, Computable general equilibrium, 148, 150 and Pacific States Conciliation, 489–90 CBI. See Caribbean Basin Initiative Conformity assessment, 200 CCC. See Community Competition Commission Contractual services suppliers, 277, 283 CEC. See Council for Environmental Cooperation Copyright protection, 395–96 CEFTA. See Central European Free Trade Agreement Copyright Treaty, 395 CEN. See Committee for Standardization Costa Rica Central America Canada-Costa Rica Trade Agreement, 328 Central American Trade Dispute Resolution Mechanism, 473 Cotonou Agreement, 154, 162, 453 preferential trade agreement participation, 54–55 Council for Environmental Cooperation, 415 Central American Common Market, 42, 55 Countervailing duties, 182–86 Central American Court of Justice, 473 Country strategy paper, 224 Index 505 CU. See Customs unions specialized institutions, 488 Cumulation, 166–68 SPS and TBT provisions, 231 CUSFTA. See Canada-U.S. Free Trade Agreement standing tribunals, 471–73, 484 The Customs Union Issue, 70 structuring systems, 475–78 Customs unions support for, 479–80 agreements in force and planned, 112 surveys, 468 collection of duties, 117 targets of complaint, 476 design and level of common external tariff, 114–16 third-party participation, 486 economic implications of, 112–14 transparency, 486–88 function of, 38, 111 United States, 468–69 ownership of collected duties, 116–17 Dispute Settlement Mechanism, 469 reasons for choosing, 113 Dispute Settlement Understanding, 467 DMS. See Dispute Settlement Mechanism D Doha Development Agenda, 257–58, 398 DDA. See Doha Development Agenda Doha Round trade negotiations, 39, 145–46 De minimis rules, 168 Dominican Republic Defragmentation, 74, 77 Central American Free Trade Agreement, 150, 230, 328 Demand-linked circular causality, 79–80 services trade, 267 Department of Labor, 436–38 Double-taxation treaties, 290 Developing countries DSU. See Dispute Settlement Understanding agricultural trade, 143–58 Duopoly, 91 capacity building for competitiveness, 104 Duty drawback, 168 designing successful reforms, 97 economic governance, 102–104 E environmental issues, 412 East Asia labor agreements, 285–88, 293–04 dispute settlement, 469 promoting market access, 101–102 preferential trade agreement participation, 55–56 reform priorities, 96–97 services trade, 262 trade liberalization strategies, 97–100 EBA. See Everything But Arms preferences Diagnostic Trade Integration Study, 334 EBRD. See European Bank for Reconstruction and Development Diagonal cumulation, 166–67 ECO. See Economic Cooperation Organization Digital Millennium Copyright Act of 1998, 396 Economic and Monetary Community of Central Africa, 155 Digital trade, 259 Economic Community of Central African States, 155 Dispersion forces, 78–80 Economic Community of West African States, 154–55, 170–71, 286 Dispute settlement Economic Cooperation Organization, 54 ad hoc panels, 485, 494–96 Economic integration agreement, 38, 50 Africa, 470 Economic issues alternative enforcement methods, 488–90 agglomeration forces, 78–80 amicus curiae submissions, 487–88 agricultural trade, 143–58 appeal, 481 conflicting priorities, 415–16 civil society participation, 486–88 dispersion forces, 78–80 collective enforcement procedures, 488–89 dynamic gains of regionalism, 74–78 complaints by stakeholders against governments, 487 financial assistance, 414–15 compliance, 490–91 locational effects of liberalization, 80–81 conciliation, 489–90 new economic geography, 78–81 consultations, 478–79 regional interindustry trade, 84–86 correction, 481 regionalism effects, 69–73 defining grounds for complaint, 475–76 technical assistance, 414–15 diplomatic, 470–71 trade arrangements, 81–84 duration of panel process, 481 Economic needs tests, 305n East Asia, 469 Economic partnership agreements, 52, 105, 174, 398, 453–54 enforcement, 470–75 EDI. See Electronic data interchange environmental issues, 417–18, 421, 422 EEA. See European Economic Area European Union, 468 EEC. See European Economic Community formal phase, 479 EEEMRA. See Electrical and electronic equipment good offices, 489 mutual recognition agreement government procurement, 