20808 August 2000 Shahid Yusuf Weiping Wu Simon Evenett Editors Local Dynamics in an Era of Globalization T H E w 0 R L D B A Local Dynamics in an Era of Globalization 21st Century Catalysts for Development Edited by Shahid Yusuf, Weiping Wu, and Simon Evenett Published for the World Bank Oxford University Press Oxford University Press OXFORD NEW YORK ATHENS AUCKLAND BANGKOK BOGOTA BUENOS AIRES CALCUTTA CAPE TOWN CHENNAI DAR ES SALAAM DELHI FLORENCE HONG KONG ISTANBUL KARACHI KUALA LUMPUR MADRID MELBOURNE MEXICO CITY MUMBAI NAIROBI PARIS SAO PAULO SINGAPORE TAIPEI TOKYO TORONTO WARSAW and associated companies in BERLIN IBADAN © 2000 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, N.W., Washington, D.C. 20433, USA Published by Oxford University Press, Inc. 198 Madison Avenue, New York, N.Y. 10016 Oxford is a registered trademark of Oxford University Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or other- wise, without the prior permission of Oxford University Press. The Chinese characters displayed on the front cover of this book signify "globalization"-quan qiou hua. They were rendered by Jun Ma in the East Asia and Pacific Region of the World Bank. Cover design by W. Drew Fasick, ULTRAdesign. Typesetting is by Barton Matheson Willse & Worthington, Baltimore. Manufactured in the United States of America First printing August 2000 1 234504030201 00 The findings, interpretations, and conclusions expressed in this study are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance of such boundaries. Library of Congress Cataloging-in-Publication Data Local dynamics in a globalizing world: 21 st century catalysts for development / edited by Shahid Yusuf, Weiping Wu and Simon Evenett. p. cm. Includes bibliographical references. ISBN 0-19-521597-4 1. Economic development. 2. Globalization. 3. Decentralization in government. 4. Urbanization. I. Yusuf, Shahid, 1949- . II. Wu, Wei-p'ing. III. Evenett, Simon J. HD75 .L63 2000 338.9-dc2l 00-036342 Preface n preparing the World Development port, the workshops were a rich source of feedback on Report 1-999/2000, the team respon- the themes we were exploring, and the grant has enabled sible for the report commissioned a us to make these papers available to a wide readership. number of papers to map the several Many people have contributed to the making of this major fields covered by the report, to volume. We want to thank our fellow team members synthesize significant findings from Anjum Altaf, William Dillinger, Marianne Fay, Vernon recent research and to explore issues that were likely to Henderson, and Charles Kenny, who assisted in the loom large in the early 21st century. These papers, by commissioning of the papers and in reviewing earlier leading specialists, were reviewed at workshops in Wash- drafts. We are greatly indebted to Rebecca Sugui for her ington, Tokyo, and Singapore, and they served as some help in the processing of the papers and, together with of the building blocks for the report. However, from the Paulina Flewitt and Leila Search, assisting with the lo- very outset we saw these papers as having a life of their gistics of the workshops. Umou Al-Bazzaz provided in- own-independent of the report-so that all those with valuable assistance in the assembly of the final manu- an interest in local and global dynamics could draw upon script. The first workshop, held in Washington during the ideas and information presented in these papers. July 1998, helped to launch the World Development Re- This companion volume to the World Development port. We owe its success to the effort put in by Jean Pon- Report assembles a selection of papers that relate to the chamni and Mani Jandu. The two workshops in Tokyo unfolding of local dynamics in an increasingly in- were flawlessly co-organized by the staff of the World tegrated world environment. A second volume in Bank's Tokyo office together with the Overseas Eco- the World Bank's Discussion Paper Series will provide nomic Cooperation Fund of Japan (now a part of the a further selection on the themes of globalization and Japan Bank for International Corporation, or JBIC) and urbanization. the Foundation for Advanced Studies in International All the papers in this volume were extensively revised Development (FASID). We would like to thank the di- and edited prior to publication. We greatly appreciate rector of the Tokyo office, Shuzo Nakamura; Mika the effort put in by each of the authors in revising their Iwasaki; and Tomoko Hagimoto; and staff from JBIC contributions and working closely with us through the and FASID for the success of these events. lengthy editing process. The research in these papers, The workshop in Singapore was co-hosted with the the convening of the workshops, and the publication of Institute of Southeast Asian Studies. Thanks are due this volume was supported by a Population and Human to the institute's director, Professor Chia Siow Yue, for Resources Development Grant from the government of providing us with the ideal physical and intellectual Japan. This financial assistance was enormously valuable environment. throughout the preparation of the World Development In editing the manuscript, we have been greatly aided Report. It helped widen the knowledge base of the re- by Sandra Hackman. Contents The World of Globalizing Forces 1. Local Dynamics in a Globalizing World: 21st Century Catalysts for Development ...................... 2 by Shahid Yusuf, Simon Evenett, and Weiping Wu 2. Globalization, the World Trade Organization, and Development Strategies of Poorer Countries .... ........ 9 by Kym Anderson 3. Toward a Global Financial Architecture for the 21st Century ..................................... 19 by Robert E. Litan 4. Prerequisites for Successful International Monetary Policy Coordination ............................. 31 by Andrew Hughes Hallett 5. The Impacts of Emigration on Countries of Origin ............................................. 45 by Stephen Castles Managing Urbanization in a Globalizing World 6. Cities and Trade: External Trade and Internal Geography in Developing Economies .................... 58 byAnthonyJ. Venables 7. On the Move: Industrial Deconcentration in Today's Developing Countries .......................... 65 by Vernon Henderson 8. The Importance of Large Urban Areas-and Governments' Roles in Fostering Them ................... 69 by Edwin S. Mills 9. Toward a New Model of Rural-Urban Linkages under Globalization ................................ 74 by Yujiro Hayami 10. Global City-Regions and the New World System ............................................... 84 by Allen J. Scott v VI CONTENTS Making Decentralization Work in a GloblI Economy 11. How to Design a Fiscal Decentralization Program .............................................. 94 by Roy Bahl 12. Strategic Fiscal Decentralization in Developing Countries: Learning from Recent Innovations .... ....... 101 by Paul Smoke 13. Decentralization Comes to Japan . ......................................................... 110 by Kengo Akizuki 14. Why Are Social Funds So Popular? ................ ........................................ 114 by Judith Tendler Building Competitive Cities for the 21st Century 15. What Distinguishes Success among Second-Tier Cities? ........................................ 132 by Ann Markusen 16. Globalization and Urbanization: The Case of Thailand ......................................... 140 by Utis Kaothien and Douglas Webster 17. The Impact of Decentralization on Cities' Fiscal Health: Lessons from the United States ..... .......... 148 by Andrew Reschovsky 18. How Singapore Regulates Urban Transportation and Land Use ................................... 159 by Sock- Yong Phang 19. To Plan or Not to Plan: Southeast Asian Cities Tackle Transport, Communications, and Land Use .... .... 164 by Peter J. Rimmer and Howard Dick 20. Public Participation in Urban Environmental Management in Japan ............................... 169 by Ryo Fujikura 21. Coping with Municipal Solid Waste in Developing Countries .................................... 175 by David N. Beede and David E. Bloom Tables 2.1 Growth in World GDP and Merchandise Exports in Real Terms, 1720-1996 .10 2.2 Growth in Capacity for Electronic Commerce, 1991-2001 .10 3.1 Foreign-Owned Banks .21 3.2 Severe Banking Crises, 1980-96 .22 7.1 Distribution of Plants in the Japanese Electronics Industry .66 7.2 Shares of Population and Manufacturing for Selected Regions of the Republic of Korea, 1983 and 1993 ... 67 9.1 Annual Growth of Employment in Manufacturing in Taiwan (China) and the Philippines .76 10.1 The World's 30 Largest Urban Areas Ranked by Estimated Population in the Year 2000 .86 11.1 Forms of Intergovernmental Grant Programs ................................................ 97 11.2 The Components of a System of Fiscal Decentralization ....................................... 98 11.3 The Champions of Fiscal Decentralization .................................................. 99 13.1 Percentage of Public Revenue after Fiscal Transfer ........................................... 112 15.1 Second-Tier City Employment, Brazil, Japan, Republic of Korea, and the United States, Selected Years 1970-90 . ............................................................. 133 18.1 Public Sector Institutions and Agencies Involved in Urban Transportation and Land Use Activities ...... 162 19.1 Urban Agglomerations by Gross Regional Domestic Product, 1995 .167 CONTENTS VII JRgunres 3.1 Net Long-Term Resource Flows to Developing Countries, 1990-97 ........ ...................... 20 3.2 Trends in Global Finance: Growth of GDP, Trade, and FDI, 1989-96 ............................. 21 6.1 Internal Transport Costs and Regional Industrialization . ........................................ 60 6.2 External Trade Barriers and Regional Industrialization . ........................................ 61 6.3 External Trade and Internal Economic Geography ............................................. 62 10.1 World Distribution of Metropolitan Areas with Populations Greater than One Million .85 15.1 Alternative Regional Industrial Structures .134 16.1 Oversupply to Remain for a Long Time.144 Bolles 3.1 The Basle Capital Accord: A First Step Toward Regulatory Harmonization .25 3.2 Banking Reform Measures for Developing Countries .26 5.1 The Benefits and Costs of Migration .47 Maps 16.1 Dynamic Urban Regions: Thailand .141 The World of Globalizing Forces Pr ,l chapter Local Dynamics in a Globalizing World: 21st Century Catalysts for Development Shahid Yusuf, Simon Evenett, and Weiping Wu The The World Bank and Virginia World The Brookings Institution Commonwealth Bank University or those who entered the last decade devolution point to a declining faith in the capacity of of the 20th century believing that the central governments to deliver high-quality public ser- central elements of the development vices, promote sustainable economic growth, and equi- question had been settled in favor of tably distribute the benefits of such growth. Further- market capitalism, the ensuing tu- more, improvements in communications technologies mult was, to say the least, disquiet- have made it easier for local actors to express and, more ing. Few anticipated that the melding together of na- important, organize dissent. Though it is a political de- tional financial systems would lead to contagion being cision made by individual countries, the devolution of transmitted across continents. Indeed, when the Berlin power to subnational entities is widespread. Wall fell who would have thought that 10 years later a Managing "localization" is a central challenge for all different type of architecture-the global financial ar- countries-industrial and developing alike-during the chitecture-would be hotly debated in the corridors of coming decades. For many developing countries this is power? And yet even these well-discussed global forces compounded by the fact that central states have never pale in comparison to the dynamics unleashed within faced such comprehensive demands to cede powers and nations by increased pressures for local autonomy. resources, even though in many instances local actors may Two events in 1999 crystallize the forces that are re- not have the administrative experience to match their as- shaping the development landscape: the failure to launch pirations. Furthermore, even though the most embryonic a new global trade round at the World Trade Organiza- institutional framework may have yet to materialize in tion's Second Ministerial Conference in Seattle, and the order to reconcile the competing demands of disparate centrifugal-and all too often deadly-forces pulling elements of society. So strong is the sentiment for local power away from the central authorities in Jakarta to the autonomy that discontented regions are seceding in regions and islands of the vast Indonesian archipelago. greater numbers: Since 1990 on average 3.1 new nations The diverse set of demonstrators at Seattle highlighted were born each year, compared to the previous 30 years just how deep are social concerns aroused by interna- where on average the number of nations grew by 2.2 per tional trade flows. Irrespective of the merits of their case, annum.2 the ramifications of trade reform for the environment, The objectives of this volume are twofold: first, to an- food safety, working conditions, and democracy received alyze how these local dynamics and global forces are prominent attention in Seattle. In addition, the capacity transforming the development landscape, are raising new of the World Trade Organization to accommodate the questions for the development agenda, and are altering differing aspirations of a membership that now exceeds the scope and effectiveness of policy alternatives. The 135 members was called into question. second goal is to propose ambitious yet feasible measures Both the increasingly violent protests in parts of In- that will enhance the prospects for developing nations donesial as well as the votes in Scotland and Wales for in the 21st century. In rising to these admittedly ambi- 2 LOCAL DYNAMICS IN A GLOBALIZING WORLD: 2IST CENTURY CATALYSTS FOR DEVELOPMENT 3 tious goals, the contributors to this volume, all promi- century and even longer if the definition is broadened nent scholars, have drawn upon the experiences of de- to embrace political globalization.3 Growth of interna- veloping and industrial countries from every continent. tional trade and capital flows, which are central to glob- A recurring theme in this volume is that traditional alization as conventionally understood, accelerated frameworks for analysis need to be rethought to take ac- markedly in the latter part of the 19th century. There count of greater interdependencies across space, time, was a sharp fall in shipping costs between 1870 and and issues. In what follows, perhaps the most prominent 1900 and a substantial narrowing of the spread of sta- example is urbanization, a phenomenon that is spurred ples prices on world markets.4 Moreover, between 1820 on by intensifying local dynamics and where integration and the first decade of the 20th century, nearly 60 mil- of cities into the world economies opens the door to lion people migrated to the Americas alone.5 In fact, faster growth, by tapping foreign funds and expertise. several observers maintain that the current phase of How these interdependencies evolve over time is of criti- globalization has yet to surpass the high point reached cal importance to humanity as at the turn of the century around the beginning of the 20th century. But a closer approximately one-half of the world's population lives in look at evidence indicates significant differences, quan- urban areas, and by 2025 this fraction is expected to rise titatively, qualitatively, and technologically. to two-thirds. The new economic geography literature, International commerce now encompasses a much which analyzes the role of cities and regions in a global larger share of tradable commodities. This share has economy, is one example of the type of development risen from 20 percent in the late 19th century to more thinking that could grow in importance. than 40 percent at the end of the 20th century.6 More- A second theme, which receives considerable promi- over, the trade in services has swelled from insignificance nence, is that the number of competing decisionmakers to nearly one-fifth of total trade-it amounts to 40 per- on the development landscape will multiply, calling for cent of the value of U.S merchandise exports. It is also the reform and often creation of new modes of gover- growing much faster than the trade in goods. Close to nance. This involves a shift away from a central, state- two-thirds of all trade is within production networks led view of development toward a more complex, and created by transnational corporations through alliances potentially untidy, coalition of actors at both supra- and and subcontracting relationships. The increase in trade subnational levels who will typically not confine them- has been supported by the steady reduction in tariff bar- selves to operating in one arena, such as the local sphere. riers, a result of seven rounds of trade negotiations over Given the growing linkages across issues, actors in any the past 50 years. Tariffs on manufactured imports in one sovereign unit are likely to find themselves dealing the industrialized countries have fallen by 90 percent with one another on matters that were not previously since World War II to an average of 4 percent.7 envisaged when their nation's constitutional arrange- Overseas capital flows into government bonds and ments were decided. This enlarges the importance of in- corporate debentures rose strongly from the mid-19th stitutional mechanisms permitting wider participation, century through the first decade of the 20th in spite of the orderly resolution of disagreements, and for induc- periodic panics and crises triggered by the collapse of ing compliance with collective decisions. major private ventures or government defaults. But the A third theme is that the combination of local dynam- current flow of capital is of a different order. The mag- ics and global forces can catalyze policy action, enabling nitude has increased manyfold and the composition has regions and cities to accelerate the pace of development. changed drastically. Financial globalization is reinforc- This is especially so for cities which build the appropriate ing the impact of the communications revolution on infrastructure and amenities to attract high-valued added global linkages. In fact, 80 percent of all international service industries and personnel from abroad. Two other data flows are generated by the financial services sector.8 implications are also worth noting: Rapid development Foreign direct investment, mainly by transnational need not occur and to the extent that it does it may well corporations, has emerged as a major item and comprised widen the gap between richer and poorer regions of the 48 percent of long-term capital flows to developing coun- world. The difference between success and failure will al- tries in 1997.9 Institutional and individual investors are most certainly become greater than ever before, and that also a source of the growing volume of private portfolio in and of itself may well focus the minds of policymakers. capital and short-term flows, and these have overshad- owed bank lending and investment in government paper. Foreign direct investment is being spurred by the easing The term globalization is of recent vintage, but the of barriers to trade, the advantages of situating produc- process it describes has been under way for well over a tion facilities close to major markets, and the greater ease 4 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION in managing dispersed operations brought about by re- of trade negotiations. The Uruguay Round envisaged cent advances in communications. Firms need no longer significant reductions in trade barriers, with sizable po- design, manufacture, assemble, and market a good in the tential gains for both industrial and developing coun- same nation; and developing nations will increasingly find tries. But given the complexity of the issues and the po- themselves evaluated as a potential location for only stages litical economy underlying the negotiations, a number in production processes, with ramifications for the nature of items were bypassed or covered only peripherally. Up- and extent of foreign direct investment that they receive. permost among these are rules for a phased liberalization The new phase of globalization is also notable for the of trade in textiles, agricultural commodities, and ser- changes in human mobility. Although the heightened vices. The tensions that surfaced at the Seattle meetings flows of goods and capital, together with advances in indicated that the members of the World Trade Organi- communications, might have been expected to dampen zation are still struggling to implement the agreements human migration, the trend is in the opposite direction. reached through the Uruguay Round. Further lowering More people are on the move than ever before. Each of trade barriers in key areas is stoutly resisted by vested year between 2 and 3 million people emigrate, with the interests, and extending multilateral disciplines to en- majority of them going to just four home countries: the compass producer services is a source of concern, espe- United States, Germany, Canada, and Australia, in that cially for developing countries. In addition, organized order. At the turn of the 21st century, nearly 120 mil- groups, particularly the international nongovernmental lion people are living outside countries of their birth. organizations (NGOs), are agitating to bring environ- Better information, the lower cost of transport, increased mental and labor-related standards into the ambit of the education, rising incomes, and a readiness to exploit trading system. worldwide opportunities are encouraging people to Financial reform and managing capitalflows. The in- travel and to migrate. The emergence of a global culture creasing frequency of banking and currency crises dur- and the spread of English as a lingua franca has further ing the 1980s and 1990s, culminating in the East Asian reduced inhibitions to movement.10 Crisis of 1997-1998, has highlighted the role of the fi- Cross-border migration will be one of the major forces nancial system in economic development. These crises shaping the landscape of the 21st century, for at least also revealed the effects of capital flows on the growth three reasons. First, migration is causing dramatic shifts and stability of financial sectors in industrializing coun- in the demographic profiles of both industrial and de- tries and the necessity of making adequate regulatory veloping countries, in particular in industrial and East- provisions when removing domestic controls on the ern European nations where fertility rates are low. Sec- banking system-especially where domestic financing is ond, the movement of highly skilled people from the combined with an easing of restrictions on international developing world affects low-income countries and re- capital flows. Robert Litan's chapter succinctly reviews cipient countries alike. Here traditional concerns about the case for actively pursuing financial deepening and the "brain drain" from developing nations are to be containing the risks of increasingly disruptive shocks. weighed against newer research that shows the benefits To be effective, institution building will have to be con- to developing nations of the remittances from national ducted simultaneously at three levels: at the level of in- residents abroad. At present such remittances exceed dividual financial entities, which must tighten their in- US$70 billion and are worth 