376–78 EFTA. See European Free Trade Area; European Free Trade Association implementation, 490–91 EIA. See Economic integration agreement infrastructure, 479–80 El Salvador investment agreements, 311 human rights issues, 460 Latin America, 469–70 Electrical and electronic equipment mutual recognition agreement, 227 mediation, 489–90 Electricity Regulatory Forum, 35n panel procedures, 480–81 Electronic data interchange, 334 panelist ethics, 486 Electronic trade, 259 participation in settlement and enforcement, 484–88 Emergency safeguard mechanisms, 249 political, 470–71 Enabling Clause, 66n preferential trade agreements disputes, 491–96 Enforce-your-own-laws standard, 431, 432 qualification and selection of decision makers, 484–85 ENP. See European Neighborhood Policy referral to ad hoc panel, 473–74 ENTs. See Economic needs tests remand, 481 Environmental issues rosters, 485–86 dispute resolution, 417–18, 421, 422 sanctions, 490–91 hard- and soft-law approaches, 416–18 special sectors, 481–84 implementation of provisions, 413–16 506 Index Environmental issues (continued) FTC. See Federal Trade Commission preferential trade agreements and, 409–11, 420–22 Full cumulation, 167–68 regional policy, 411–13 trade and, 408–09 G EPAs. See Economic partnership agreements GAFTA. See Greater Arab Free Trade Area Equivalence, 204 Galileo satellite system for global navigation, 334 ESM. See Emergency safeguard mechanisms Gambling, online, 262 Euro-Mediterranean Conference, 154 GATS. See General Agreement on Trade in Services Euro-Mediterranean free trade agreements, 154, 220–21 GATT. See General Agreement on Tariffs and Trade Europe General Accountability Office, 437 European Coal and Steel Community, 23 General Agreement on Tariffs and Trade European Court of Justice, 24, 471–72 Article III requirements, 19 preferential trade agreement participation, 50–54 effect of regionalism, 123–24 European Bank for Reconstruction and Development, 333, 345n Enabling Clause, 42, 66n European Economic Area, 151, 472 human rights, 445–48 European Economic Community, 138n preferential trade agreement notifications, 39–42 European Free Trade Area, 472 General Agreement on Trade in Services European Free Trade Association investment provisions, 308, 312 agricultural provisions, 151 market access issues, 20 human rights, 452–53 services standards, 248–49, 261–62, 266–69 preferential trade agreement participation, 20, 52–53 General Secretariat of the Andean Community, 24 European Neighborhood Policy, 105–106 Generalized system of preferences, 19–20, 37, 145 European Union Geographical indications, 398, 404n agricultural trade agreements, 154–55 Global Forum on Migration and Development, 287 Caribbean Forum of African, Caribbean, and Pacific States Global free trade, 124–25 Economic Partnership Agreement, 328 GLOBAL-GAP, 229 cultural cooperation and aid for trade, 263 Global multilateral tariff liberalization, 82 dispute settlement, 468 Global safeguards, 182–86, 191–92 environmental issues in preferential trade agreements, 421–22 Government procurement European Union-Chile Free Trade Agreement, 221–22, 328 buy national provisions, 369 European Union-Morocco preferential trade agreement, 220–21 costs of implementing provisions, 378–79 European Union-South Africa Trade and Development Cooperation dispute settlement, 376–78 Agreement, 222–24 economic and developmental dimensions of, 367–73 Everything But Arms preferences, 19–20 flexible provisions, 375 human rights provisions, 453–56 international instruments, 372 labor agreements with developing countries, 295–96, 302 multilateral trade provisions, 380–82 labor mobility in preferential trade agreements, 282–83 policy options, 382 market integration and economic development, 105–107 preferential trade agreements, 371–83 rules of origin in existing trade agreements, 168–71 provisions in preferential trade agreements, 373–80 services trade, 266–69 technical assistance, 379–80 tourism liberalization, 260 thresholds, 374 trade facilitation, 341–42 trade negotiating strategy, 382 Trade-Related Aspects of Intellectual Property Rights, 397–400 transparency, 370–71 worker rights