50 percent more to devel- formation and risk management practices to ensure oping nations than official aid. Third, the international their viability in a globalizing environment; at the coun- diaspora has tremendous business potential, facilitating try level through the defining and enforcement of rules the dissemination of market opportunities, funds, and that, for instance, set benchmarks for capital adequacy, best practices across national borders. help sustain franchise value, and orchestrate a process of Given the accelerated and unpredictable pace of liberalization; and at the international level through change, the critical question is how to develop rules, in- a revamped Basle Accord backed up by monitoring ar- stitutions, and policies that enable nations to respond rangements that will result in commonality of standards to the many opportunities and risks presented by glob- and adherence to good regulatory practices. alization as it unfolds. In this volume, the following four Coordinating monetary policies. While multilevel regu- chapters delineate reforms that will raise the likelihood latory institutions will be essential for stability under of developing nations benefiting from globalization. conditions of financial globalization, they will need Rules governing international trade. Kym Anderson ac- to be complemented by some degree of policy coordina- counts for the factors that determined the current mul- tion both between major countries and at the regional tilateral trading system, the legacy of the Uruguay Round level where international linkages are greatest. Andrew LOCAL DYNAMICS IN A GLOBALIZING WORLD: 2IST CENTURY CATALYSTS FOR DEVELOPMENT 5 Hughes Hallett explores the desirability of and prerequi- the growth of services and tourism. In some countries, sites for coordination of monetary policies, the recent such as China, openness has also prompted rural devel- record, and the most fruitful pathways for the future. opment by enabling rural producers to attract foreign During the 1 980s and 1 990s, exchange rate volatility be- direct investment, to expand processing industries, and tween key currencies had significant national and re- to enter foreign markets. gional consequences, but attempts to harmonize macro- The globalizing world is a world of cities and their economic policies met with limited success. Hughes hinterlands. Anthony Venables draws upon the models Hallett emphasizes that successful international coopera- generated by the new economic geography to explore tion requires flexible domestic institutions which can ac- the spatial structure of development. He shows how an commodate the macroeconomic policies needed to sup- initial clustering of urban activity linked to agglomera- port the cooperative endeavor. Care must also be taken tion effects could be followed by dispersion as econo- in devising cooperation agreements so that the gains are mies open up and expand their trading links, transport not asymmetrically distributed across participants. costs decline, and congestion offsets the benefits of con- Harnessing international migration. After providing an centration in one or a few large urban centers. However, overview of the principal trends in international migra- he notes that economic development is likely to be an tion, Stephen Castles' chapter discusses the many costs uneven process that is concentrated in a few areas that and benefits associated with the movement of people have been quick to grasp opportunities, create backward across national borders. Of particular interest are two and forward linkages, and respond to declining transac- case studies, on Turkey and the Philippines, where it is tion costs. Venables' chapter sets the stage for the four argued that migration has contributed more to the lat- chapters that follow, each of which examines a particu- ter's development strategy than to the former's. One rea- lar facet of urban development. son for this difference is that the Philippines has taken Vernon Henderson examines the emergence of a hi- extensive measures to facilitate the reintegration of ex- erarchy of cities and the role of megacities in develop- patriates into their home country. This has encouraged ing countries. Henderson points out that megacities the transfer of funds, essential for establishing small and might initially dominate the landscape but rising costs medium-sized businesses, as well as transferring practices in such cities along with the building of transport infra- and tapping global contacts. Drawing from these expe- structure can expedite the dispersion of activities to riences Castles proposes a series of international mea- medium-sized cities with a more specialized production sures that could enhance the benefits of cross-border mi- structure. Edwin Mills extends the points made by Hen- gration. Should policymakers wish to redress the striking derson by showing how the productivity advantages of imbalance between the extensive rules on trade and cap- large urban areas are dependent upon policies that de- ital flows and the paucity of such rules on the movement termine the quality of infrastructure and effective man- of people, then Castles' chapter provides considerable agement of land use. food for thought. Globalization also opens up new markets to hinter- Localization land producers previously more tied to domestic mar- kets than producers in the vicinity of coastal regions. Yu- Global integration frames and, to a degree, spurs the on- jiro Hayami discusses how the emergence of trading going localizing tendency. The multiplication of inter- opportunities was responsible for rural industrialization national rules, standards, and conventions; the growing in Japan and Taiwan (China) in the earlier part of the influence of multilateral agencies; transnational corpo- 20th century and more recently in some of the South- rations and nongovernmental organizations; and the east Asian economies. He also explores the significance force of global public opinion reinforced by modern of urban-rural trading linkages that have served as the communications have all circumscribed the authority of relays transmitting global demands to the rural areas. national governments. Greater openness has also en- Hayami's contribution usefully complements others in larged the freedoms enjoyed by subnational entities to this section by showing that the growth impulse deriv- exploit trading opportunities, to raise funds on overseas ing from globalization can lead to a more balanced de- markets, and to cultivate direct relations with interna- velopment of rural and urban areas. tional agencies and cities. Over the past quarter century, another form of ur- The economic effects of openness and easier travel banization has acquired prominence. Allen Scott ana- have spurred urban economies, thereby reinforcing ur- lyzes the formation of economic city-regions through a banization in developing countries. Direct gains have coalescence of growing, outward-oriented metropolitan been observed from exports, and indirect benefits from areas. He also suggests that globalization enhances the 6 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION possibilities of increased geographic differentiation and deal with the flow of goods, capital, labor, and informa- locational specialization. A new set of industries, includ- tion among countries, so also localization raises a host ing high-technology manufacturing, business and finan- of policy and institutional issues relating to decentral- cial services, neoartisanal industries, and cultural prod- ization and the creating of competitive and livable cities. ucts industries, are becoming especially important in Effective decentralization. This involves a number of major world metropolises such as New York, London, steps that distribute responsibilities between the center Paris, and Los Angeles. The economies of scope and and subnational entities, assign sources of revenue com- scale of city-regions are reinforced by local labor market mensurate with these responsibilities to different levels processes, by innovation and learning effects, and by the of government, create administrative capacity to dis- conventions, cultures, and interdependencies that con- charge the newly assigned functions, and establish the stitute intrinsic elements of any active industrial system. ground rules for interaction between the center and The concentration of economic power in cities and lower level governments in a decentralized milieu. regions has strengthened demands for decentralization- Rules for decentralization and recent experience in a fiscal and administrative-demands that have drawn im- number of countries that could guide other economies petus from political trends in industrial countries, where travelling down this path are the topics of the chapters the momentum toward decentralization has been gath- by Roy Bahl and Paul Smoke. Bahl synthesizes a wide ering force for some time. This is now finding an echo range of experience into 11 rules on the pace and nature in developing countries where central governments are of decentralization, taking account of the capacity in acceding to the pressure to devolve authority to lower most developing countries. He emphasizes a rule-bound levels. However, in several instances, revenue constraints process, kept simple and continuously monitored, which are also encouraging central authorities to shed obliga- gives local authorities significant taxing powers, but then tions to subnational entities by way of decentralization. subjects them to hard budget constraints. Similarly, Paul Political openness and a heightened awareness of eth- Smoke draws lessons from the decentralization episodes nic or regional identity also have contributed to the in Brazil, Ethiopia, Indonesia, Kenya, and Uganda; con- trend toward greater political and administrative respon- cluding that although there is no single model for fiscal sibilities being placed on the shoulders of local authori- decentralization, seven principles should form the build- ties. The importance of local development and gover- ing blocks of any strategy to accommodate increased nance is further enhanced by the assertiveness of civil pressures for local autonomy. But the key rule of decen- society in many countries. Such a focus on local devel- tralization is that it must respond to the specific attri- opment is conducive to greater participation, as the in- butes of the country and local communities. centives are stronger and the consequences of action Successful decentralization also is contingent upon taken are more visible and immediate at the local level. effective local governance. Accountability of local offi- But in order to ensure that newly empowered local gov- cials to the population determines whether the fruit ernments are in fact responsive and accountable to their of decentralization will be realized in the form of more constituents, transparency and incentives for democratic responsive provision of local services and local self- participation will be required. determination. Kengo Akizuki describes Japanrs steps to- With increased levels of political autonomy and demo- ward devolving authority down to prefectures and lower cratic participation, municipalities are now facing greater levels of government since the mid- I 990s. He discusses challenges in delivering services to urban populations. The the bureaucratic restructuring that this is entailing, the exclusive responsibility of the public sector to provide a administrative capacity gaps that still exist, but also the range of services is giving way to private participation, alacrity with which prefectures are grasping the possi- public-private partnerships, and community initiatives. bilities offered by globalization. Reaffirming the importance for municipalities of develop- Judith Tendler reminds us that decentralization ini- ing appropriate institutions that involve the private sector, tiatives do not necessarily lead to improved quality of community-based organizations, and nongovernmental public-service delivery to local constituencies, particu- organizations, promising improvements have emerged in larly those less advantaged. She challenges the effective- several developing countries in all major sectors of urban ness of social funds-programs usually administered by services. It is also apparent that for many municipalities a separate agency within a central government that allo- the new responsibilities entail considerable financial hard- cates grants directly to myriad communities-as a safety ship and stretch, if not exceed, their technical capabilities. net for the poor and, more significantly, about the Just as globalization requires the community of na- greater desirability of social funds as an alternative to tra- tions to devise policies and constitute institutions to ditional government provision. At best, such funds may LOCAL DYNAMICS IN A GLOBALIZING WORLD: 21ST CENTURY CATALYSTS FOR DEVELOPMENT 7 represent a deconcentrated version of supply-driven ser- credits Singapore's effective implementation of such vice that constructs small infrastructure projects more planning to a network of competent and reliable orga- quickly at lower costs. nizations that implement public policies. However, the Managing urbanization. As engines of growth, cities transferability of Singapore's experience needs to be jux- are becoming the focal points of global and national taposed with local political and social context because economies, despite the likely strain on urban governance of the island city's unique governmental system and its and infrastructure. ability to control the entry of migrants who can impose Ann Markusen's study of secondary cities in Brazil, a strain on urban services. Japan, the Republic of Korea, and the United States In comparing two sets of large cities in Southeast points to some of the prominent contributors to suc- Asia, Rimmer and Dick reaffirmed the need for in- cessful urbanization. These include specialization in in- tegrated land use and transportation planning. For cities novative or income elastic economic sectors, favorable that have applied modern transport technologies piece- national government policies in such areas as investment meal and uncoordinated by any consistent long-term and infrastructure, economic development leadership strategy, the outcome has inevitably been heavy con- from either local public or private sectors, and flexibil- gestion and urban sprawl. The mobility of the urban ity and openness toward new and emerging industrial poor has been impeded and dangers from vehicle traffic ensembles. National governments should continue to worsened. The authors further suggest that although monitor urban growth patterns, selectively intervening cities of varying sizes often need to focus on different in dynamics that are worsening regional disparities. priorities, all would gain from greater access to local re- Adding to this list, Utis Kaothien and Douglas Webster sources and more responsiveness to local constituencies. believe that cities must address quality of life, develop One option for large cities-a technique that Japan has human resources, harness global capital flows, and focus used with success-is to encourage private companies on software (policy and programming) instead of hard- to finance infrastructure by relying on earnings from re- ware (physical systems). But municipalities have diffi- lated development. culty in achieving multiple objectives because of the Smart cities are efficient at providing services and massive scale of periurbanization-as in the case of at putting in place infrastructure sufficient to achieve Thailand-and the multiplicity of local governments in competitiveness in an open environment. Cross-country the extended urban regions. experience with municipal development suggests that Successful urban development in the context of de- the quality of local governance, especially the ability centralization must come to terms with mobilizing fi- of local authorities to establish partnerships with the nancing for municipalities and long-term financing for private sector and community groups and to mobilize infrastructure. Municipal revenues are still mainly de- organizational skills, strongly influences the degree to pendent upon transfers from provincial and central au- which municipalities deliver on their promises. In a re- thorities. But there are growing incentives to utilize local vealing chapter on Japan's urban environmental move- tax handles such as property and sales taxes and user ments, Ryo Fujikura takes up the interlinked issues charges for municipally supplied services. Success in this of governance and environmental quality. In particular, regard depends on the building of administrative capac- Kitakyushu City's experience shows the importance ity to levy and collect both taxes and user charges, as ex- of public awareness and environmental education, emplified by the experience of the United States in An- spearheaded by women's groups, in building support for drew Reschovsky's chapter. Term financing rests in large environmental protection. Local leaders also under- part on the ability to tap bond markets. Globalization stand their communities better than higher-level gov- is enhancing access to such financing, but cities must ernments, and careful advice from local officials to in- first fulfill certain criteria and establish a good policy dividual firms has solved many problems of industrial and payment record. pollution. Building livable cities. Urban success requires strate- The issues of service provision by municipalities are gic planning, which can guide key investments and as- further pursued by David Beede and David Bloom, sist in allocating resources. Backed by integrated land specifically in solid waste management. Economic use and transport policies that accommodate private ini- growth, poverty reduction, and good governance are tiative to fill the gaps, the process could help cities avoid likely to enhance countries' capacities to address the the worst outcomes of unplanned growth. This has waste problem. Among the available solutions, recycling clearly been a key source of Singapore's success, as dis- and composting may be better alternatives for most de- cussed in the chapter by Sock-Yong Phang. She also veloping cities as they also capitalize on the resource 8 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION value of the waste. But composting will not fulfill its development opportunities presented by a globalizing promise until both quality control and demand improve world of urban regions. significantly. Many economic schemes-including charg- ing service fees, deposit-refund mechanisms, and pri- Notes vate concessions-can help municipalities pay for proper 1. Economist "Brutality and Budgets in Indonesia," January management of solid waste and cut the amount illicitly 22, 2000; Far Eastern Economic Review, "Bleak Prospects," De- dumped into the natural environment. cember 16, 1999, discusses the effects of a breakdown in law and order in Indonesia in its mining industry. Concluding Observations 2. Juan Enriquez, "Too Many Flags." Foreign Policy, Fall 1999, pp. 30-50. The contributions in this volume identify some of the 3. David Held, Anthony McGrew, David Goldblatt, and major forces transforming the world economy in the Jonathan Perraton. 1999. Global Transfornations. Palo Alto, CA: early 2 1 st century. They also provide a framework for Stanford University Press. thinking about the challenges ,,o,,fronting policymk 4. Kevin H. O'Rourke and Jeffrey G. Williamson. 1999. thinking about the challenges confronting policymak- Globalization and History. Cambridge, MA: MIT Press. ers. They offer tools for analyzing key problems and 5. Timothy J. Hatton and Jeffrey G. Williamson. 1998. The suggest policy or institutional initiatives that allow Age ofMass Migration. New York: Oxford University Press. countries to make globalization and localization yield 6. Michael D. Bordo, Barry Eichengreen, and Douglas A. benefits rather than becoming obstacles to development. Irwin. 1999. "Is Globalization Today Really Different than The integration of economies through trade, capital Globalization a Hundred Years Ago?" In Brookings Trade Pol- icy Forum, Washington, D.C.: Brookings Institution. flows, and communications, the emphasis on local au- 7. Business Week, "Global Growing Pains," December 13, tonomy, the emergence of regional economic entities, 1999. and the growing political and economic salience of 8. Graham, Stephen. 1999. "Global Grids of Glass: On urban centers cannot be viewed in isolation. The inter- Global Cities, Telecommunications and Planetary Urban Net- relationships need to be recognized and factored into the works." Urban Studies 36, 5-6 (May): 929-949. 9. World Bank. 2000. Global Economic Prospects 2000, Table making of policy. 'while each author concentrates on a A2.10, Washington, D.C. particular aspect of a large picture, taken together, the 10. David Crystal. 1997. English as a Global Language. Carn- 21 chapters offer a wide-ranging perspective on the bridge, U.K.: Cambridge University Press. chapter 2 Globalization, the World Trade Organization, and Development Strategies of Poorer Countries Kym Anderson School of Economics and Centre for International Economic Studies, University of Adelaide, Australia : he pervasive effects of the process of in the fourth section; and the final section of the chap- globalization are being felt in poorer ter explores the implications of these developments for I q economies at least as much as in policies and strategies of developing countries, includ- j richer ones. Globalization is also being ing the poorer economies in transition from socialism. affected by, and is affecting, interna- Particular attention is given to rural development strate- * -: tional institutions such as the World gies because globalization is likely to strengthen urban- Trade Organization (WTO). The WTO's evolution in ization and thereby add to the burden of adjustment turn is altering the policy options of developing and that rural areas traditionally bear. transition economies. In addition, the ability of those The implications of globalization for poorer coun- countries to influence the WTO is beginning to in- tries, and for the poorer rural areas within them, are of crease. This three-way interaction between globaliza- significance to people in rich countries as much as to tion, the WTO, and the development policy strategies those in poor countries. This statement is true because of developing and transition economies is explored in globalization threatens to increase the gap between the this chapter, with a view toward drawing out implica- haves and have-nots to such an extent as to cause politi- tions for further action by governments of the world's cal tensions within and between countries and even a developing countries as the next century dawns. Those backlash against a key contributor to globalization, actions certainly involve getting rid of, or not introduc- namely, the opening up of national economies. As ing, certain types of government interventions in mar- David Landes (1998, p. xx) put it in his latest book: kets, especially those whose inappropriateness is increas- ing with globalization. But attention is also given to the The old division of the world into two power question of what new government actions may be re- blocs, East and West, has subsided. Now the big quired-at international as well as at national levels- challenge and threat is the gap in wealth and to maintain or improve the smooth functioning of eco- health that separates rich and poor. These are often nomic and political markets. styled North and South, because the division is ge- The chapter has five sections. The first defines glob- ographic; but a more accurate signifier would be alization and briefly describes indicators of its growth; the West and the Rest, because the division is also the second summarizes technological aspects of global- historic. Here is the greatest single problem and ization of relevance to poorer countries; the third notes danger facing the world of the Third Millennium. the changes in national economic policies that have con- tributed to globalization; the General Agreement on Tariffs and Trade (GATT) and now the World Trade Definition and Extent of Globalization Organization (WTO) have added significantly to those For present purposes, globalization is defined as the de- governmental contributions to globalization, as discussed cline in transactions costs or barriers to doing business 9 10 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Table 2.1 Growth in World GDP and Merchandise Exports in Real Terms, 1720-1996 (percent per year) 1720-1820 1820-70 1870-1913 1913-50 1950-73 1973-90 1990-96 Real GDP 0.8 1.9 2.5 1.8 5.4 2.7 2.9 Export volume 1.4 4.5 3.9 0.5 9.8 4.0 5.9 Export growth/GDP growth 1.7 2.4 1.6 0.3 1.8 1.5 2.1 Source: WTO (1998, p. 34) or otherwise interacting with people of other nations number of telephone lines, a 25-fold increase in the num- around the world. Its effect is to enhance the integra- ber of cellular phones, a near quadrupling in the number tion of markets for goods, services, technology, ideas, of personal computers, and an expectation that two-thirds capital, and labor, reducing the differences in prices for of those PCs will have Internet access by 2001. those products and factors across space. Both techno- The Technological Dimension of Globalization logical and governmental barriers contribute to the costs of interacting internationally. Falling transport costs, the There have been three technological revolutions in huge decline in communication and information costs, transport and communication costs in modern times. and cuts in tariff and nontariff barriers to trade in goods, The cost of transporting goods was lowered enormously services, financial capital, and, to some extent, labor in the 19th century with the advent of the steam engine, have combined in the late 20th century to accelerate which created the railway and steamship. Steel hulls for globalization to an unprecedented speed that shows no ships and refrigeration further lowered the cost of ocean sign,of abating. transport late in the 19th century, particularly for per- The extent of the acceleration in globalization cannot ishable goods. The telegraph helped as well (O'Rourke be captured in a single statistic, but several provide par- and Williamson 1999). The second technological revo- tial indications of what is involved. A standard indicator lution also greatly lowered the cost of moving people. It is the comparison between trade and gross domestic prod- was dominated in the middle half of the 20th century uct (GDP) growth. As Table 2.1 shows, while merchan- by the falling cost of transport by car and airplane, dise trade has grown faster than output for all periods ex- caused by mass production of such goods and associated cept between the two world wars, the gap has been larger services. Ocean freight rates (helped by containeriza- in the 1990s than in any earlier period since the mid- 19th tion) and telephone charges also fell massively over this century. More than one-fifth of global output is now ex- period.' The third revolution in transport and commu- ported, double the proportion in the 1950s. Also, annual nications technology, at this end of the 20th century, has outflows of foreign direct investment grew more than six- been digital. Aided by deregulation of telecommunica- fold between 1983 and 1990 and have continued to grow tions markets in many countries, the digital revolution more than twice as fast as goods trade in the 1990s. In- trafirm trade among multinational corporations is esti- mated to account for one-third of world trade, and an- Table 2.2 other one-third is multinational corporation (MNC) Growth in Capacity for Electronic Commerce, trade with nonaffiliates. During the 1990s international 1991-2001 portfolio investment was growing equally as fast as for- (million units) eign direct investment (WTO 1996, pp. 44-46). In just the past five years, the annual value of cross-border merg- 1991 1996 2001a ers and acquisitions has trebled, growing from US$100 Telephone main lines 545 741 1000 billion to US$300 billion (UNCTAD 1998, p. 21). Daily Cellular phone subscribers 16 135 400 foreign exchange transactions now exceed global currency Personal computers 123 245 450 Internet host computers 1 16 110 reserves, with international capital flows more than 50 PCs with internet access 5 60 300 times the value of international trade flows. The 1990s a Projected by the International Telecommunications Union, Geneva have also seen an explosion in the world's capacity for Source: WTO (1998b, p.8). electronic commerce. Table 2.2 shows a doubling in the GLOBALIZATION, THE WTO, AND DEVELOPMENT STRATEGIES OF POORER COUNTRIES II has enormously lowered long-distance communication not in place. Two examples, to be discussed later, have costs, and especially the cost of rapidly accessing and to do with prudential regulation of the financial sector processing knowledge, information, and ideas from any- and environmental concerns. There is also a risk that the where in the world.2 A side effect of the Internet's ex- market-opening reforms of the postwar period, and es- pansion has been the growth in the use of the English pecially the past 15 years, could be reversed by govern- language. It has been claimed that there are now more ments as domestic political circumstances change. As ex- people using English as a second language than there are plained in the next section, during the past 50 years that people for whom it is a first language (Cairncross 1997). risk has been contained by the GATT and, since 1995, This too is lowering costs of communicating between by the WTO. countries. Together with the technological revolution, these The Governmental Contribution to Globalization policy reforms have brought about a more integrated global trading system, a global capital market, and more The previously mentioned developments have been re- international transactions that formerly took place be- inforced by government decisions to liberalize trade and tween independent entities are being internalized within investment regimes. Following the protectionist inter- single firms or corporate alliances. The increasing mo- war period, this liberalization began with the lowering bility of the productive assets of firms enables them to of import tariffs on trade in manufactures between in- minimize their corporate income tax exposure by strate- dustrial economies. Within Western Europe that trade gically locating their headquarters and using transfer was especially liberal following the Treaty of Rome and pricing in their intrafirm international trade. This mo- the formation of the European Free Trade Area. In the bility also encourages governments to compete for the 1980s trade reform was followed by extensive liberaliza- presence of firms with regulatory reforms and invest- tion of foreign exchange markets and of restrictions on ment incentives (including restraints on the adoption of capital flows, leading (with the help of new digital tech- higher environmental and labor standards). On the one nologies) to the development of new financial instru- hand, this could leave governments with less tax revenue ments. At the same time many non-Organisation for to implement social policies at a time when the demand Economic Co-operation and Development (OECD) for such policies is rising with income growth and with countries-including China and ultimately the Soviet disruptions in the market for low-skilled labor. On the bloc-began moving away from inward-looking to other hand, if regulatory reform is growth enhancing outward-oriented trade and investment policies. The and includes the privatization of state-owned enter- 1980s also saw the deregulation of domestic markets in prises, government revenue could expand. a growing number of countries, which reinforced the ef- These technological and governmental revolutions fects of deregulating transactions at national borders. have contributed increasingly to the drift toward urban- These reforms benefit most the countries making ization. The first helped launch the industrial revolution them, but they also benefit their trading partners. Hence in Western Europe, but this was done partly by lower- the more countries open up and reform, the greater is ing the cost of exploiting natural resources abroad, the gain to other countries from doing likewise. In par- which also allowed primary sectors in less densely pop- ticular, they expand the opportunities for developing ulated and tropical countries to expand. The second rev- and transition economies to access goods and services olution accelerated industrialization in the West and its markets, investment funds, and technologies, thereby spread to the Far East, including by means of what Ver- raising the payoff to those economies for joining the non (1966) described as the product cycle. The current bandwagon of liberalization. Those that have already revolution is increasing the scope to subdivide the done so have grown much faster than the rest and have processes of production and distribution into parts that seen their incomes converge toward OECD income lev- can be relocated anywhere in the world according els (see, for example, Dollar 1992; Edwards 1993; Sachs to ever-increasing changes in comparative advantages and Warner 1995; and WTO 1998a, pp. 62-63 for a over time. That outsourcing can be via various means bibliography). The reasons for faster growth of more including subcontracting, licensing, joint ventures, and open economies have to do with the dynamics of trade direct foreign investment by multinational corporations liberalization, which are not just predictions derived (Markusen et al. 1996). from new trade and growth theories (Grossman and The resulting productivity growth in the industrial Helpman 1991) but that are well supported empirically and service sectors is altering the key source of wealth of (USITC 1997). However, greater openness can carry nations, which is moving ever faster away from natural some risks, especially if appropriate domestic policies are capital to human capital (that is, from raw materials and I2 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION physical capital per worker to human skills and knowl- ment on Tariffs and Trade was signed by 23 large trading edge). In particular, wealth creation in the 21st century countries in 1947. The GATT provided not only a set of will depend especially on the ability to access and make multilateral rules and disciplines but also a forum within productive use of expanding stocks of knowledge and which to negotiate tariff reductions and rules changes. It information, and to build on them through creative re- also provided a mechanism to help settle trade disputes. search and development (World Bank 1998). How well Eight so-called rounds of negotiations took place in the and how quickly people of different countries are able subsequent 46 years, the last one (the Uruguay Round) to do that will increasingly determine relative economic culminating in the "interim" GATT Secretariat being growth rates. But for all countries the extent and speed converted into the World Trade Organization. with which economic events abroad are transmitted to The GATT, and now even more so the WTO, con- domestic markets will increase inexorably, and govern- tribute to globalization in several crucial ways. The ments will have less and less capacity to isolate their WTO has four key objectives: to set and enforce rules economies from such trends, as derivatives and elec- for international trade; to provide a forum within which tronic commerce have made clear in the cases of inter- to negotiate and monitor trade liberalization; to im- national financial flows and a widening range of traded prove policy transparency; and to resolve trade disputes. goods and services. Apart from the transparency role, these were also the key objectives of its predecessor before the WTO came into being; but the WTO is much more comprehensive than History shows that the risk of market opening being re- the GATT. For example, GATT's product coverage in versed is much more likely in the absence than in the practice was confined mainly to manufactures (effec- presence 6f international constraints on national trade tively not including textiles and clothing), whereas the policy actions. For example, the Cobden-Chevalier WTO encompasses all goods (including.sensitive farm Treaty of 1860 between England and France contained products), services, capital to some extent, and ideas (in- a most-favored-nation (MFN) clause which required tellectual property). In addition, following the conclu- that the agreed cut in the tariff on each item in their bi- sion of the Uruguay Round negotiations, the interim lateral trade was also to be applied to their imports from GATT Secretariat was converted to a permanent WTO other countries. This clause also meant that every Euro- Secretariat with greatly strengthened trade policy review pean country that subsequently signed a trade treaty and dispute settlement mechanisms. It also has a new with either England or France (and most did by 1867) role: cooperating with the International Monetary Fund signed onto the MFN treaty. The effect was a network (IMF) and World Bank with a view toward achieving of treaties that greatly lowered the level of tariff protec- greater coherence in global economic policymaking. tion in Europe (Kindleberger 1975), allowing world GATT/WTO rules to govern international trade output and trade to boom for several decades until the serve at least three purposes. First, they protect the wel- First World War intervened (Table 2.1). Following that fare of small and weak nations against discriminatory war, efforts to restore liberal trade centered on interna- trade policy actions of large and powerful nations. tional conferences but did not lead to renewed trade GATT Articles I (most favored nation) and III (national treaties with binding commitments to openness based treatment) promise that all WTO members will be given on MFN. Then when recession hit in the late 1920s, the same conditions of access to a particular country's governments responded with beggar-my-neighbor pro- market as the most favored member, and all foreign sup- tectionist trade policies that drove the world economy pliers will be treated the same as domestic suppliers. into depression. The volume of world trade shrank by These fairness rules are fundamental to instilling confi- one-quarter between 1929 and 1932, and its value fell dence in the world trading system. In particular, they by 40 percent. The first attempts to reverse that protec- lower the risks that are associated with a nation's pro- tion were discriminatory, as with the Ottawa Confer- ducers and consumers becoming more dependent on ence of 1932, which led to preferential tariffs on trade foreigners-risks that otherwise could be used by a among members of the British Commonwealth. country as an excuse for not fully opening its borders. Out of the interwar experience came the conviction Second, large economies have the potential of ex- that liberal world trade required a set of rules and bind- ploiting their monopoly power by taxing their trade, but ing commitments based on nondiscriminatory princi- we know from trade theory that the world as a whole is ples. While there was not enough agreement to create an made worse off by such trade taxes. Thus while each international trade organization, at least a General Agree- large economy might be tempted to impose trade taxes, GLOBALIZATION, THE WTO, AND DEVELOPMENT STRATEGIES OF POORER COUNTRIES 13 the effect of many of them doing so simultaneously may most all of them are found wanting (Corden 1997). The well be to leave most if not all of them worse off, not to most compelling explanation relates to a country's po- mention the welfare reductions for smaller countries. litical economy. It has to do with the national income Here then is the value of agreeing not to raise trade bar- redistribution feature of trade policies: the gains are con- riers and instead to "bind" them in a tariff schedule at centrated in the hands of a few who are prepared to sup- specified ceiling levels. This rule is embodied in GATT port politicians who favor protection, while the losses Article II whereby WTO members are expected to limit are sufficiently small per consumer and export firm and trade only with tariffs and are obligated to continue to are distributed sufficiently widely as to make it not provide market access never less favorable than that worthwhile for those losers to get together to provide a agreed to in their tariff schedules. Again, the greater cer- counterlobby, particularly given their greater free-rider tainty that this tariff-binding rule brings to the interna- problem in acting collectively (Hillman 1989; Gross- tional trading system adds to the preparedness of coun- man and Helpman 1994; Anderson 1995). Thus the ob- tries to become more interdependent and of business served pattern of protection in a country at a point in people to invest more. time may well be an equilibrium outcome in a national The third and perhaps most important contribution political market for policy intervention. of multilateral rules disciplining trade policy is that they That political equilibrium in two or more countries can help governments ward off domestic interest groups might be able to be altered, however, for the better seeking special favors. This comes about partly via Arti- through an exchange of product market access. If coun- cle II, which outlaws the raising of bound tariffs, as well try A allows more imports it may well harm its import- as via numerous other articles aimed at ensuring that competing producers if there are no compensation nontariff measures are not used as substitutes for tariffs. mechanisms; but if this liberalization is done in return This benefit of the system is sometimes referred to as the for country A's trading partners lowering their barriers to "Ulysses effect": It helps prevent governments from A's exports, the producers of those exports will enjoy this being tempted to "sin," in this case to favor special in- additional benefit. The latter extra benefit may be suffi- terest groups at the expense of the rest of their economy.3 ciently greater than the loss to A's import-competing pro- While no one would argue that GATT rules have ducers that Ns liberalizing politicians also become net been applied without exception, the fact that they are gainers in terms of electoral support. Likewise, politi- there ensures the worst excesses are avoided. They there- cians in the countries trading with A may well be able to fore bring greater certainty and predictability to inter- gain from this trade in market access, for equal and op- national markets, enhancing economic welfare in and posite reasons. That is, a new opportunity for trade ne- reducing political tensions between nations. More than gotiations can stimulate trade liberalization by altering that, by promoting interdependence, the GATT/WTO the incentives to lobby politicians and thereby the polit- indirectly has raised the price and hence reduced the ical equilibrium in trading nations.4 likelihood of countries going to war. Such gains from trade negotiations involving ex- But why do countries need the WTO to negotiate change of market access will tend to be greater nation- freer trade? One of the clearest lessons from trade theory ally and globally, the larger the number of countries in- is that an economy unable to influence its international volved and the broader the product and issues coverage terms of trade cannot maximize its national income and of the negotiations; hence the wisdom in negotiating economic growth without allowing free trade in all goods multilaterally with more than 100 countries over a wide and services. Consumers lose directly from the higher do- range of sectors and issues, as in the Uruguay Round, mestic prices of importables, while exporters lose indi- despite the process being cumbersome. Now that there rectly because import barriers cause the nation's currency is so much more product coverage under the WTO than to appreciate (there is less demand for foreign currency under the GATT, and the number and extent of partic- from importers) and raise the price of labor and other ipation by member countries keeps growing, the scope mobile resources. More open economies also grow faster. for exchange of market access has increased dramatically. Why, then, do countries restrict their trade, and why do That is especially true for exchanges between more- and they need to get together to agree to liberalize those pro- less-developed economies now that agriculture, textiles, tectionist trade regimes multilaterally, when it is in their and clothing are back in the GATT mainstream and ser- national economic interests to do so unilaterally? vices and trade-related intellectual property have been Numerous reasons have been suggested as to why a added, making a wider range of intersector tradeoffs country imposes trade barriers in the first place, but al- possible. I4 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION This is not to deny the many challenges confronting mit to a permanently open international trade and pay- the global trading system. Digesting and implementing ments regime and to provide secure property rights (in- the agreements that came out of the Uruguay Round ne- tellectual as well as physical). The stability of the com- gotiations is but one of them. Launching a new round mitment to openness is much more crucial now than of multilateral trade negotiations next year is another, even just 15 years ago because otherwise capital inflows especially given the lack of agreement on the extent to and investments will be only short-term in nature and which new issues should be added to the WTO's agenda. will be susceptible to withdrawal should confidence Another obvious and urgent challenge is the Asian finan- waver. It is for this reason, and because of the compre- cial crisis. This crisis is of interest internationally not just hensiveness of the Uruguay Round agreements, that lib- because of humanitarian concern for those directly hurt- eral trade policy commitments under the WTO are so ing in East Asia. What it has demonstrated is how important. They are valued by would-be investors be- volatile short-term private capital can be in an integrated cause WTO commitments involve (a) legal bindings and global capital market where herd mentality and panic (b) most-favored-nation treatment by trading partners. can take hold and spread quickly thanks to the digital The legal bindings mean a WTO member cannot return revolution in communications. The consequent risk of to a more protectionist regime by raising tariffs above contagion to other regions is thus now greater than in the bound rates listed in the member's schedules of com- even the recent past. Such international spillovers mag- mitments, nor does that member risk facing higher than nify the adverse consequences of the crisis. Among other MFN bound tariffs in exporting to its trading partners things, that multiplies the probability that affected if they are WTO members. countries in Asia and elsewhere will slow or reverse their The security of a stable trading environment instills market-opening policy stance even though the opposite a confidence in investors that is noticeably less in coun- is what is required to resume or sustain