provisions, 435 Government Procurement Agreement, 249, 372–76 Everything But Arms preferences, 19–20, 52 GPA. See Government Procurement Agreement Externalities, 22 Graduate trainees, 283 Extra-preferential trade agreements, 59–63 Gravity models, 150 Greater Arab Free Trade Area, 331 F GSP. See Generalized system of preferences Factor-market competition, 79 Guatemala Factory Asia, 85–86 human rights issues, 460 FAO. See Food and Agriculture Organization FDA. See Food and Drug Administration H FDI. See Foreign direct investment Haberler, Gottfried, 69–70 Federal Trade Commission, 359 Harmonization Florence Forum, 35n services trade, 247–49 Food and Agriculture Organization, 201 standards, 204–206 Food and Drug Administration, 198 Harmonized System, 162 Foreign Agricultural Resource Management Services, 305n Harmonized System-8 product level, 135 Foreign direct investment, 318–19 Human Resources and Skills Development Canada, 288 Foundation of Enterprises for the Recruitment of Foreign Labor, 305n Human rights Frankel and Wei model, 128 background, 447–49 Free trade agreements. See also specific agreement by name knowledge gaps, 460–61 agriculture provisions, 156 linkages, 460–61 customs unions, 111, 113–16 obligations of trading nations, 444–45 human rights provisions, 454–55 (See also specific countries by name) preferential trade agreements and, 449–60 purpose of, 38 trade agreement provisions, 445–47 resource transfers, 24 Free trade area, 71–72 I Free Trade Commission, 415–16 ICCPR. See International Covenant on Civil and Political Rights Frictional barrier liberalization, 72–73 ICESCR. See International Covenant on Economic, FTA. See Free trade area Social and Cultural Rights FTAs. See Free trade agreements ICSID. See International Center for the Settlement of Investment Disputes Index 507 IFC. See International Finance Corporation L ILO. See International Labour Organization Labor mobility Immigration and Naturalization Service, 438 agreements with developing countries, 293–99 Independent professionals, 277, 283 bilateral labor agreements, 287–90 Information asymmetries, 22–23, 199 categories of labor included in trade agreements, 277 Information Technology Agreement, 139n concept of, 276–77 Input-cost-linked circular causality, 80 distributional effects of Mode 4 liberalization, 278–81 INS. See Immigration and Naturalization Service economic benefits, 277–81 Intellectual property rights estimates of gains, 292–93 bilateral investment rules, 401–402 importance of, 275 enforcement of, 396–97 natural persons definition, 276 negotiating considerations, 388–90 preferential trade agreement provisions, 300–304 preferential trade agreements, 17 in preferential trade agreements, 281–87 Trade-Related Aspects of Intellectual Property Rights, 18 recommendations, 290–92 TRIPS+ provisions, 390–401 Labor rights Interindustry trade, 84–86 bipartisan agreement, 433 International Center for the Settlement of Investment economics of, 428–29 Disputes, 311, 485 implementation and enforcement of provisions, 435–39 International Center for Trade and Sustainable Development, 403n International Labour Organization standards, 427 International Covenant on Civil and Political Rights, 444 provisions in preferential trade agreements, 429–30 International Covenant on Economic, Social and sanctions authorized for violations, 436 Cultural Rights, 444 worker rights provisions, 430–35 International Finance Corporation, 345n, 429, 439 Workers Rights Project, 438 International Labour Organization, 305, 427, 429 LAIA. See Latin American Integration Association International Organization for Migration, 287 Latin America International Organization for Standardization, 197–98 dispute settlement, 469–70 International Trade and Labour Grants and Latin American Integration Association, 55, 469–71 Contributions Program, 434 Liberalization International Trade Organization, 445–47 global multilateral tariff liberalization, 82 Intra-preferential trade agreements, 59–63 locational effects, 80–81 Intracorporate transferees, 277 multilateral liberalization, 121–38 Investment of parts and components, 83–84 economic impact, 317–20 preferential, 83 impact of bilateral investment, 324 regional, 76–78 incorporation into preferential trade agreements, 307–309 unilateral, 82–83 international agreements, 315–16 welfare effects, 78 liberalization of bilateral investment, 313–15 Local-market competition, 79 multilateralization of provisions, 316–17 nature and scope of provisions, 309–313 M preferential trade agreements covering investment, 322–23 Management systems, 198 protection provisions, 317 Market fragmentation, 74 types of preferential trade agreements, 311–13 Mauritania typology of investment measures, 309–311 human rights issues, 460 WTO rules, 308 Meade, James, 70–71 IPRs. See Intellectual property rights Mediation, 489–90 ISO. See International Organization for Standardization Mediterranean countries Israel-U.S. Free Trade Agreement, 495 Euro-Mediterranean free trade agreements, 154, 220–21 ITO. See International Trade Organization Mercosur, 136, 227–28, 339, 482, 495–96 Meta-standards, 198 J Metrology, 200 Jamaica Mexican migrant workers, 438 services trade, 268 Mexico Japan human rights issues, 460 labor agreements with developing countries, 296–97, 302 trade agreements, 398–99 labor mobility in preferential trade agreements, 283 MFA. See Multifibre Arrangement Trade and Investment Liberalization Fund, 340 MFN. See Most favored nation policies worker rights provisions, 435 Middle East Johnson-Cooper-Massell proposition, 86–88 preferential trade agreement participation, 56–57 Johnson diagram, 86–88 Migrant workers, 438 Joint Initiative for Priority Skills Acquisition Act of 2004, 289 Milk Marketing Board, 153 Jordan MLAT. See Mutual legal assistance treaty U.S.-Jordan preferential trade agreement, 437–38 MMB. See Milk Marketing Board workers rights, 439 MNEs. See Multinational enterprises Jordan standard, 431, 432 Monopoly power, 21–22 Juggernaut building block logic, 126–28, 138n Monopoly profit maximization, 89–90 Morocco K European Union-Morocco preferential trade agreement, 220–21 Kemp-Wan theorem, 70, 128 Most favored nation policies Kennedy Round, 34n imported liberalization, 130–31 Korea-U.S. Free Trade Agreement, 157 investment provisions, 309–310, 316–17 KORUS. See Korea-U.S. Free Trade Agreement labor agreements, 290 Krugman, Paul, 78, 85 liberalization of, 27, 101 Kyoto Protocol, 411 multilateral liberalization, 133–34 508 Index Most favored nation policies (continued) O rules of origin and, 131–32 Obama, Barack, 432–33, 436 tariff rates, 20 OECD. See Organisation for Economic Co-operation and Development trade services, 243 Oligopoly, 91–92 MRAs. See Mutual recognition agreements One-stop border posts, 330–31 MTNs. See Multilateral trade negotiations Online gambling, 262 Multiannual indicative program, 224 Organisation for Economic Co-operation and Development, 96, 198 Multifibre Arrangement, 99, 429 Multilateral liberalization P bargaining model, 132–33 Pacific region Frankel and Wei model, 128 preferential trade agreement participation, 55–56 induced liberalization and protection, 130–32 Packaging requirements, 198 juggernaut building block logic, 126–28 PAFTA. See Pan-Arab Free Trade Agreement Kemp-Wan theorem, 128 Pan-Arab Free Trade Agreement, 56 most favored nation policies, 133–34 Partial-scope agreement, 38 reciprocal trade agreements, 121–23 Patent-registration linkage, 391, 394 regionalism, 123–24, 134–35 Patents protection, 391, 394 rules for preferential trade agreements, 137–38 Performances and Phonograms Treaty, 395 stumbling block logic, 124–26 Permanent Appeals Tribunal, 482 unilateral liberalization, 130 Permanent workers, 276–77 veto avoidance, 128–29 Peru welfare consequences of preferential trade agreements, 135–37 U.S.