growth. The pos- tries that are not WTO members. For such countries a sibility of policy reversals underscores the importance of key ingredient in achieving good economic governance getting another comprehensive WTO round under way is to seek speedy accession to the WTO. Already there in the year 2000. are 134 countries that have chosen to join the new or- Implications for Development Strategies of ganization. The WTO is thus approaching the status of Poorer Countries a truly global trade organization except for underrepre- sentation by two groups: the former centrally planned Greater openness of and interdependence between na- economies (CPEs) seeking to transform from plan to tional economies provides wonderful opportunities for market orientations, and some of the smallest and poor- poorer economies, but it is not without its challenges. est economies. Globalization is raising the rewards to economies choos- Most of the CPEs not already members are seeking ing good economic governance but is also raising the cost WTO accession, the most notable being China (whose to economies with poor economic governance. Just as finan- accession would allow Taiwan to join) and Russia. Their cial capital can now flow into a well-managed economy accession negotiations have proceeded slowly in part be- more easily and quickly than ever before, so it can cause members want more access to those countries' mar- equally quickly be withdrawn if confidence in that econ- kets than their governments have been willing to give. omy's governance is shaken-as the East Asian crisis has This is especially so with respect to bound tariffs, sign- demonstrated all too clearly during 1997-98. In this sec- ing the WTO's plurilateral government procurement tion just two aspects of good economic governance in agreement, and assurances over intellectual property the wake of globalization are discussed: commitment to rights. Additional problems include their lack of policy a liberal international trade and payments regime, and transparency and their high degree of state trading, not growth-enhancing domestic policies that are not biased to mention the need to overcome political opposition by sectors. Together these will enable producers to take (for human rights reasons) in the U.S. Congress and else- maximum advantage of new and prospective export op- where to their joining. In China's case the "concessions" portunities following Uruguay Round implementation. available to developing country WTO members are also being sought. The United States and others are very re- Commitment to a Liberal Trade and Payments Regime luctant to allow China those "concessions," however, be- The previous discussion suggests the first priority for a cause that could effectively make meaningless the nego- poor country seeking to achieve sustainable economic tiated access to Chinese markets (Anderson 1997). development in the 21st century is to practice good eco- The other group feeling marginalized is the world's nomic governance generally, and in particular to com- least developed countries (LDCs), particularly those GLOBALIZATION, THE WTO, AND DEVELOPMENT STRATEGIES OF POORER COUNTRIES 15 that are not yet WTO members. For them the cost of been made in the Uruguay Round to lower agricultural the accession process, and subsequently of maintaining and textile import barriers, only modest reductions re- a mission in Geneva that is large enough to cover the ex- sulted by the turn of the century. Pooling of negotiating panding number of items of key concern to them, is efforts to open those markets more is one strategy worth prohibitive without some financial and technical aid. A pursuing, as the Cairns Group of agricultural exporting program of multilateral assistance does exist and was ex- countries demonstrated during the Uruguay Round. panded following a high-level meeting between LDCs, On the third question, the answer is that while there the WTO, and five other international agencies in late are ample opportunities for less-advanced economies to October 1997. Many bilateral assistance programs also become engaged in WTO activities, such as chairing exist (OECD 1998). But with so many new countries committees, they are taken up infrequently. Michalo- seeking membership and so many more issues to deal poulos (1998a) suggests that is because poor and espe- with following the Uruguay Round, the budgets for cially small countries have few, if any, delegates in those programs may need to expand further, especially Geneva, and those that are there are inadequately ser- if a new round is launched soon. viced by their national capitals and so are always over- In addition to the 30-plus countries currently in the stretched. The pooling of efforts by members forming queue for membership, perhaps another 20 will apply a group has been one way of coping, as the Association soon. Hence within a decade, the WTO will have much of Southeast Asian Nations (ASEAN) member coun- the same membership number and composition as the tries do, for example. Perhaps further aid funding is United Nations. The WTO's predecessor, the GATT, warranted for the smallest and least developed coun- began as a club of industrial countries, but by the start tries in order to raise their quality and quantity of of the Uruguay Round those countries' share had fallen representation. to one-quarter and could be as small as one-sixth dur- As to the final question concerning the pace of acces- ing the next round. sion of new members, the answer is unclear. Certainly To understand how well the WTO club is managing an average time of six years to accede to the WTO its own globalization, consider the following four ques- sounds long, and certainly politics may have con- tions: To what extent are less-advanced economies (1) tributed to delays, as with China.5 But much of the opening up to trade, (2) able to get their exports into delay appears to be on the part of the acceding country. markets of more-advanced economies, (3) engaged in Sometimes this is because of a lack of internal political WTO activities such as improving the rules, and (4) able support to push ahead with reform commitments. More to accede expeditiously? often it is because of insufficient bureaucratic horse- On the first question regarding openness, the answer power to get on top of the issues and to move the nec- is that many developing and transition economies are essary papers forward faster (Michalopoulos 1998b). opening up substantially. During the past decade or so This situation is unfortunate because the reforms re- an ever-larger number of developing countries-includ- quired to join the WTO will become even greater dur- ing those in Africa-have embraced trade liberalization. ing the next round of WTO multilateral trade negotia- Some of those reform programs have been adopted with tions. One partial solution is to raise the quantity and reluctance as conditions for receiving IMF or World quality of trade policy staffing in national capitals, and Bank loans, while others have been unconditional uni- in particular to boost training. Further education is lateral decisions. Until they are bound under the WTO, needed not only about the WTO institution but also in though, there is a risk of backsliding in the future. Fur- analytical capability and skills in advocating the virtues thermore, tariffs need to be bound at levels close to ap- of liberal markets. The domestic political commitment plied rates in order to be taken seriously, unlike during to do that may not be in place, however, and the ques- the Uruguay Round when many developing countries tion again arises as to whether more development assis- just committed to ceiling bindings at several times the tance funds need to be directed to that cause. level of applied rates. On the second question of market access for devel- Growth-Enhancing Domestic Policies That Are Not oping country products, the answer is that not enough Biased by Sectors has been done. The two sectors of most interest to less- The extent to which liberalizing one's own trade and advanced economies are agriculture and textiles/cloth- payments regime and securing greater market access op- ing, and protection levels in more-advanced economies portunities for one's exports boosts a developing coun- for those items are as much as ten times the average for try's economic growth depends importantly also on do- other merchandise. Even though commitments have mestic policy environment. Sound, predictable, stable i6 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION macroeconomic policies and taxation policies that are better-educated workers, rural areas would be more attrac- not biased by sectors are essential. Uruguay Round re- tive to investors in low-skill intensive manufacturing and re- forms abroad will make agricultural and textile/clothing lated service activities. Hence we would see clothing facto- exports more profitable for developing countries. Trade ries and the like becoming more common in rural areas. liberalization at home will tend to reinforce that, be- That would boost off-farm earnings of farm households, al- cause many developing and transition economies have lowing a more efficient and fuller use of the rural work force, traditionally protected heavy manufacturing industry at particularly in nonpeak seasons. In addition to its contribu- the expense of light manufacturing and primary produc- tion to output, this rural development strategy would slow tion-a pro-urban industrial bias. the growth of urban pollution and congestion, and reduce Those past trade and other pro-urban policy biases the incidence of poverty and hunger (since most of the poor discouraged investment in infrastructure and human are in or have recently migrated from rural areas).6 capital in rural areas. Unleashing the productive poten- tial of the rural sector requires major upgrading of es- Conclusion sential rural infrastructures, thereby lowering transac- In conclusion, it is worth reiterating that market re- tions costs of doing business there, and it requires sponses to trade liberalization at home and abroad may investment in the people involved. The crucial infra- not be sufficient to fully reap the rewards from those re- structures include rural roads, electricity, telecommuni- forms. As explained by Krugman (1998), the reason is cations, and radio transmission so that costs of trans- that history matters: Earlier protection toward heavy in- port, communications, and information (about market dustry in developing countries boosted their urban cen- conditions, new technologies, and the like) become ters, as did the restrictions on those countries' access to more affordable. Investments in these items probably rich-country markets for farm products. With the grow- will be more expensive per capita than in urban areas, ing importance of scale economies in many nonprimary but that needs to be weighed against the net long-term sectors, and inadequate environmental taxes on nega- benefits from expanding output faster from rural areas. tive externalities such as urban pollution and conges- In numerous countries irrigation investments also need tion, resources will continue to want to locate in large to be facilitated, even if they are mostly funded at the urban centers. In societies where lobbying politicians local level with the help of loans (which require secure has a high payoff, that provides a further reason for cap- property rights over land use so land can be used as col- italists to retain a presence in the capital city. First-best lateral with lenders). solutions could include imposing environmental taxes The crucial investments in people include basic and outlawing rent-seeking. But if that is too difficult, schooling (for girls as much as for boys), basic health second-best measures may well be justified. They could services, and agricultural research and extension. All the include temporarily subsidizing investments in in- empirical evidence points to the social rates of return frastructure and human capital in rural areas until the from such public investments being very high in devel- centrifugal market forces are sufficient to offset the cen- oping countries even when price and trade policies dis- tripetal forces of geographic concentration, to use Krug- criminate heavily against the rural sector; hence they are man's terminology. likely to be even higher as and when those policies are Bibliography reformed. If those social rates of return are significantly Anderson, K. 1995. "Lobbying Incentives and the Pattern of above private rates, a case might be made for govern- Protection in Rich and Poor Countries." Economic Develop- ment subsidies (Schultz 1975). Drawing on a great deal ment and Cultural Change 43(2): 401-23, January. of empirical research, Schultz argues that the case is es- Anderson, K., ed. 1996. Strengthening the Global Trading Sys- pecially strong with human capital investments in rural tem: From GA TTto WTO. Adelaide: Centre for International areas. Such investments not only raise farm incomes and Economic Studies. so reduce the need for farmers to leave agriculture, but Anderson, K. 1997. "On the Complexities of China's WTO Ac- 50 ~~~~~~~~~~~~~~~~cession." The World Economy 20(6): 749-72, September. they also increase the prospects for those who migrate Anderson, K. 1998a. Vietnam's Transition Economy and WTOAc- to nonfarm jobs. In both respects, the social tensions cession. Adelaide: Centre for International Economic Studies. that are inevitably associated with rapid economic Anderson, K. 1 998b. Lao Economic Reform and WTO Accession. growth and structural change are lessened. Adelaide: Centre for International Economic Studies. Efforts to reduce underinvestment in rural infrastructure Baldwin, R. E., and P. Martin. 1999. "Two Waves of Globaliza- tion: Superficial Similarities and Fundamental Differences." also would benefit not only agricultural production. With NBER Working Paper 6904, Cambridge MA, January better transport and communications infrastructure and (http://www.papers.nber.org/papers/w6904). GLOBALIZATION, THE WTO, AND DEVELOPMENT STRATEGIES OF POORER COUNTRIES 17 Blainey, G. 1966. The Tyranny of Distance. Melbourne: Sun OECD (Organisation for Economic Co-operation and Devel- Books. opment). 1998. Survey of DAC Members' Go-operation for Bordo, M. D., B. Eichengreen, and D. A. Irwin. 1999. "Is Glob- Capacity Development in Trade. DCD/DAC(97)24/REV1, alization Today Really Different Than Globalization a Hun- Paris: OECD, March. dred Years Ago?" in S. Collins and R. Lawrence eds. Brook- O'Rourke, K. H., and J. G. Williamson. 1999. Globalization ings Trade Policy Forum 1999, Washington, D.C.: Brookings and History: The Evolution of a Nineteenth Century Atlantic Institution. Economy. Cambridge, MA: MIT Press. Cairncross, F. 1997. The Death ofDistance: How the Communi- Petersmann, E-U. 1991. Constitutional Functions and Constitu- cations Revolution Will Change Our Lives. London: Orion tional Problems of International Economic Law. Fribourg: Fri- Business Books. bourg University Press. Corden, W. M. 1997. Trade Policy and Economic Welfare, 2d ed. Roessler, F. 1985. "The Scope, Limits and Function of the Oxford: Clarendon Press. GATT Legal System." The World Economy 8(3): 287-98. Dollar, D. 1992. "Outward-Oriented Developing Economies Sachs, J. D., and A. Warner. 1995. "Economic Reform and the Really Do Grow More Rapidly: Evidence From 95 LDCs, Process of Global Integration." Brookings Papers on Economic 1976-85." Economic Development and Cultural Change 40: Activity 1: 1-118. 523-44, April. Schultz, T. W. 1975. "On the Ability to Deal With Disequilib- Edwards, S. 1993. "Openness, Trade Liberalization and Growth ia. "JournalofEconomic Literature 13(3): 827-46, September. in Developing Countries." Journal of Economic Literature UNCTAD (United Nations Conference on Trade and Develop- 31(3): 1358-93, September. ment). 1998. World Investment Report 1998: Trends and De- Goldstein, J. 1998. "International Institutions and Domestic terminants. New York and Geneva: United Nations. Politics: GATT, WTO, and the Liberalization of Interna- USITC (United States International Trade Commission). 1997. tional Trade." In A. O. Krueger, ed., The WTO as an Inter- The Dynamic Effects of Trade Liberalization: An Empirical national Trade." tin. 0. Kruege ed. The Won as aier- Analysis. Publication 3069, U.S. International Trade Com- national Organization. Chicago and London: University of misin Wahntn D.. Ocobr Chcg Prss mission, Washington, D.C., October. Cimcago ress. Vernon, R. 1966. "International Investment and International Grosma, G. M. an E. Hepn 1991. Innvto an Trade In the Product Cycle." Quarterly Journal of Economics Growth in the Global Economy. Cambridge, MA: MIT Press. the Product Cy Grossman, G., and E. Helpman. 1994. "Protection for Sale." 80: 190-207, May. World Bank. 1997. Rural Development: From Vision to Action. American Economic Review 84(4): 833-50, September. ESSD Studies and Monograph Series 12, Washington, D.C.: Grossman, G., and E. Helpman. 1995. "Trade Wars and Trade The World Bank. Talks." Journal of Political Economy 103(4): 675-708, World Bank. 1998. World Development Report 1998/99: Knowl- August. edge for Development. New York: Oxford University Press. Hillman, A. L. 1989. The Political Economy of Protection. New WTO (World Trade Organization). 1996. Annual Report 1996. York: Harwood Academic. Geneva: World Trade Organization. Hoekman, B. M., and M. Kostecki. 1995. The Political Economy WTO (World Trade Organization). 1998a. Annual Report 1998. of the World Trading System: From GA TT to WTO. London Geneva: World Trade Organization. and New York: Oxford University Press. WTO (World Trade Organization). 1998b. Electronic Commerce Hufbauer, G. C. 1991. "World Economic Integration: The Long and the Role of the WTO. Geneva: World Trade Organization. View." International Economic Insights, May/June. Kindleberger, C. P. 1975. "The Rise of Free Trade in Western Europe, 1820-1875." Journal of Economic History 35(1): Niotes 20-55, March. 1. Between 1920 and 1980, the real charge per tonne for Krugman, P. 1998. "The Role of Geography in Development." ocean freight fell by almost three-quarters and between 1960 and Paper presented to the Annual World Bank Conference on 1980 the real cost of a telephone call from New York to London Development Economics. Washington, D.C., April 20-21. fell by 90 percent. Meanwhile, between 1930 and 1980 the real Landes, D. 1998. The Wealth and Poverty of Nations. London: cost of air travel fell 85 percent (Hufbauer 1991). Little Brown and Company. 2. Two book tides summarize this 150-year history: Blainey's Maddison, A. 1982. Phases of Capitalist Development. London: Tyranny of Distance which refers to Australia's early isolation Oxford University Press. from the Old World prior to steamships, and Cairncross' Death Markusen, J. R., A. J. Venables, D. B. Konan, and K. Zhang. of Distance which refers to the latest communications revolution 1996. "A Unified Treatment of Horizontal Direct Invest- (Blainey 1966; Cairncross 1997). For comparisons of the nine- ment, Vertical Direct Investment, and the Pattern of Trade teenth and late twentieth century episodes of globalization, see in Goods and Services." NBER Working Paper 5696, Cam- Baldwin and Martin (1999) and Bordo, Eichengreen and Irwin bridge, MA. (1999). Michalopoulos, C. 1998a. "The Participation of Developing 3. Petersmann (1991, p. 83) goes so far as to say that "the pri- Countries in the WTO." Policy Research Working Paper mary regulatory function of the GATT.... [is] the welfare- 1906, World Bank, Washington, D.C., March. increasing resolution of domestic conflicts of interest within Michalopoulos, C. 1998b. "WTO Accession for Countries in GATT member countries among individual producers, im- Transition." Policy Research Working Paper 1932, World porters, exporters and consumers." Similarly, Roessler (1985, Bank, Washington, D.C. p. 298) claims that "the principal function of the GATT as a sys- i8 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION tem of rules is to resolve conflicts of interest within, not among, 5. In particular, the concern in the United States about human countries. The function of the GATT as a negotiating forum is rights in China has often meant the US Congress is unsympathetic to enable countries to defend the national interest not against the to China. One manifestation of that is the high demands on China national interests of other countries but against sectional inter- in terms of policy reform and market opening before the US will ests within their own and other countries." conclude its bilateral negotiations on its WTO accession. Vietnam 4. Elaborations of this economists' perspective can be found and Laos have similar hurdles to jump in their bids to join WTO, in Grossman and Helpman (1995), Hillman and Moser (1995), in their cases because of the sensitive issue of US soldiers missing Hoekman and Kostecki (1995), and Anderson (1996, Ch. 1). in action during the Vietnam war (Anderson 1998a,b). Political scientists are beginning to take a similar view. See, for 6. For more on the benefits of removing the anti-rural bias in example, Goldstein (1998). many countries' development strategies, see World Bank (1997). chapter 3 Toward a Global Financial Architecture for the 21st Century Robert E. Litan Vice President and Director Economics Studies Program The Brookings Institution ~ oing into the 21st century, the chal- evaluating information about investment projects and lenge for financial policymakers their sponsors, commercial banks and other financial in- is to learn from the past while con- termediaries can help efficiently allocate capital to its structing institutions that better en- most productive uses. able economies to reap the benefits of These are not merely theoretical propositions. Empir- modern finance without the pitfalls ical evidence confirms that the level of financial devel- that have plagued so many countries in this century, opment-measured by, among other things, the size of vividly illustrated by the Asian financial crisis of the past the banking sector relative to overall economic activity- two years. The task is complicated by the need to antici- is highly correlated with economic growth among devel- pate and take account of trends and developments that oping countries, controlling for other factors (Levine will continue to shape the financial landscape throughout 1996). At the same time, if incentives for the efficient the world. use of capital are lacking, then even sophisticated finan- This chapter seeks to address this challenge. It begins cial systems cannot insulate an economy against adver- by surveying the role of finance in both developed and sity, as the Asian crisis demonstrated. developing economies and its increasing cross-border or The focus on banking in developing countries is not global character. It then describes the origins and key an accident. Developing economies rely heavily on features of the recent financial crises, as a predicate for commercial banks to perform the critical function of a description of some of the key issues that financial intermediating funds between savers and investors. In policymakers around the world must now confront. contrast, nonbanking institutions active in the capital Along the way, it addresses the role for financial super- markets play a more important intermediating role in vision, the relative costs and benefits of various ex- more developed economies. The relative importance of change rate regimes, and the ways in which interna- banks and markets as instruments of intermediation is tional financial institutions may assist in achieving more a function, in large part, of the state of economic de- orderly growth of both the real and financial sectors in velopment. In developing economies, savers tend to the years ahead. lack sophistication in financial matters and are mainly interested in investing their funds in safe vehicles, such The Economics and Globalization of Finance as bank deposits. Meanwhile, few borrowers in devel- Finance is important not just for its own sake but be- oping economies are of sufficient size and credit- cause it is central to the process of economic growth. Fi- worthiness to justify the expense of floating securities, nancial intermediaries and markets enhance both saving which in turn reduces the demand for securities mar- and investment, enlarging the capital stock and thus in- kets. However, as economies mature and their firms creasing the productivity of workers. In addition, by grow in size and sophistication, so do their capital mar- taking advantage of economies of scale in collecting and kets, provided both the private and public sectors erect '9 20 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION a suitable technical and legal infrastructure to support Figure 3.1 them. Net Long-Term Resource Flows to Developing Financial intermediation is not confined to national Countries, 1990-97 borders. Like water that will flow downhill if given the chance, financial capital will-if allowed-move to proj- 300billions of dollars ects offering the highest risk-adjusted returns wherever * official development flows a debt flows they may be located. In addition, modern portfolio the- 250 - 0 foreign direct investment ory demonstrates that there are advantages to investing 0 portfolio equity flows in a wide variety of projects, even in different countries, 200 - provided those endeavors' returns are not perfectly cor- related with one another. 