-Peru Free Trade Agreement, 397 Multilateral trade agreements Pharmaceutical test data, 391, 394–95 agricultural provisions, 144–46 Plurilateral trade agreements Multilateral trade negotiations, 126–27 differentiated from bilateral agreements, 43–44 Multinational enterprises, 318 membership, 64–65 Mutual legal assistance treaty, 350 networks, 51–54 Mutual recognition agreements, 26–27, 207, 220, 247–49 trade developments, 57–59 Mutual recognition of standards, 204 Pollutant release and transfer registry, 414, 420 Positive integration, 18 N Preferential liberalization, 83 NAAEC. See North American Agreement on Environmental Cooperation Preferential rules or origin NAALC. See North American Agreement on Labor Cooperation absorption principle, 168 NACEC. See North American Commission best-practice suggestions, 164 for Environment Cooperation change of tariff classification, 162–63 NAFTA. See North American Free Trade Agreement cumulation, 166–68 Nash tariffs, 132 defining origin, 161–64 National Labor Committee, 439 duty drawback, 168 Natural persons economic development and, 174–75 defined, 276 economic implications of, 171–75 Natural trade blocs, 136 in existing preferential trade agreements, 168–71 NCTA. See Northern Corridor Transit Agreement outward processing, 168 NCTS. See New Computerized Transit System quantifying costs associated with, 173–74 Negative integration, 18 restrictiveness index, 171, 173 New Computerized Transit System, 334 specific manufacturing process, 163–64 New economic geography, 78–81 substantial transformation determination, 162–64 New trade theory, 85 tolerance rules, 168 New Zealand trade preferences and, 164–71 Australia-New Zealand Closer Economic Relations utilization of trade preferences, 171–73 Trade Agreement, 152 value added, 163 China-New Zealand Free Trade Agreement, 226–27 Preferential trade agreements. See also specific issues by name labor agreements with developing countries, 297–99, 303 at-the-border policies, 18–19 labor mobility in preferential trade agreements, 283–84 behind-the-border policies, 18–19 NLC. See National Labor Committee benefits of, 1 No-trade-to-free-trade liberalization, 76–77 composition of, 42–43 Nontariff measures, 47 customized problem solving, 30–31 North Africa deep integration preference, 17–18, 33–34 preferential trade agreement participation, 56–57 disadvantages of, 1 North America economic benefits, 19–23 preferential trade agreement participation, 54–55 externalities, 22 North American Agreement on Environmental Cooperation, 409, 414 flexibility, 30–31 North American Agreement on Labor Cooperation, 431, 437, 482 geographic configuration, 43–45 North American Commission for Environment Cooperation, 414, 482 geopolitical objectives, 23–24 North American Free Trade Agreement global landscape, 49–57 agriculture provisions, 149, 156 impact on third parties, 26–27 creation of, 123 impact on trade, 57–63 dispute settlement, 493–95 implementation, 31–33 environmental issues, 414 information asymmetries, 22–23 investment provisions, 311–13 institutions for reform, 24–25 labor standards, 431–32, 437–38 monopoly power, 21–22 rules of origin, 165 motivations for deep integration, 19–25 Northern Corridor Transit Agreement, 345n negative integration, 18 NTMs. See Nontariff measures North-South agreements, 95–108 Index 509 policy anchoring, 25 international trade provisions, 217–19 policy complementarities, 28–30 preferential trade agreements provisions, 219–29 political-economy benefits, 23–25 technical assistance needs, 229–31 positive integration, 18 trade facilitation, 329–30 preferences, 26–28 WTO standards and guidelines, 218 product standards, 203–210 SARS. See South African Revenue Service proliferation of, 39–42 Seasonal Agricultural Workers Program, 288 regulatory objectives, 28–30 Secure Trade in the APEC Region initiative, 343 scope of, 45–49 SEP. See Strategic Economic Partnership societal benefits, 23 Services trade specificities of deep integration, 25–33 assessing depth of preferential liberalization, 254–69 structural configuration, 43 costs and benefits of preferential treatment, 237–39 systemic effects, 27–28 digital trade, 259 trade diversion, 26 domestic regulation, 247 trade remedy provisions, 179–95 dynamic economies of scale, 239 trends, 37–39 economics of, 236–42 Prisoners’ Dilemma, 132 emergency safeguard mechanisms, 249 Process standards, 198 fixed cost measures, 238 Product standards government procurement, 249 cost effects, 201 harmonization, 247–49 designing and implementing, 199–200 investment, 250 economic rationale for, 198–99 liberalization in preferential trade agreements, 