150 - These basic theoretical propositions have been put to work in recent decades, as finance has become truly 100 - "globalized," through two types of cross-border financial flows. So-called portfolio capital consists of investments 50- * that are potentially short-term in nature, including loans between banks and investments in traded bonds and 0 equities. Cross-border interbank lending now exceeds 1990 1991 1992 1993 1994 1995 1996 1997 US$6 trillion, up from US$1.2 trillion as recently as Source. World Bank Debtor Reporting System. 1983 (Bank of International Settlements 1992). The globalization of stock markets also has facilitated the movement of portfolio capital between nations. The shares of portfolios invested in foreign stocks and bonds was up by more than 125 percent from the beginning of have been rising steadily in the United States, Western the decade. Indeed, during the 1989-96 period, the Europe, and even Japan (Tesar and Werner 1998). growth of incoming FDI into developing countries out- In contrast, foreign direct investment (FDI) consists paced the growth of trade and gross domestic product of long-term holdings of equity, where the aim is to con- (GDP) by a large margin, as well as the growth of in- trol or exercise significant influence over local firms. FDI coming FDI into developed countries (Figure 3.2). has become increasingly linked to trade as it no longer An important aspect of the globalization of finance is suffices simply to ship many products across borders to the increased presence of banks in multiple countries. customers waiting in foreign lands. Consumer and busi- Table 3.1 highlights the extent to which foreign banks ness products often require local servicing operations, as have made inroads into the banking systems of certain well as research and development and marketing enter- developing countries. Notably, none of the countries prises, if they are to be sold effectively in other countries. with a foreign bank share in excess of 20 percent experi- Moreover, financial and professional services, among enced banking and financial problems either directly or others, typically can only be supplied by a presence on indirectly during the Asian financial crisis. Two reasons the ground in other countries. It is not surprising, there- help explain this fact. Foreign institutions bring cutting fore, that since the dominant share of trade of developed edge financial technologies and practices that not only economies is with other developed economies, a similar improve the performance and health of financial institu- pattern has developed with respect to FDI. tions in the host countries but impart up-to-date knowl- The globalization of finance has not been confined to edge of banking practices to their regulators. More- developed economies. One of the more important trends over, because they are often active in numerous parts of of the 1990s is that investors from developed economies the world, foreign financial institutions are typically far have poured steadily larger sums into developing coun- more diversified in their lending than local institutions, tries. This process has been facilitated by dramatically and thus less likely to fail when economic conditions in lower costs of collecting, analyzing, processing, and com- local markets turn down (for evidence, see Claessens and municating information. Figure 3.1 illustrates that as late Glaessner 1998). as 1990, official funds accounted for the majority of the Financial integration has not been universal, how- roughly US$ 100 billion flowing to developing countries. ever. The vast majority of developing countries still re- By 1997, about US$300 billion went into the develop- ceive relatively little private capital (World Bank 1997). ing world, of which only about 15 percent was supplied Much of the flows have been to Asia, where net foreign officially. FDI accounted for the largest share (40 per- portfolio investments rose steadily from US$20 billion cent), followed closely by borrowing (34 percent), which in 1990 to nearly US$120 billion in 1993. Since 1993 TOWARD A GLOBAL FINANCIAL ARCHITECTURE FOR THE 21ST CENTURY 21 Figure 3.2 Table 3.1 Trends in Global Finance: Growth of GDP, Trade, Foreign-Owned Banks and FDI, 1989-96 Government Percentage share of total assets Advanced Countries Hong Kong, China 78.Oa annual percentage change India 7.3 8- Indonesia 3.7 Rep. of Korea 5.1 Malaysia 15.9 6- Singapore 80.0 Taiwan, China 4.7 Thailand 7.1 Argentina 21.7 Brazil 9.4 Chile 21.4 2- Colombia 3.6 Mexico 1.2 _E _-_Venezuela 1.2 GDP trade FDI Russian Federation 2.2 Israel 0.0 South Africa 3.3 Developing Countries Germany 3.9 annual percentage change Japan 1.8 24 - United States 22.0 Note: Figures refer to latest available year. 20 - Refers to all overseas-incorporated authorized institutions. Sources: OECD, central banks, ministries of finance, and Goldstein and 16 - Turner (1996). Adapted from Goldstein 1997. 12 - 8 - cially in an increasingly global environment, financial 4 _ crises can lead to significant disruption of real econo- _0_E l mies. Large-scale banking problems, for example, can o - ~ ~ ~ contribute to "credit crunches," or severe contractions in GDP trade FDI the availability of credit at the worst possible time, dur- Source: World Bank Investment Report, 1997, and World Economic ing a recession. This problem can be especially serious in Outlook, May 1998. developing economies, which as already noted, rely much more heavily on banks to perform intermediation than capital markets (Lindgren et al. 1996). Indeed, fi- portfolio inflows have fluctuated considerably, collaps- nancial crises have occurred so frequently in some many ing to less than US$40 billion in 1998 (International parts of the world that some economists believe that fi- Monetary Fund 1998). Such fluctuations imply that fi- nancial panics (sudden "runs" by investors in capital nancial integration can be a two-edged sword: If foreign markets or depositors from their banks) or "near crises" capital comes in too quickly, it can overwhelm the ca- (financial difficulties short of a panic but which never- pacity of local financial institutions to put it to produc- theless depress economic activity) are an inevitable fea- tive use, especially where government guarantees distort ture of market economies (Kindleberger 1978; Minsky investment choices. This can sow the seeds of a later fi- 1982). nancial crisis, especially if local bank failures or other In the 20th century, the deepest financial crisis oc- shocks to the economy trigger a loss of confidence in the curred during the Depression when thousands of banks currency. In that event, portfolio capital tends to run failed in the United States and other countries, and out much more quickly than it came in. stock prices plunged to fractions of their prior levels. For several decades after World War II, the financial arena Financial Crises was relatively free from major difficulty. But since 1980 When it functions well, finance is critical to facilitating the industrialized and developing world alike have ex- economic growth. But when it is managed poorly, espe- perienced a rash of banking crises, with notable exam- 22 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Table 3.2 noted, the 1987 stock market crash reverberated around Severe Banking Crises, 1980-96 the world. The Mexican and Asian financial crises of the 1990s had international effects because both episodes Estimate of total lossestcosts were characterized by sharp plunges in exchange rates, Country (time of crisis) (percentage of GDP) which inevitably affected trade flows, and because the Latin America financial difficulties of these countries stemmed in large Argentina (1980-82) 55 * f Chile (1981-83) 41 part from an inablity to service debt linked to or de- Venezuela (1994-95) 18 nominated in foreign currencies (sovereign debt in the Mexico (1995) 12-15 Mexican case and private borrowings in the Asian case). Africa Similarly, the United States and to a lesser extent other Benin (1988-90) 17 industrialized countries suffered near crises in the 1980s Cote d'lvoire (1988-91) 25 when their large banks suffered losses on sovereign lend- Mauritania (1988-93) 15 ing to developing (especially Latin American) countries. Tanzania (1987-95) 10 In most other cases, however, banking crises have had entirely or largely domestic origins and impacts. Middle East Israel (1 977-83) 30 Challenges Ahead Transition Countries Developing sound financial policy is much like manag- Bulgaria (1990s) 14 ing auto traffic on highways: While there always will be Hungary (1995) 10 accidents, traffic engineers try to avoid especially costly Industrial Countries ones-crashes at key intersections that cause gridlock United States (1980s) 3-4 throughout an entire region. The number and severity Spain (1977-85) 17 of auto crashes can be reduced by educating drivers, de- Japan (1990s) 10 signing safer cars, monitoring speed levels, and building safer highways. The market alone cannot carry out all these functions. Countries have found that a helpful ples listed in Table 3.2. The crises in the developed hand from government is necessary. world include the well-known savings and loan disaster The objectives and tools for controlling financial traf- in the United States, the ongoing difficulties in Japan, fic are similar. The objective of financial policy can be and the banking problems that plagued the Scandina- viewed as facilitating the orderly flow of money between vian countries in the late 1980s and early 1990s. The actors, wherever they may be located, without experi- crises in the developing world have been both more nu- encing "crashes" resulting in financial gridlock. This, merous and more severe, as measured by the cost of their too, requires a combination of market discipline and resolution as a share of local GDP. In all, roughly three- government oversight. Market participants must have quarters of the IMF's member countries have experi- timely access to accurate information; those running fi- enced significant banking problems in the past two nancial institutions must be educated in the tools of risk decades (Lindgren et al. 1996). management; mechanisms must be in place to ensure The 1980s and 1990s have been marked not only the safety and soundness of individual financial institu- by banking crises, but by panics in capital markets as tions; and the "highways" of finance-the nuts and well. The October 1987 stock market crash afflicted bolts of the back offices of financial institutions and markets throughout the world, although prices recov- clearing and settlement systems-must be built to with- ered in many markets a year later. That was not true in stand both heavy volumes of traffic and occasional traf- the case of Japan, whose stock prices collapsed in 1989 fic accidents. and at present stand 40 percent below those prior to the While the highway analogy is highly useful, it actu- crash. Similarly, stock prices plunged in Mexico during ally understates the difficulties in designing financial sys- its peso crisis of 1994-95 and throughout Southeast tems, especially in an increasingly global financial mar- Asian markets during the recent financial crisis there. ketplace. Although automobile accidents may cause In all of these cases, the collapse in stock prices was di- highway traffic to back up many miles from the scene, rectly or indirectly linked to weaknesses in domestic accidents in one part of a county, state, or country do banking systems. not tie up traffic in other geographic locations. In other Some, but certainly not the majority of, financial words, there is no highway analogy to a "contagious crises have had international consequences. As just run," which can occur in financial systems when prob- TOWARD A GLOBAL FINANCIAL ARCHITECTURE FOR THE 21ST CENTURY 23 lems in one or more institutions or in the currencies of widely. In contrast, when lending is concentrated in one or more countries trigger the flight of depositors or banks, especially when the banking industry itself is investors from other institutions or currencies. highly concentrated as it is many developing coun- Nonetheless, it is useful to keep the highway analogy tries, then an economic downturn that renders many in mind in considering the following four trends that borrowers unable to service their debts can signifi- are likely to pose challenges to financial policymakers in cantly weaken the banking system. This in turn can the years ahead: aggravate the recession to the extent that banks are then unable to extend credit. * Continued improvements in communications and X Just as commerce continues to "globalize," so will fi- computer technology will lead to even more rapid nancial institutions. Indeed, many large banks head- transmission of information while increasing both the quartered in developed countries already have fol- volume and speed of financial transactions. Here lowed their corporate customers into the countries technology may be both a boon and a bane. The grad- in which they have made direct investments. This ual implementation of technology-based "real-time process will only gather force in the next century as gross settlement" systems for clearing and settling trade expands and firms increasingly cross borders to funds transfers (including currency trades) between service existing customers and to develop new mar- banks will eliminate clearing risk, helping insulate kets. The Asian crisis has given a strong impetus to banking and securities markets against chain reactions countries that previously had inhibited FDI to relax if one or more large financial institutions fail. The po- those restrictions. tential downside of technological advances, however, is that precisely because they make it easier to move Cross-border investment by financial institutions, in money from one place and one instrument to the developing countries in particular, increases competition other, they also facilitate more rapid exits from mar- and innovation in these markets while helping to stabi- kets when investor moods suddenly shift. lize their financial systems. At the same time, however, * Financial technologies also will continue to evolve, the globalization of finance throughout both the devel- producing new and even more complex financial in- oped and developing worlds will pose continuing and struments and trading strategies. To the extent that new challenges to national regulators. As a Group of the new technologies facilitate risk management by fi- Thirty report points out, "there is no international frame- nancial institutions and investors, they will enhance work for dealing with the supervisory, legal and financial the safety of those institutions and of the system as a problems that would arise in a cross-border insolvency of whole. However, complex new instruments and trad- any kind, and a major cross-border insolvency in the ing strategies can detract from safety and soundness. financial sector could therefore pose a substantial risk It can take time for top management and regulators to the international financial system" (Group of Thirty to fully understand the risks of the newer instru- 1998). ments, especially how they might interact with one another. The risks to individual institutions, and more importantly, to entire financial systems may rise Industrial economies with relatively sophisticated finan- in the meantime. Moreover, while the fact that most cial systems that are already open to capital movements customized derivatives contracts are concentrated of all kinds face challenges that are both similar to and among the world's largest financial institutions gives different from those confronting developing countries. some observers comfort, it makes others uneasy, fear- The similar challenge is to maintain and effectively ing the potential chain reaction-like effects if one or enforce sound capital standards for financial institu- more large counterparties become unable for any rea- tions, in particular banks whose deposits are insured ex- son to honor their commitments. plicitly (up to some limit) or implicitly (if the bank is * Markets will continue to replace banks as vehicles of large). This challenge is more complex and difficult than intermediation, not just in the developed world where it may appear because capital standards by themselves this process has been ongoing (especially in the are merely numbers and useless as a device for disciplin- United States), but also in developing and emerging ing banks against excessive risk-taking, as well as insu- market countries where it is in its relative infancy. lating them against economic shocks of various sorts, Markets disperse lending risks throughout an invest- unless bank assets and liabilities are recorded at levels ment community, potentially worldwide in scope, so reasonably close to their market values. Full-dress mar- that even when prices fall, the impact is diffused ket value accounting is probably some way off, even in 24 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION industrial economies, because many bank loans con- countries. To be sure, there have been exceptions, such tinue to be customized and are not traded (although at as the failure of the Bankhaus Herstatt in 1974 that led some point in the next century, it is likely that some to the creation of the Basle Committee and the more re- form of market value accounting for banks will be cent collapse of the infarmous Bank of Credit and Com- adopted, either at the behest of regulators or the mar- merce International (BCCI), which had cross-border ket). Nonetheless, even under historical cost accounting ramifications around the world and ultimately led to standards, loan loss reserves-which reduce a bank's re- tightened regulation of foreign banks by the United ported capital-must reflect current conditions if bank States and other countries. With the increasing global- capital is to even come close to being an accurate mea- ization of finance, however, failures of internationally sure of the institution's net worth. This, in turn, requires active institutions-with offices and activities located effective regulatory supervision to ensure the reasonable- in numerous countries-may become more common. ness of the financial data the banks and their auditors While international failures may pose challenges to fi- report. nancial regulators throughout the world, they should be But even accurate information is not sufficient to of special interest to regulators in industrial economies ensure the effectiveness of a supervisory regime. Both where global financial institutions are likely to be most market participants and regulators must act on informa- active. tion suggesting bank weakness, and must do so promptly. The rapid changes in financial technologies described Otherwise, weak and insolvent institutions have strong earlier also pose special challenges for regulators in in- incentives to continue gambling, especially when they dustrial economies where those technologies, instru- know that because of deposit insurance schemes or their ments, and trading techniques are most intensively used. size, the government is likely to guarantee their deposi- The member countries of the Basle Committee have tors' holdings. The United States learned the lesson of been attempting to keep up with these fast-paced devel- regulatory "forbearance" the hard way during its savings opments by continuously refining the minimum capital and loan and banking problems in the 1980s, which ul- standards to take account of the newer risks these devel- timately required taxpayer funds amounting to roughly opments may pose (see Box 3.1). But the way they are 3 percent of GDP to reimburse depositors (a sum that responding highlights the growing recognition of the at the time seemed enormous, but with the passage of limits to detailed regulation in an increasingly complex time, looks moderate by international standards). Simi- and fast-moving financial environment. larly, large U.S. banks that had weakened capital posi- A key problem confronting the Committee-and in- tions as a result of uncollectible loans to less developed deed all national bank regulators-is the fact that mar- countries were not forced by regulators to recognize the ket discipline cannot function effectively in the presence full extent of their losses, and instead were allowed to of government guarantees that distort decisionmaking. take risky bets-primarily in commercial real estate- Although industrial economies have limited the reach that also turned sour and later brought a number of the of formal deposit insurance protection, the perception banks near to insolvency (Litan 1994). As a result of remains that their governments nonetheless would pro- these experiences, the United States adopted a "prompt tect all depositors, and possibly all creditors, of those corrective regime" in its banking law in 1991. A central large banks whose failure might be perceived to jeopar- challenge for the United States in the next century is not dize financial stability. Even the United States, which to weaken this regime if and when the economy turns formally prohibited government guarantees of unin- down again and political pressure for regulatory forbear- sured depositors in 1991, left open the possibility for ance returns. broader guarantees in the event all bank regulators con- If there is any doubt about the dangers of regulatory clude it is necessary to prevent systemic risk. As finan- forbearance, the Japanese reaction to its own banking cial institutions grow in size and scope, often by merger, troubles in the 1990s should dispel it. By not taking the so-called too big tofail problem is likely to become forceful, timely action at the first sign of trouble to con- more pronounced over time. To the extent very large in- strain its banks from risk-taking or to require raising ad- stitutions benefit from such implicit government guar- ditional capital, the Japanese banking authorities have antees, they have lower funding costs than their smaller allowed that country's banking problems to grow to vast competitors, which constitutes an artificial bias toward proportions. concentration within and across markets. For the most part, banking failures to date have been For those countries seeking to address this problem, domestic in nature in that the institutions involved have one possible solution is to require large banks to back a been engaged primarily, if not exclusively, in their home certain portion of their assets with subordinated debt, TOWARD A GLOBAL FINANCIAL ARCHITECTURE FOR THE 21ST CENTURY 25 Box 3.1 The Basle Capital Accord: A First Step Toward Regulatory Harmonization The Basle Committee on Banking Supervision was established in weighted total. The minimum standard was set at 8 percent. In sub- 1976 by G-10 countries and Luxembourg in the wake of the failure sequent years, the Basle Committee has supplemented the initial of the Bankhaus Herstatt, a West German bank whose unfulfilled standards with separate computations for other types of risk, such foreign currency obligations to American banks sent a shudder as interest rate risk and trading risks. through the U.S. and other financial systems. For its first decade, Most recently, in June 1999 the Basle Committee proposed a the Basle Committee focused on principles for the supervision of sweeping set of revisions to its risk-based standards, adding more internationally active banks and for cooperation among bank regu- risk classifications and tying them to the ratings of borrowers pro- lators to help govern the activities of multinational banks. By the vided by credit rating agencies. The intention behind the proposal mid-1 980s, concern grew among the members both about the in- was to incorporate more market-based information in setting the adequate levels of capital at major banks and the seeming unfair- risk weights. ness of different capital standards in different countries. Spurred However, the June 1999 proposal is open to criticism on sev- by an initial agreement on capital standards between the United eral grounds. For one thing, by retaining a complicated set of risk States and the United Kingdom in 1986, the Basle Committee even- weights, the Basle standards continue to ignore the fact that the tually adopted in 1988 the first international capital standards for best measures of bank risk center on the risks of their entire port- banks. folios rather than the sum of the risk-weighted assets. Indeed, the The initial standards were "risk-weighted" in that assets of dif- risk weights that the Committee continues to use are arbitrary in ferent categories (as well as off-balance sheet liabilities) were given nature. These flaws are not corrected by relying in private credit rat- different risk weights (government bonds in the member countries ing agencies, which often lag rather than lead to market develop- being assigned a zero risk weight, ordinary loans a 100 percent ments. Furthermore, many borrowers-especially those outside weight) and then capital was computed as a percentage of the risk- the United States-are not even rated. or uninsured instruments "subordinate" to the rights of next century. One will be to upgrade-as quickly as pos- depositors. Unlike deposits, which can be withdrawn at sible-financial laws, practices, and institutions up to any time and thus are potential sources of instability, minimum standards, such as those embodied in the holders of subordinated debt cannot regain their princi- Core Principles established by the Basle Committee. pal until the instruments mature (or only by selling The second will be to decide at what pace to link their them to other investors on the secondary market). In ad- financial systems with the industrialized world, or using dition, unlike shareholders who benefit when their the highway analogy spelled out earlier, at what pace to banks become more profitable, subordinated debt hold- build out and fully pave their financial on-ramps to the ers are entitled only to the interest on the bonds, and be- global financial turnpike. cause their instruments are uninsured, they bear the risk of losing their investments. For all these reasons, subor- Upgrading Financial Supervision dinated debt holders have powerful incentives to moni- Developing countries have several reasons for upgrad- tor the risks taken by bank managers and shareholders. ing their financial standards and practices. First, the A subordinated debt requirement would supplement Asian crisis has demonstrated that while it may seem to regulation with market-based discipline and thus imple- work for some time, the personalized, nontransparent, ment the "multiple eyes" approach to bank surveillance and heavily government-influenced credit culture preva- that one recent World Bank study has recommended for lent not only in Asia but in other parts of the world has all countries (Caprio 1998). The idea has been recom- its limits. When those making credit decisions do not mended in a report of the Treasury Department of the believe they bear any significant risk if they make a mis- United States (Litan 1997) and has been supported by take, they are likely to lend too much for the wrong pur- numerous academic scholars. It is being seriously con- poses, especially when foreign capital rushes in, provid- sidered by U.S. regulatory agencies and should be on the ing local banks with ample funds with which to make agendas of other countries as well (Board of Governors those mistakes. The result is not just that scarce capital 1999). ends up being channeled toward unproductive uses, but that in the process, asset prices are bid up artificially, Financial Policy Challenges for only to collapse when investors suddenly pull out, send- ing the economy into a recession. Financial policymakers from countries outside the in- Second, in the wake of the Asian crisis, upgrading fi- dustrialized world, but seeking to join that club as nancial standards (and enforcing them) will be necessary rapidly as possible, confront two sets of challenges in the if developing countries hope to attract direct investment 26 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION from abroad, which as already noted, holds the best the need to upgrade experience within their regulatory promise for quickly bringing cutting-edge financial ex- agencies, there is a strong case for not mixing banking pertise to both the private sector and financial supervi- and commerce in developing countries, and even for sors in these economies, as well as for weaning them off separating commercial and investment banking for official finance. Although many foreign investors jumped some transition period in order not to stifle the devel- into Asian countries in the years preceding the crisis opment of capital markets. without caring about the lack of transparency or the A number of developing countries already have made presence of Western-style bank regulation and supervi- major strides toward improving their banking systems, sion, the crisis has since acted as a wake up call: Smart notably Argentina and Brazil. Indeed, Argentina is money is not likely to return to Asia nor is it likely to ahead of the developed world in being the first country move in great quantities into other developing countries to require its banks to back a certain portion of their until there is demonstrated progress toward upgrading assets with subordinated debt, a recommendation ad- the financial infrastructure. Moreover, countries saddled vanced earlier in this chapter. with major bank insolvency problems, in Asia and East- More broadly, it will be in the interest of developing ern Europe in particular, have also discovered that they countries to adopt other reforms to strengthen their cap- can minimize the taxpayer costs of resolving them if they ital markets so that they can become less reliant on are able to attract foreign financial institutions to join banks for intermediation. Among other things, this will the bidding for the insolvent institutions and their non- require enforcement of disclosure standards for firms performing assets. whose securities are listed in exchanges, effective clear- Goldstein (1997) has laid out a comprehensive ing and settlement mechanisms so that trades can be agenda for upgrading developing countries' banking su- completed promptly without undue risk, and effective pervision systems (see Box 3.2). Significantly, Gold- bankruptcy systems that can quickly sort out claims of stein's list envisions banks in these countries not just creditors when things go wrong. In addition, it is espe- meeting the Basle capital standards, but capital require- cially important for governments to develop their bond ments 50 percent more stringent than the Basle rules, markets, issuing government debt in maturities extend- to reflect the greater risks currently associated with lend- ing beyond one year. This should not only make fund- ing in these economies. In addition, the agenda for re- ing of the government more stable but also help develop form extends beyond well-enforced capital guidelines to a longer-term bond market for private securities as well. include rules designed to avoid conflict of interest and It is much easier to spell out what may be required connected lending of the sort that has contributed to for developing and emerging market countries to do to banking problems throughout the world. Indeed, given upgrade their financial supervision and regulation, how- ever, than for these countries actually to implement the Box 3.2 necessary steps. For economies that have long operated Banking Reform Measures for Developing Countries under different rules and institutions, the transition will be long and difficult. It would be wishful thinking to * Implement and enforce capital standards for banks that are believe otherwise. Nonetheless, the experience of the above those for developed countries United States in overcoming its banking difficulties in * Publish timely, accurate, and audited information on the finan- recent years provides a beacon of hope to developing cial condition of individual banks . * Adopt internationally recognized loan classification and provi- countries. sioning practices * Enforce requirements that banks maintain internal procedures Policy Toward Capital Movements and safeguards Precisely because it will take some time for developing * Make government involvement in the banking sector more transparent (through disclosure of government costs involv- v t ing the banking system, data on nonperforming loans of state- tion and supervision, governments in these countries owned banks, and disclosing the nature and extent of govern- must also determine how open they should be in the ment instructions to banks on the allocation of credit) meantime to shorter-term portfolio capital. Clearly, one * Limit bank lending to "connected" parties * Limit deposit insurance and offset its "moral hazard" through a system of "prompt corrective action" (now embodied in developing country has decided to fix its exchange rate U.S. banking law) but has a weak financial sector, it runs great risks if it Source: Goldstein (119971. freely allows the influx of short-term funds, especially if they are borrowed and denominated in foreign currency. Moreover, notwithstanding the theoretical case for TOWARD A GLOBAL FINANCIAL ARCHITECTURE FOR THE ZIST CENTURY 27 openness to foreign capital-that it facilitates the effi- ward markets for currencies, disrupt trade. For this rea- cient allocation of capital around the world-there is son, even if a country decides not to fix its exchange little evidence that controls on portfolio capital have pe- rate, it may want to provide some buffer to its domestic nalized economic growth in countries that have used economy from volatile foreign capital movements. them. At a minimum, therefore, it is prudent for coun- In that case, allowing foreign portfolio capital to move tries with weak banking systems and methods of finan- freely in and out of equities while putting some limits on cial regulation to provide disincentives for banks and or disincentives against short-term foreign currency bor- private firms to borrow in foreign currency. rowing by private sector firms may steer a sensible mid- What about countries that have been forced off fixed dle course. Examples of the latter include reserve require- exchange rates, already have heavy foreign currency ments on short-term foreign currency borrowing that obligations, weak or essentially insolvent banking sys- have been applied in Chile and Colombia or direct lim- tems and meager foreign currency reserves, but so far its on foreign currency borrowing by banks in particular have allowed portfolio capital to move freely in and out that have been used in Malaysia (which also has used re- of the country? Should they now reinstate some form of serve requirements). controls on foreign portfolio capital? Or is the prover- It cannot be emphasized too strongly that any limits bial horse already out of the barn? on short-term capital movements should not be ex- For countries in the middle of an exchange rate cri- tended to foreign direct investment, which, for reasons al- sis, either having floated or watched the currency plum- ready discussed, is an important means by which coun- met, introducing new controls on incoming capital runs tries can increase their rate of growth. Indeed, empirical a severe risk of aggravating the crisis. After all, in the evidence suggests that FDI tends to act as a complement middle of a currency crisis, governments should want to, rather than substitute for, domestic investment. And foreign (and domestic) investors to buy or keep their not to be forgotten is the fact that the Southeast Asian currency rather than sell it. countries, where the financial crisis first appeared, con- Furthermore, not all short-term capital is alike in ef- tributed to their own predicament by restricting foreign fects. At worst, flows into and out of equity instruments direct investment, which provided artificial induce- can produce swings in the prices of domestic stocks and ments for banks and private firms to borrow short term may interrupt some public offerings. Nonetheless, for- in foreign currency instead. eign investors in the equities of domestic firms know One further caveat is in order. The case for even tar- that their investments are risky, and they bear the full geted capital controls on foreign currency borrowing consequences of any currency depreciations. Accord- must recognize that the presence of any controls invites ingly, there seems little reason for developing countries rent-seeking and corruption (Summers 1998). The ulti- to restrict portfolio investments in equities. mate aim for all countries is to have their economic and That is not the case with movements of foreign capi- financial systems mature to the point where even these tal into debt instruments. Foreign capital pouring into limited controls, which can themselves lead to distor- banks, in the form of interbank deposits, can fuel lend- tions, prove no longer necessary. Indeed, the preferable ing booms. And if the foreign deposits or bonds are de- course for countries seeking to discourage foreign cur- nominated in foreign currencies, as they were in Asia rency borrowing during some transition phase is to and Russia, they can become an especially heavy burden avoid quantitative restrictions or prohibitions and to use for domestic borrowers if the currency declines in value, disincentives instead (such as the Chilean-style reserve putting the country's financial condition at risk if the requirement or the removal of tax deductibility of inter- domestic borrowers and foreign investors lose confi- est on such borrowings). dence in the currency and thus help make their fears a self-fulfilling prophecy. Exchange Rates in the Global Financial Marketplace As will be discussed, countries nonetheless may be No discussion of an appropriate global financial archi- able to reduce their risks by not fixing their exchange tecture can ignore the question of what exchange rate rates. When exchange rates are flexible, capital controls policy countries should follow. The 20th century has may be less necessary to insulate the economy against seen various types of exchange rate policies, but essen- excessive foreign currency borrowing. However, wild tially they reduce to three: some type of fixed or pegged swings in investor sentiment about the country's com- rate, freely floating rates, and variations in between, mitment to fighting inflation, among other economic sometimes called "managed floating" or adjustable or objectives, may produce unwelcome volatility in ex- crawling pegs. It is becoming the conventional wisdom, change rates that, in the absence of well-developed for- however, that as capital becomes increasingly mobile, 28 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION countries will be forced to choose one of the two ex- the currency could fluctuate, domestic firms and their tremes, in particular monetary union with one or more creditors would have entered into very different borrow- countries (such as some members of the European Union ing arrangements than actually occurred. Creditors have decided to do) or purely flexible rates (which comes would have demanded higher interest rates to compen- close to the current policy of the United States and that sate them for the currency risk. Higher rates, in turn, of Asian countries that have been forced to abandon would have discouraged many Asian firms and banks their pegs). from engaging in "currency arbitrage": borrowing dol- Choosing between the two extreme exchange rate op- lars at low interest rates and then investing the proceeds tions, however, is difficult and depends on a host of eco- in domestic projects offering potentially much higher nomic and, perhaps even more important, political con- returns, but in domestic currency. The net result is that siderations (Eichengreen 1994). Generally speaking, it foreign currency borrowing may not have risen so far is in the economic interest of small, open economies or out of line relative to foreign currency reserves and the larger economies with flexible labor markets to belong crisis could have been avoided (or its severity consider- to a monetary union-that is, to irrevocably link their ably minimized). In short, the Asian financial crisis has currencies to the value of one or more others, or to yield strengthened the case for developing countries to float control over their currencies to a multijurisdictional or- their exchange rates. ganization. Small countries can benefit by "free-riding" It is important not to oversell the benefits of flexible on the reputations of other currencies (in the case of exchange rates, however. For developing countries in fixed rates) or on the reputations of a multicountry cur- particular-which generally do not have well-developed rency block (European Monetary Union). The more forward markets enabling parties to hedge their currency open an economy is the more its participants benefit, risks-the uncertainty associated with movements in through lower transactions costs, by having its currency flexible rates may impose significant transactions costs fixed in terms of other currencies. And the more flexible on importers and exporters. Moreover, even floating the labor market is, the less countries need flexible rates rates can be subject to speculative attack or "bandwagon to adjust real wages between regions experiencing differ- effects" if, for any number of reasons, domestic or for- ent levels of unemployment (accounting for the benefits eign investors engage in panic selling, dumping the cur- of a single currency in the 50 states of the United States rency before they believe it will fall further in value. Al- where labor mobility is high, but raising a question ternatively, excessive optimism can lead to speculative about the wisdom of European countries adopting a sin- bubbles in exchange rates, just as they can with the price gle currency at a time when few citizens of European of any asset. Fluctuations in exchange rates not justified countries move across national borders to find jobs). by underlying fundamental economic conditions can At the other extreme, flexible rates are in the eco- complicate economic policy management as exchange nomic interest of countries that have strong institutions rates affect domestic inflation rates through their effect in place to contain inflation (and thus do not need the on import prices. Indeed, a strong argument for fixing discipline afforded by fixed exchange rates) or are vul- the exchange rate-either pegging to another one (or a nerable to macroeconomic shocks not experienced else- basket of currencies), adopting a currency board, or join- where (perhaps requiring changes in monetary policy re- ing a monetary union-is that it eliminates this source gardless of their impact on the exchange rate). Even with of domestic inflation. a flexible rate regime, countries still have a strong inter- In the end, political considerations often prove even est in ensuring that they and their major trading part- more important than the foregoing economic factors ners follow sustainable macroeconomic policies that in determining which exchange rate regime countries lead to stable inflation and real growth. For it is precisely choose. A political barrier to monetary union is the when macroeconomic policies depart from this objec- demonstrated desire of governments and their citizens tive that financial markets force adjustments in the ex- to maintain their own currency. That hurdle can be change rates. Along with multimember groupings such overcome, but only when some other political objective, as the G-7, the IMF can help countries coordinate their such as the desire of the European members of the EMU macroeconomic policies so that wild swings in exchange to cement peaceful relations among themselves, assumes rates are avoided. overriding importance. In fact, had the Asian countries at the center of the 1997-98 crisis permitted their currencies to float much The Role of International Financial Institutions earlier in the decade, they might never have found Finally, what role should the major international finan- themselves in difficulties in the first place. Knowing that cial institutions-the World Bank and the IMF-play in TOWARD A GLOBAL FINANCIAL ARCHITECTURE FOR THE 2IST CENTURY Z9 facilitating the construction of resilient global financial ful information to private actors on a timely basis so that architecture? This is a broad topic, but several functions markets can allocate funds more effectively. remain especially important and deserve highlighting. Minimizing Moral Hazard The World Bank Although IMF lending packages do not protect in- A central objective of the World Bank should be to grad- vestors, in recent years they have tended to insulate for- ually wean client states off official assistance and encour- eign creditors from significant losses. The IMF can age them to mature to the point where they are funded reduce this well-known source of moral hazard by en- entirely by private markets. The funds of the Bank are couraging improvements in national financial super- scarce and so the more borrowers that graduate from the vision and regulation. But the Fund shouldn't stop Bank's programs, the more funds become available for there. the most needy countries. The Bank can best achieve its Eichengreen and Portes (1995) have outlined a series goals by providing technical assistance, preferably in co- of possible steps, such as allowing the IMF to lend to ordination with the IMF, to help countries upgrade the countries even when their private banks and other bor- legal, technical, and managerial infrastructure that is re- rowers are behind in their loan payments to foreign cred- quired for financial institutions to participate in global itors (which would reduce the power of those