242–54 economics of preferential standard market access, 247 liberalization, 203–204 modalities of liberalization, 251–54 externalities, 198–99 most favored nation treatment, 243 facilitating market access, 204 mutual recognition, 247–49 harmonization of standards, 204–206 negotiating approaches, 253 information asymmetries, 199 number of service providers restriction, 238 inventory methods, 203 regulatory cooperation, 247–49 management systems, 198 regulatory intensity, 240–41 meta-standards, 198 rules of origin, 250–51 mutual recognition of, 206–207 static economies of scale, 239 preferential trade agreement provisions, 209–210 subsidy disciplines, 249 process standards, 198 sunk costs, 238–39 regional and international dimensions, 200–201 third-country effects, 242 regional standards, 210–13 tourism sector, 260 review of, 207–208 transparency, 243, 247 trade and, 197–203 variable cost measures, 237–38 trade effects, 202 SIAs. See Sustainability impact assessments voluntary standards, 201 Singapore Accreditation Council, 209 Protocol of Olivos, 482 Singapore-Australia Free Trade Agreement, 224–25, 343 PRTR. See Pollutant release and transfer registry Single Market Economy, 285 PTAs. See Preferential trade agreements “Small PTA� diagram, 88–89 Smith, Adam, 69–70 R South Africa Reciprocal trade agreements, 121–23 European Union-South Africa Trade and Development Regional economic integration organization exception clause, 316 Cooperation Agreement, 222–24 Regional interindustry trade, 84–86 South Africa Pesticide Initiative Program, 224 Regionalism South Africa Strategic Partnership, 223 agricultural provisions, 151–53 trade agreements, 398–99 dynamic gains of, 74–78 South African Revenue Service, 340 effects of, 69–73 South America environment and, 411–13 preferential trade agreement participation, 54–55 multilateral liberalization and, 123–24, 134–35 South Asia trade services regulatory cooperation, 239–42 preferential trade agreement participation, 55–56 REIO. See Regional economic integration organization exception clause South Asian Free Trade Area, 384n Revised Kyoto Convention, 328–29 Southern African Customs Union, 223, 330, 358 Revised Treaty of Chaguaramas, 353 Southern African Development Community, 155, Risk-sharing funds, 345n 174–75, 223, 330 Roll-up principle, 168 Southern Cone Common Market, 23, 169, 339 ROOs. See Rules of origin SPS Agreement. See Agreement on the Application of Sanitary Rules of origin, 27, 88, 131–32, 161–76 and Phytosanitary Measures SPS standards. See Sanitary and phytosanitary standards S Stagflation, 287 SACU. See Southern African Customs Union Standing tribunals, 471–73, 484 SADC. See Southern African Development Community STAR initiative. See Secure Trade in the SAFTA. See Singapore-Australia Free Trade Agreement APEC Region initiative Sanitary and phytosanitary standards Stolper-Samuelson effect, 138n Agreement on the Application of Sanitary and Strategic Economic Partnership, 42 Phytosanitary Measures, 210–13 Sub-Saharan Africa dispute settlement, 231 preferential trade agreement participation, 57 environmental issues, 409, 421, 422 Sustainability impact assessments, 408 implementation costs, 229–31 Sweatshops, 429, 439 510 Index T prohibition of, 190–91 TAFTA. See Thailand-Australia Free Trade Agreement trade and protection diversion, 191–95 Tariff-rate quotas, 145, 158n TradeMark Southern Africa, 330 Tariff rates, 20 TRAINS. See Trade Analysis and Information System TBT. See Technical barriers to trade Trans-European Networks on Transport, 334 TBT Agreement. See Agreement on the Technical Barriers to Trade Trans-Kalahari Corridor, 330, 345n TDCA. See Trade and Development Cooperation Agreement Trans-Pacific Partnership, 155 Technical barriers to trade Trans-Pacific Strategic Economic Partnership Agreement, 42, 209, 434 Agreement on the Technical Barriers to Trade, 210–13 Transaction spillover, 22 dispute settlement, 231 Transit corridors, 337–40 environmental issues, 409, 421, 422 Transparency Mechanism, 46 implementation costs, 229–31 Transparency Mechanism for Regional Trade Agreements, 66n international trade provisions, 217–19 Treaty of