creditors markets. to "hold up" new loans until they are provided govern- An equally important job for the World Bank is to ment guarantees of their loans) and amending Article provide financial and technical assistance to enable 8.2(b) of the IMF Articles of Agreement to authorize the countries to establish better social safety nets, such as Fund to impose "standstills" on debt payments owed to unemployment insurance and job training, that are foreign creditors until the loans are renegotiated. common in the West but not elsewhere. This challenge A more aggressive step would be to condition IMF is much more closely related to the financial agenda lending on borrower countries imposing a "haircut" on than may first appear because financial crises often bring foreign currency denominated credits to their banks un- substantial increases in unemployment in their wake. less the creditors roll over their loans or not withdraw Social safety nets act as a cushion that brakes the impact their deposits. Less drastically the IMF could impose of such crises on aggregate demand, as well as meeting higher penalty rates on loans to countries that have not very real human needs for aid in difficult times (Stiglitz adopted such a system (Shadow Financial Regulatory 1998). In the absence of an adequate safety net, politi- Committee 1998). The haircut's virtue is that it ensures cal pressures can build within countries to reject move- that foreign currency creditors take some losses, which ments toward markets. could be tied to the discount on sovereign debt of the The Bank and the Fund should work together to country prior to the IMF lending package or to some promote the adoption and use of the tools for effective preset amount. At the same time, by only imposing the bank and financial supervision already outlined. In par- haircut if creditors withdraw their funds, this proposal ticular, both institutions can assist regulators in these would supply powerful incentives on creditors not to countries with techniques for valuing loans (especially run once the IMF steps in and thus would exert a stabi- nonperforming loans) and measuring interest rate and lizing effect on the currency and the banking system. To currency risk. They can and should also assist countries prevent domestic governments from using the proceeds with identifying and measuring areas of lending and ac- of the IMF lending packages to guarantee lenders, IMF tivity displaying unusually rapid growth, and providing loans could also be conditioned on the government not advice about the best means for discouraging excesses in extending guarantees to any more than the net amount this regard. of the foreign currency loans (the nominal principal minus the haircut). IMF An alternative approach to limiting the moral hazard While the IMF remains critical to provide emergency associated with IMF lending is to impose limits on cri- assistance to help stabilize currencies in distress, espe- sis lending. A 1999 Council on Foreign Relations study cially where falling currencies threaten systemic conse- recommended, for example, that ordinary IMF borrow- quences to the global financial system, the Fund also ing be limited to perhaps two or three times a country's must play a central role in reducing the likelihood of credit tranche. Any larger packages-designed to ad- future crises. It can do this by taking steps to minimize dress global systemic risks-would have to be provided the moral hazard entailed in future emergency lending by a new facility, subscribed to by large industrialized packages, while encouraging the provision of more use- countries, and approved by a supermajority vote. 30 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION One suspects that whether any formal means are Board of Governors of the Federal Reserve System. 1999. "Using adopted to limit IMF lending, the Fund is likely to do Subordinated Debt as an Instrument of Market Discipline." so on its own initiative, but on a case-by-case basis. In- Staff Study 1972. Washington, D.C.: Board of Governors. Caprio, Gerard Jr. 1998. "Banking on Crises: Expensive Lessons deed, the Fund has already shown an unwillingness to from Recent Financial Crises." Policy Research Working Paper provide lending to Russia in 1998 and Ecuador in 1999 1979. Washington, D.C.: World Bank Development Research because these countries did not change their economic Group. policies in ways that conformed with the Fund's policy Claessens, Stijn and Thomas Glaessner. 1998. Are Financial Sec- recommendations. Given the Fund's limited funds, and tor Weaknesses Undermining the Asian Miracle? Washington, D.C.: World Bank. the unlikely prospect that industrialized countries will Council on Foreign Relations. 1999. Safeguarding Prosperity in augment them any time soon, the Fund is therefore a GlobalFinancial System. Carla Hills and Peter G. Peterson, likely to take a harder line in the future in its emergency Co-Chairs, Task Force Report. lending than it has in the past. Eichengreen, Barry. 1994. International Monetary Arrangements for the 21st Century. Washington, D.C.: The Brookings In- stitution Press. Enhancing Information Eichengreen, Barry and Richard Portes. 1995. Crisis, What Cri- A necessary but not sufficient condition for markets to sis? Orderly Workouts for Sovereign Debtors. London: Centre operate properly is for actors to have comprehensive, re- for Economic Policy Research. liable information on a timely basis to inform their de- Goldstein, Morris. 1997. The Case for an International Bank- cisions. One response to the Mexican peso crisis was for ing Standard. Washington, D.C.: Institute for International cisions. One response to the Mexican peso crisis was for Economics. the IMF to encourage countries to provide more finan- Group of Thirty. 1998. International Insolvencies in the Finan- cial information about their economic and financial cial Sector. Washington, D.C. conditions-indeed, to post it regularly on the Internet. International Monetary Fund. 1998. Balance of Payments Statis- Nonetheless, the Asian crisis demonstrates that infor- tics Yearbook 1998. Washington, D.C.: International Mone- mation was lacking about two key financial indicators tary Fund. mation was lacking about two key financial indicators Kindleberger, Charles P. 1978. Manias, Panics and Crashes. New that are highly relevant to investors' perceptions about York: Basic Books. the status of a country's currency: the country's net in- Levine, Ross. 1996. "Foreign Banks, Financial Development, ternational reserve position (gross reserves minus for- and Economic Growth." In Claude E. Barfield, ed., Interna- ward crrenc commiments and te forign curencyti.onal Financial Markets: Harmonization Versus Competition. ward currency commitments) and the foreign currency Washington, D.C.: The AEI Press. debts of the country's domestic firms and their foreigJn Wahntn'..:TeAIPes d o c ' s rf Lindgren, Carol-Johan, Gillian Garcia, and Matthew I. Saal. affiliates. These data are not consistently gathered now 1996. Bank Soundness and Macro-economic Policy. Washing- by many developed economies, let alone those that have ton, D.C.: International Monetary Fund. been at the center of market turmoil in 1997-98. Litan, Robert E. 1994. "Financial Regulation." In Martin Feld- Nonethless,the eperiece ove the ast to yeas hasstein, ed., American Economic Policy in the 1980s, pp. 5 19-557. Nonetheless, the experience over the past two years has Chicago, IL: National Bureau of Economic Research and Uni- amply demonstrated the need for investors to know the vest ofCiaoPes , ,, ^ , ~~~~~~~~~~~~versity of Chicago Press. reserve and debt positions of countries in advance, not Litan, Robert E. (with Jonathan Rauch). 1997/1998. American after the fact. A major challenge for the Fund, and all its Finance for the 21st Century. Washington, D.C.: U.S. Trea- members, will be to put systems in place to gather these sury Department, 1997; Washington, D.C.: Brookings Insti- data and have them publicly reported on a timely basis. turion Press. Minsky, Hyman. 1982. Can "It" Happen Again? Essays on Insta- bility and Finance. New York: M.E. Sharpe. Conclusion Shadow Financial Regulatory Committee. 1998. "International Monetary Fund Assistance and International Crises." Stare- In sum, there is a continuing role for the International ment No. 145 (see www.aei.org). Financial Institutions, as the world enters the 21st cen- Stiglitz, Joseph. 1998. "The Role of International Financial In- tury. However, there also is a need for both the World stitutions in the Current Global Economy." Address to the Bank and the International Monetary Fund to concen- Chicago Council on Foreign Relations, February 27, 1998. trate their missions on problems and issues likely to be Summers, Lawrence. 1998. "Remarks before the International Monetary Fund." March 9, 1998. characteristic of the next century, drawing upon the Tesar, Linda L., and Ingrid M. Werner. 1998. "The Internation- ample lessons from recent events. alization of Securities Markets since the 1987 Crash." In R. E. Litan and A. Santomero, eds., Brookings-Wharton Papers Bibliography on Financial Services, 1998, pp. 281-349. World Bank. 1997. Private Capital Flows to Developing Coun- Bank of International Settlements. 1992. "Recent Developments tries: The Road to Financial Integration. A World Bank Policy in International Interbank Relations." Basel, Switzerland. Research Report. Washington, D.C.: World Bank. chapter 4 Prerequisites for Successful International Monetary Policy Coordination Andrew Hughes Hallett University of Strathclyde, Glasgow, and Centre for Economic Policy Research, London he second half of the 1990s has seen gional or a worldwide affair, and it might involve agree- a series of financial crises, first in ments on fiscal and monetary discipline or measures of Central America, but then more seri- financial support and lender of last resort facilities. But ously in East Asia. These disturbances it might also have to deal with issues of policy mix within caused major disruption in terms of economies, exchange rate management, market flexibil- falling incomes, currency deprecia- ity (competition and trade policies, including policies on tions, inflation, and losses in investment activity and out- market access and corporate control), and structural re- put. Worse, they spilled over into neighboring countries, forms, as well as with reforms to a country's financial ar- until then unaffected, through contagion in the financial chitecture. These are all local issues requiring local deci- markets and disruptions in trade-developments which sions, which must be taken in conjunction with the then exaggerated the crises at home and began to threaten international coordination designed to counter the diffi- the markets of those who might organize loans or a culties created by the effects of increasing globalization. bailout. This raises the question of whether some form of One objective of this chapter is to evaluate the bene- monetary and fiscal cooperation might not have reduced fits of monetary coordination for stabilizing and en- the severity of the crisis once it had happened, or whether hancing economic performance among the smaller and some mild forms of cooperation, involving exchange rate developing economies in the face of increasing global- regimes or a redesign of the financial architecture in par- ization. We are also interested in what drives those ben- ticular, might not have reduced the probability of a crisis efits and how they would be distributed. A second ob- in the first place-or at least have allowed the burden of jective is to stress that monetary conditions are not the adjustment to be shared and the risks to be spread. But only economic feature that needs to be coordinated. A there are additional features which may also be impor- strengthening and more focused use of local policy in- tant: Are the domestic policies well coordinated with each struments is perhaps a prerequisite for successful mone- other and with the financial policies of the G8 economies tary coordination; that is, we need to ensure sufficient which provide much of the financial and capital flows to support from competition and free trade policies and these economies? And are there sufficient supporting from certain structural reforms in the labor and finan- policies in place to provide the market flexibility needed cial markets. for making adjustments if a crisis does strike? These two objectives create, in effect, a distinction This chapter reviews the advantages and likely char- between absolute coordination and relative coordination. acteristics of more effective coordination between the The former has to do with the overall stance of policy economic and financial policies in countries subject to in a region or within the G8 countries and is designed large economic or financial disturbances-particularly to capture the efficiency gains of better-designed poli- those that come through the increased competitive pres- cies and a better economic performance. That may well sures of globalization. Such coordination might be a re- involve greater fiscal and monetary discipline so that, in 3' 32 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION pursuing their separate interests, countries do not pro- resulting in a crisis (for example, in the problem of in- duce excessively tight, loose, or risky conditions for a re- ternational debt, the Stock Market crash of 1987, Ger- gion as a whole. But against that, we have relative (or in- man Unification, or the liquidity squeeze following the ternal) coordination, which is concerned with getting financial crisis in Asia). The institutional implications the policy mix right and financial stability right within of this are that the world's major economies should in- each economy. It tends to focus on exchange rate behav- stitute suitable disaster relief plans, hold sufficient finan- ior, market flexibility, market access, trade balances, and cial resources in reserve, or impose suitable surveillance supply side reforms. mechanisms and require that certain financial safety A Taxonomy of Monetary Coordination measures be observed at all times. But they or their agen- cies should not interfere in any country's policies or pol- Most people would probably think of economic policy icy stance on a day-to-day basis. Thus a lender of last re- coordination as the business of the G8 world economic sort facility (but at a penal rate, and to central banks only summit meetings, the International Monetary Fund in order to minimize moral hazard) or an agreed set of (IMF)-World Bank meetings, or regional summits like capital adequacy standards for risk management in the those conducted by the European Union's council of min- financial markets and a jointly operated surveillance sys- isters or by the Asia-Pacific Economic Council (APEC) tem could all be part of this arrangement. But notions of ministers. However, there is a whole range of simpler and a world central bank, a policy of restructuring banks politically more manageable forms of coordination, many without a free market in corporate control, or programs of which can be conducted on a routine basis. We de- of structural adjustment and restrictions on monetary scribe them next, in ascending order of sophistication. and fiscal policy, or target zones for exchange rates, That said, the potential benefits of coordinating policies would not be. internationally, or between institutions, will always de- Avoiding conflicts over shared target values. Shared tar- pend on the level and degree of cooperation between the gets arise where countries actually target the same vari- agencies responsible for overseeing trade, or the capital or able (for example, a mutual exchange rate) or where they financial markets worldwide, and also between the coun- target variables linked by some identity which is not cap- tries that operate in those markets. able of being relaxed by policy interventions (for exam- Information exchanges. Here countries (institutions) ple, a set of current accounts). Here coordination could would exchange information freely about their targets, take the form of agreements that prevent countries or in- priorities, and information (including expectations) stitutions targeting mutually inconsistent values for what about external events, as well as about how they expect is actually the same variable, or agreements that prevent each of those factors to affect economic performance in them attempting competitive policy changes (for exam- the countries and markets in their charge. However, ple, devaluations) that cannot all be achieved simultane- given that information and including any announced or ously. In either case, policy effort would be saved and proposed policy changes elsewhere, they would continue better target values achieved for both parties if countries to make their decisions in a decentralized autonomous and institutions did not push against each other in vain way. By coordinating their information, policymakers efforts to improve their own (national) outcomes. Insti- will improve outcomes by eliminating incomplete or tutional forms of this would include target zone regimes faulty information over the intentions, expectations, or for exchange rates, an absence of capital controls, and an policy responses of others. On the other hand, there is al- effective scheme for modifying policies so as not to pass ways the danger that by sharing faulty information about current account deficits or liquidity shortages around be- external shocks, policymakers could spread those errors tween trading partners. around the whole system. Nevertheless, existing research Intermediate targeting schemes. Similarly, a limited (for example, Hughes Hallett 1986a) has suggested that amount of coordination may be achieved when coun- the major gains come from this aspect of coordination, tries agree to jointly control the variables forming the and that those gains may often be rather robust to shared main links between their economies or that cause the information errors (Hughes Hallett 1987c). main spillovers (externalities) from one to another. In Crisis management. Here coordination would take this case, cooperation is achieved by targeting the link place in response to episodes of particular difficulty in or spillover variables directly. Coordination is therefore the international economy and would involve policy achieved by using certain variables that affect the out- changes particular to that episode. It might involve ad comes in several economies, as a surrogate for the targets hoc policy adjustments, where misalignments and cur- that policymakers really care about. Such intermediate rent national policies or regional policy regimes interact targets may or may not be shared variables, although the PREREQUISITES FOR SUCCESSFUL INTERNATIONAL MONETARY POLICY COORDINATION 33 scope for fruitless competition over shared targets make ordination, which might prove important in practice. them the most obvious examples. Thus exchange rate But it does not say whether coordination can be ex- targeting would be one example, but monetary target- pected to yield significant benefits overall, or who would ing and the supply of liquidity or credit would be an- gain and why, or whether coordination would be an ef- other, while deregulating markets would be a third, and fective mechanism for absorbing financial shocks and debt management a fourth. preventing the kind of economic disruptions that we Partial coordination. Partial coordination takes place have seen in the Asian economies recently. when countries cooperate in achieving certain targets, Nor, more importantly, does such a classification of but may choose other targets noncooperatively or ac- policy coordination schemes give any idea how these cording to some preassigned national rules. It is often schemes might actually be implemented in practice. suggested that countries need to coordinate their mone- One of the lessons from the G3 and European attempts tary policies, leaving fiscal policies to the domestic poli- at coordination is that it is often extremely hard to get cymakers. This reflects the thinking in Europe's mone- agreement on measures that cover all the instruments tary union. On the other hand, fiscal policy needs to be and targets of policy, or to secure agreement on how to sustainable if it is not to cause liquidity shortages and treat those who deviate from their part of the agreement. high interest rates for others if it becomes excessively Similarly, a proper coordination of policy usually makes expansionary.' That implies a minimal degree of fiscal greater demands on information and modeling of eco- coordination. The latter might be achieved by agreeing nomic responses than do simpler forms of policymak- on policy assignments: fiscal policy for internal targets ing. It is therefore often thought to be more vulnerable (growth and employment) and monetary policy for ex- to unexpected shocks or to errors of information or ternal and link targets (liquidity, capital, trade). The dif- specification. Indeed, it can be very hard to secure agree- ficulty here is that a great deal of design work may be ment on how to respond to such shocks and errors, es- needed to devise agreed, but country-specific policy pecially when they are distributed asymmetrically across rules that can provide gains for each participant over the participants. For all these reasons, it is important to have best they can attain otherwise. On the other hand, an idea of how policies may be coordinated, what sim- agreed rules make monitoring-and transparency-rel- plifications are effective in practice, and where the gains atively straightforward. are likely to come from. Full coordination. Full cooperation takes place when countries agree to a certain bargain describing how all The Benefits of Full Coordination their targets and their fiscal, monetary, and exchange Studies that have attempted to evaluate the likely gains rate instruments shall behave. This would aim to maxi- from full coordination have generally concluded that mize the gains over the best outcomes that could be at- those gains exist but are small. However, those studies tained from noncooperative policymaking, subject to a have been restricted (almost exclusively) to the major in- reasonable distribution of those gains. As in the previ- dustrialized economies. For example, in their pioneer- ous two cases, the key is that the bargain is between a ing study, Oudiz and Sachs (1984) estimated that of the number of freely contracting parties, be they countries gains among the Group of Three countries in the mid- or institutions. That means each party must be able to 1970s, coordination would be worth no more than 0.5 make gains over what they could achieve with indepen- percent of gross national product (GNP) to each coun- dent policies. However, full cooperation may be very try compared with the best noncooperative outcomes. hard because it is politically difficult to reach agreement It may, of course, be debated whether 0.5 percent of across all the targets and instruments of policy-espe- GNP is only a "small" gain. It would represent a signifi- cially when there is significant uncertainty about the cant amount of extra productive capacity if fully in- exact impact of those instruments or whether the exter- vested in each period. But it is not large compared with nal variables will turn out as expected. And, because the annual growth rates. Moreover, if gains of this size are policy changes are unlikely to follow any simple rules in not much bigger than the forecast standard error of the this case, monitoring or penalizing those who deviate target variables (that is, if the gains are small relative from the agreement may also be difficult. to the imprecision with which policies can be