the Functioning of the European Union, 356 nontariff barriers, 72–73 Triangle trade, 85–86 preferential trade agreements provisions, 219–29 TRIMS. See Trade–Related Investment Measures Agreement technical assistance needs, 229–31 Trinidad and Tobago WTO standards and guidelines, 218 services trade, 269 Temporary visa programs, 288 TRIPS agreement. See Trade-Related Aspects of Intellectual Property Temporary worker programs, 291 Rights Temporary workers, 276–77 TRQs. See Tariff-rate quotas Terms-of-trade approach, 139n TFAP. See Trade facilitation action plan U TFEU. See Treaty of the Functioning of the European Union UDHR. See Universal Declaration of Human Rights Thailand UN International Convention on the Protection of the Rights of human rights issues, 460 Migrant Workers and Their Relatives, 305n Thailand-Australia Free Trade Agreement, 225 UNCITRAL. See United Nations Commission on Third-country fabric rule, 149, 173–74 International Trade Law TIFAs. See Trade and investment framework agreements UNCTAD. See United Nations Conference on Trade and Development TILF. See Trade and Investment Liberalization Fund Unilateral liberalization, 82–83, 130 Tinbergen, Jan, 18 United Nations Commission on International Trade Law, 311, 371–72 TKC. See Trans-Kalahari Corridor United Nations Conference on Trade and Development, 167, 203, 315 Tolerance rules, 168 United States Tourism liberalization, 260 agricultural trade agreements, 155–57 TPA. See Trade Promotion Act Australia-United States Free Trade Agreement, 219–20 Trade Analysis and Information System, 203 Canada-U.S. Free Trade Agreement, 136, 153, 493–95 Trade and Development Cooperation Agreement, 222–24 Chile-United States Free Trade Agreement, 225–26 Trade and investment framework agreements, 155 dispute settlement, 468–69 Trade and Investment Liberalization Fund, 340 environmental issues in preferential trade agreements, 420–21 Trade diversion, 26 human rights provisions, 455–59 Trade facilitation Israel-U.S. Free Trade Agreement, 495 cooperation for regional facilitation, 329–31 Korea-U.S. Free Trade Agreement, 157 delivering regional public goods, 336–37 labor agreements with developing countries, 293–94, 300 dynamic effects, 332–36 labor mobility in preferential trade agreements, 281–82 economic dimensions, 331 rules of origin in existing trade agreements, 168–71 institutional arrangements, 337–41, 340–41 trade facilitation, 342 interplay between regional and multilateral trade, 335 Trade-Related Aspects of Intellectual Property Rights, 390–97 provisions in preferential trade agreements, 341–43 U.S.-Canada Auto Pact, 131–32 regional initiatives, 327–31 U.S.-Jordan preferential trade agreement, 437–38 static effects, 331 U.S.-Peru Free Trade Agreement, 397 trade facilitation action plan, 343 worker rights provisions, 430–33 transit corridors, 337–40 Universal Declaration of Human Rights, 443–46 World Trade Organization facilitation, 329 Uruguay Round, 99, 133 Trade facilitation action plan, 343 U.S. Agency for International Development, 436, 439 Trade facilitation agenda, 28 U.S. Department of Labor, 436–38 Trade Facilitation Principles, 343 U.S. Digital Millennium Copyright Act of 1998, 396 Trade liberalization U.S. Federal Trade Commission, 359 regulatory objectives, 28–29 U.S. Food and Drug Administration, 198 strategies, 97–100 U.S. General Accountability Office, 437 Trade Policy Review Mechanism, 103–104 U.S. International Trade Commission, 449 Trade Promotion Act, 408, 456 U.S. Trade Representative, 397 Trade Promotion Authority Act, 155 USAID. See U.S. Agency for International Development Trade-Related Aspects of Intellectual Property Rights, 18, 390–401 USITC. See U.S. International Trade Commission Trade-Related Investment Measures Agreement, 20, 308 Trade remedy provisions V antidumping measures, 182–86, 194 Veto avoidance logic, 128–29 countervailing duties, 182–86 Viner, Jacob, 70 cross-regional pattern, 187–90 Visas, temporary, 288 diversity among preferential trade agreements, 181–84 function of, 180 W global safeguards, 182–86, 191–92 Walrasian model, 70–71 hub-and-spoke pattern, 187–90 Walvis Bay Corridor group, 330, 345n incidence of actions, 181 WCO. See World Customs Organization Index 511 West African Economic and Monetary Union, 154–55 human rights, 445–48 WIPO. See World Intellectual Property Organization labor rights, 430 Worker rights. See Labor rights rules on investment, 308 Workers Rights Project, 438 TBT and SPS standards and guidelines, 218 World Bank Institute, 403n trade facilitation, 329 World Customs Organization, 328–29 Trade Policy Review Mechanism, 103–104 World Health Organization, 403n Trade-Related Aspects of Intellectual Property Rights, 18, 390–401 World Intellectual Property Organization, 395 Transparency Mechanism for Regional Trade Agreements, 66n World Trade Organization WTO+ agenda, 17, 241 Agreement on Agriculture, 145–46 WTO extra, 17, 241 Agreement on Rules of Origin, 163 WTO. See World Trade Organization Agreement on Safeguards, 461n chronological development of preferential trade agreements, 39–42 Codex Alimentarius Commission, 201 Y Yale Law School, 438 disciplines on regional standards, 210–13 dispute settlement model, 473–74 Dispute Settlement Understanding, 467 Z Government Procurement Agreement, 372–76 Zoellick, Robert, 455 ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. The • 24 trees Office of the Publisher has chosen to print • 10 million BTUs of Preferential Trade Agreement Policies for total energy Development: A Handbook on recycled paper • 2,430 pounds of with 50 percent postconsumer fiber in accor- net greenhouse dance with the recommended standards for gases paper usage set by the Green Press Initiative, a • 10,959 gallons of nonprofit program supporting publishers in waste water using fiber that is not sourced from endan- • 694 pounds of gered forests. For more information, visit solid waste www.greenpressinitiative.org. TRADE AND DEVELOPMENT SERIES “The ongoing wave of preferential trade agreements (PTAs) strongly suggests a need to revise long-held conceptions about the multilateral trading system and to embrace a more complex world composed of networks of multilateral, plurilateral, regional, and bilateral arrangements. This handbook is a timely and much needed review of the analytical underpinnings and practice of modern PTAs, from the liberalization of agriculture trade to the promotion of human rights. I expect the handbook to be of particular relevance to trade specialists in the policy-making and academic communities confronted with these new and di cult policy challenges.� — ROBERT M. STERN Professor Emeritus of Economics and Public Policy University of Michigan “Preferential trade agreements are not only prevalent in today’s international trade system but most importantly have shaped the reform agenda of many developing countries in a number of behind-the-border areas. This handbook provides a comprehensive, systemic, and thorough analysis of PTAs by both addressing key conceptual issues underlying their negotiation and exploring their contents across several agreements around the world. The book makes a very valuable contribution for policy makers, practitioners, academics, and readers interested in further understanding the impact of deep integration PTAs from a development perspective.� — H.E. ANABEL GONZ�LEZ Minister of Foreign Trade Costa Rica “The impact of preferential trade agreements on development has been a priority policy issue for developing countries in recent years. This handbook gives a thoughtful and comprehensive overview of PTAs, their economics, and their implications for the multilateral trading system. It provides interesting reading for anyone concerned with the evolving international trade and development agenda.� — SUPACHAI PANITCHPAKDI Secretary-General United Nations Conference on Trade and Development (UNCTAD) About our Program The World Bank’s International Trade Department produces and disseminates policy-oriented knowledge products and forges partnerships on trade to advance an inclusive trade agenda for developing countries and to enhance developing countries’ trade competitiveness in global markets. Learn more about the World Bank’s trade portfolio at: www.worldbank.org/trade. ISBN 978-0-8213-8643-9 SKU 18643