imple- mented), then it may not be easy to persuade policy- lnstitutional Prerequisites to Secure the Gains from mkr ftervle makers of their value. Monetary Coordination Later studies have confirmed that the gains from coor- The previous discussion gives an idea of the types of dination among the Organisation for Economic Co- monetary policy coordination, with or without fiscal co- operation and Development (OECD) economies are 34 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION likely to be small. In studies that allowed for dynamic financial crises, a collapse of export markets, or a li- decisionmaking,2 Hughes Hallett (1986a, 1986b, 1987a) quidity crunch), or where governments need to be re- found the gains to be somewhat larger-between 0.5 per- strained from implementing a sequence of entirely in- cent and 1.5 percent of GNP for the United States, the appropriate policies (whether designed for their own European Union, and Japan. Subsequent work by Can- private advantage, a general lack of "reputation," or zoneri and Minford (1986), Minford and Canzoneri through ignorance of the pressures on their economy). (1987), and Currie et al. (1987) have also suggested rela- tively small gains in the absence of major shocks, based on In any of those cases, coordination may prove more calculations for the United States and the EU or OECD, advantageous. We consider each in more detail below. respectively. However, all those results have turned out to vary significantly with the size and persistence of external How the Gains from Coordination Are Distributed shocks and the perceived reputations of the governments Little work has been done on how the gains from coor- concerned. In fact, persistent shocks and the existence of dination might be distributed (relative coordination), as "reputation" among government and central bank policy- opposed to how they might be created (absolute coordi- makers appear to increase the value of coordination more nation). This is a pity since it will be difficult to persuade than anything else (see Currie et al. 1987). policymakers to cooperate if the gains are not distributed Regrettably, there have been very few exercises that in a reasonable manner and are not assured to be posi- were not concerned with the major industrial economies tive for each participant. In the event, Oudiz and Sachs or with the coordination of the standard instruments of (1984) found gains distributed roughly 2:1 in favor of macroeconomic policy. So it is hard to tell how well Germany, relative to the United States. Hughes Hallett's these "stylized facts" carry over into cases involving de- (1986b) study of the United States and Europe again veloping or newly industrialized countries, or to coun- suggests gains distributed 2:1 in favor of Europe using a tries with underdeveloped capital and financial markets, wide range of bargaining models. Later work showed this or where there are trade barriers or other market imper- result to be somewhat sensitive to alternative types of ex- fections. However, small gains on average do not always ogenous shocks (Hughes Hallett 1987b), although in no mean small gains for everyone. Coordination tends to case was the position of Europe as main gainer over- be more important for the following: turned. Hughes Hallett et al. (1989) also found the gains to be asymmetrically distributed among the Group of * Developing countries, or those on the weaker side of Five countries in the late 1980s and point out that it can an asymmetric relationship with their trading partners be extremely difficult to find ways of improving the lot or the suppliers of capital (see Currie and Vines 1988; of those countries that benefit least from coordination. Hughes Hallett 1988; and Brandsma and Hughes These are important results because they suggest that Hallett 1984) (1) it will be hard to secure and maintain a coordination * Countries subject to persistent shocks or a lack of rep- agreement in the face of significant uncertainties, and utation in policymaking (2) if those who make the gains and those who shoulder * Situations where there are significant market distor- the burden of adjustment are different sets of people, se- tions or restrictions, and in particular where free trade curing any agreement at all will be politically difficult. is abandoned or where free trade zones discriminate against outsiders (Helkie et al. 1989; Hughes Hallett Coordination Substitutes: Information Exchanges 1992a; Hughes Hallett 1994a) The gains from coordination relative to noncoordina- * Situations where coalitions among subsets of players tion may well be substantially smaller than the gains are easily formed of efficient noncooperative policies over strategies that * Episodes where exchange rate stability is important in ignore predictable foreign policy changes altogether its own right, as opposed to it being simply a means (Hughes Hallett 1986a). Thus, coordination in the sense to an end (that is, where the exchange rate is an inter- of information exchanges, rather than detailed coordina- mediate target designed to achieve a superior perfor- tion across all variables, may supply the major part of the mance in the other targets of economic policy; improvements available from policy coordination. If this Hughes Hallett 1992b) is so, an important function of international forums for * In cases where optimalnoncooperative behavior is not policy discussion will be the exchange of information the alternative; for example, where large exchange rate among policymakers concerning their policies and the fluctuations or large unanticipated shocks disrupt a state of their economies. Interestingly, this conclusion sequence of rationally chosen policies (for example, in can hold even when the information exchanged is found PREREQUISITES FOR SUCCESSFUL INTERNATIONAL MONETARY POLICY COORDINATION 35 to contain prediction errors (Hughes Hallett 1987b). It cooperative alternative without targeting (see Currie and has also been confirmed in a series of experiments car- Wren-Lewis 1989; Hughes Hallett 1992b). That sug- ried out by Minford and Canzoneri (1987) on the larger gests that exchange rate targeting yields gains, not be- OECD economies. Hence poor information may be less cause it is a particularly effective substitute for full co- of a problem than a mismatch of policies. ordination, but because the historically chosen policies Thus, information exchanges do appear to be a key were usually inappropriate for the job in hand. That, in part of the coordination process, irrespective of the model turn, means that the major advantage of exchange rate or time period. One might also suppose that the wider targeting schemes is that they restrain governments from the range of policies reviewed, the greater the benefits choosing silly policies-whether in their own private in- of information exchange to the decisions subsequently terest or through sheer ignorance. That is a particularly made (Bryant 1987). Certainly prior consultation would important result in the context of large financial shocks alert policymakers to potential and self-defeating con- which threaten to pervade the whole system. flicts such as incompatible exchange rate or trade balance However, these results all refer to coordination includ- targets that would lead to competitive appreciations or ingexchange rate targets. If exchange rate targeting is un- depreciations, or inconsistent fiscal and monetary pro- dertaken with no explicit coordination, the gains are grams. Prior consultation could also help policymakers much smaller: typically 0.5 percent of GNP, and then not avoid any losses attributable to conditioning their own necessarily for all participants (Hughes Hallett et al. decisions on erroneous information about other policy- 1989). That runs the risk of creating an incentive incom- makers' intentions (for example, what priorities they patible, and hence unsustainable, regime. Subsequent have, what target paths they aim at, what model they use studies have been able to confirm analytically that ex- for policy selection). Whatever the difficulty of predict- change rate targeting can be no more than an imperfect ing the true state of the world, errors owing to mistakes substitute for proper economic coordination (Hughes should be avoidable, and there is no point in adding Hallett 1994b), but that there are ranges of circumstances them to the unavoidable (genuinely random) errors. in which it would be a good substitute and others where it would be a poor substitute. For example,3 exchange Coordination Substitutes: Exchange Rate Targeting rate targeting is helpful in the following circumstances: None of these early studies of policy coordination con- sidered exchange rates to be a target of policy. Yet much * In economies that are structurally dissimilar but sub- of the recent policy debate has been concerned with ex- ject to symmetric shocks or in economies that are struc- change rate management, with the aim of either stabi- turally similar but subject to country-specific shocks. lizing exchange rates or of making controlled realign- This plays off the fact that the potential for gains from ments. This may serve as a means of improving relative coordination is driven (in large measure) by the asym- coordination among countries. metries between economic responses (see below, p. 36). The Hughes Hallett (1992b) study points out that * Also in cases where the exchange rate target can be set although exchange rates may be included among the tar- to reinforce the positive spillovers between economies gets during policy selection, they can either be included (and reduce the negative), so that either smaller inter- in the associated objective function evaluations (in which ventions are needed to support the policy targets (that case exchange rate stability is a target in its own right) is, the intervention costs saved allow the instruments or excluded from those evaluations (in which case ex- to be used more intensively on the targets they are change rates are,just an intermediate target, instrumen- most effective at achieving, and the gains from that tal in securing improvements elsewhere). If the former are more than enough to pay for the sacrifice required holds, the gains from coordination appear larger than to reach the extra target), or that better target values before, about 3-6 percent of GNP (equivalent to 10-20 are achieved for the same level of interventions (with percent improvements in the targets themselves) as esti- the same implication that the target improvements mated across seven multicountry models. If exchange more than pay for the interventions needed to sup- rates are treated merely as intermediate targets, the gains port the extra target). are significantly smaller and much the same as in the * Similarly where the shocks are destabilizing (that is, earlier literature (1-1.5 percent of GNP). they reinforce the spillover effects between economies; Perhaps the most interesting result here is that the this will happen if, but not only if, the shocks or the gains from exchange rate targeting over the policies his- responses are asymmetric between countries), but, torically chosen are typically several times larger than conversely, exchange rate targeting is unhelpful if the the gains of exchange rate targeting over the best non- shocks are predominantly stabilizing. 36 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION * Finally, whether exchange rate targeting is helpful or invite speculative attacks, since they present speculators not must also depend on the sensitivity of the ex- with a "one-way bet" in that they need to commit only change rate to policy changes or shocks (because that limited funds to push a currency beyond its interven- determines the sacrifice necessary to support the extra tion boundary. And they may be able to trigger such an target), and on the responsiveness of the target vari- attack before the policymakers have had time to mount ables to the exchange rate (because this determines the a convincing counterattack. Widening the bands, by potential gains through better coordination), andon contrast, is unlikely to lead to much deterioration in ex- the target values chosen for the exchange rate path. change rate stability because it doesn't present the spec- Some authors have emphasized that exchange rate ulators with such an easy one-way bet. Larger funds will regimes are very unlikely to be helpful if the target val- have to be committed to the attack, and there is a much ues are chosen far from their equilibrium or "funda- better chance that the authorities will be able to inter- mentals" (see Wren-Lewis et al. 1991). That leaves vene and reverse the attack before the currency reaches open the question of how best to calculate those equi- its boundary. In that case the speculators will lose librium values.4 One answer to the latter problem is money, not gain. Knowing this in advance, speculators to make the target values respond to the deviations of will be less willing to risk an attack. Moreover, the au- other peoples' targets from your own, as well as to your thorities will have time to mount (and demonstrate their targets and the economic "fundamentals" (Hughes commitment to) corrective policies before the bound- Hallett 1998). It is in the latter case that exchange rate ary is reached-which is more convincing to the mar- targeting appears to provide a good measure of the kets than merely saying they would defend those bound- gains available from coordination. aries should the need arise. Speculators are therefore more likely to anticipate a return toward parity rather Three other observations are important. First, there than speculate on a realignment at the boundary as long are obviously circumstances where exchange rate target- as governments have the time to make their policy ac- ing is a poor substitute for coordination-and in partic- tions credible and transparent. ular it only takes one participant to be damaged for the Policy Assignments in an Increasingly system to break down, whereas all have to benefit if it is Golize Econmy to work. So the distribution of gains is crucial. It can be quite difficult to arrange that successfully. Second, sym- Assigning Your Own Policies to Your Own Targets metry obviously plays a role, but not in a very direct One result that appears in almost every study is that co- way. There are circumstances where asymmetries make ordination seems to generate more stability in the tar- coordination and exchange rate targeting helpful and gets and more continuity in the policy interventions circumstances where it is damaging. Equally there are (Hughes Hallett 1986a). The explanation, according to cases where symmetry helps and cases where it harms. Cooper (1969), is that ignoring independence within or Nevertheless, it seems that whereas asymmetries typi- between economies leads to oscillations and overshoot- cally appear to increase the potential gains from coordi- ing because the implied assignments (country by coun- nation, they tend to exaggerate the characteristics that try) ignore the international side effects of decisions cause exchange rate targeting to fail. By contrast, sym- made in the domestic interest until after they have ap- metries tend to reduce those characteristics. Finally, the peared. Corrections have then to be applied. degree of commitment by the participating govern- These are the costs of imposing policy assignments on ments is important. Success depends on the commit- an exclusively national basis (that is, domestic policy in- ment of all participants, not just some of them. But the struments to domestic targets). They are analogous to advantage of exchange rate targeting, if it works, is fewer the costs imposed by creating one-to-one policy assign- policy conflicts and a reduction in the scale of the inter- ments internally, for example, monetary policy to infla- ventions needed to resolve domestic or foreign policy tion, and fiscal policy for output stabilization, as hap- problems. pens when the central bank is granted independence. Finally, there is the question of speculative attacks. Typically one would expect internal coordination (rela- Any target zone scheme may suffer speculative attacks, tive coordination) to be more important than the coor- although, given a credible commitment to defend the dination of policies or regimes between countries, since intervention boundaries and a parity in line with the any changes or shocks go directly through the domestic economy's "fundamental equilibrium," there is no par- goods and financial markets in the former case. But in ticular reason to expect them. An important point, how- the latter, they have to go through the markets for traded ever, is that exchange rate bands that are too tight will goods or capital movements before they get to the do- PREREQUISITES FOR SUCCESSFUL INTERNATIONAL MONETARY POLICY COORDINATION 37 mestic markets. Hence countries that are particularly terns, but it can only be partly fixed by greater coordina- sensitive to trade flows, exchange rate movements, or tion at a global level. It also requires increasing coordina- capital flows (driven by interest rate differentials, ex- tion and flexibility among policymakers at the local level pected devaluations, or the threat of debt, insolvency, or (that is, within countries or within particular markets). liquidity problems) may find that their stability is also Indeed, without the latter, coordination between coun- particularly sensitive to international coordination. Con- tries or blocs of countries will become much more diffi- versely, other types of countries would be more sensitive cult. We therefore distinguish between the following: to the advantages of relative (internal) coordination. Thus coordination, both internal and external, is impor- * No coordination. Domestic instruments are assigned tant in the following circumstances: to domestic targets on a one-to-one basis (as in Mun- dell's assignment of fiscal policy to achieving internal * When trade contributes a large share of national income targets, and monetary for achieving external or mon- * Where capital inflows or imports are a key component etary balance). of domestic production, or exports a key element in * Internal coordination. Domestic instruments are used the development strategy simultaneously to achieve domestic targets, but they * Where the exchange rate can be significantly affected take no account of foreign targets. The assignments by changes in the policy mix in the major trading are national. partners * Full coordination. Domestic and foreign instruments * And, most of all, where the domestic financial and are used together to achieve the set of all domestic and capital markets are incomplete or too small, with the foreign targets. result that domestic investment and production have to be financed by foreign capital inflows. Thus coordination is an important element in any well-balanced set of policies. The first regime represents Finally, "vertical" coordination, that is, linkages that decentralization both internally and externally, whereas make elements in the internal coordination problem ad- the other two regimes represent policy arrangements that just to support the external one more effectively, can also are noncooperative and cooperative externally. Studies be important for the final outcomes and their stability. of these different regimes have shown the following: Strengthening an economy's local policy instruments to provide that support is therefore a necessary part of any m Target overshooting and oscillations are removed by coordination exercise. internal coordination, which no longer ignores side It is worth reflecting on the intuition behind these effects until after they have happened. The same hap- observations. Essentially the message is that assignments, pens if the assignments of national instruments to na- whether internal or international, render domestic poli- tional targets are eliminated. cies less effective as policymakers have to correct past * Increasing degrees of coordination (for a fixed degree "mistakes" and counter the undesired impacts of policies of interdependence) damp out the transitory effects introduced by others for the benefit of others. Moreover, of shocks more rapidly. On the other hand, increas- because the strength of these spillovers increases with the ing degrees of interdependence (with policies fixed) degree of interdependence (globalization), the effective- extends the size and duration of the target distur- ness of policy in decentralized economies declines along bances. Thus coordination increases the power of pol- with it. Decreasing policy effectiveness shows up in the icy interventions, while interdependence reduces the size of the policy interventions needed to restore targets effectiveness of policy. to their desired levels and in the speed with which the * One example of the increased costs of intervening in targets approach their desired levels after a shock. So in- an open economy is that interdependence slows the creasing integration means that in the absence of changes correction of trade imbalances (because other coun- in the way economies are run or policies made, one will tries take countermeasures such as competitive tight suffer either worse outcomes, or longer periods away money or competitive depreciations in order to pro- from equilibrium (exaggerated cycles), or greater inter- tect their own targets). The result is that one must vention costs in the form of what is necessary to restore hold a higher level of reserves to last out the slower the targets to their desired values. adjustments. The required level of reserves or foreign The key point is that all this is an unavoidable con- inflows is lower with coordination both because the sequence of increasing globalization, whether in trade, domestic adjustment processes are stronger (policy ef- capital flows, technological transfer, or consumption pat- fectiveness) and because smaller interventions mean 38 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION that the side effects (externalities) of national policies especially important to the smaller and developing coun- are reduced. tries to prevent sudden changes in those variables. * In the wider context of greater integration within and between economies, Mundell's policy assignment On the Importance of Internal Coordination proposition only survives in the form of policy spe- (i) Getting the policy mix right. The previous arguments cializations between or within interdependent econo- suggest that internal and vertical coordination-the lo- mies. Indeed, the assignment of national policies to calization of policy-are likely to be a necessary initial national targets is already a generalization of Mundell's condition for successful coordination at an international original idea, and international coordination is just to level. Hughes Hallett and Ma (1996a) provide examples exploit the same idea yet further. Moreover, Patrick of why it may be an important necessary condition. (1972) has shown that there is no generalization of They find that the coordination of fiscal and monetary Mundell's proposition to systems with many targets policies within one economy (that is, getting the policy and instruments or to systems where there is uncer- mix right) is important because policy assignments-or tainty over the policy responses. That means we will at least the absence of mutually supporting or consistent be compelled to rely on policy specializations rather policies-are costly in terms of lost performance and than exclusive assignments. Thus "comparative policy generation of larger oscillations and overshootings in the advantage," defined next, remains the key to success- targets (see also Demertzis et al. 1999a). Not only is that ful policies in economies that are interdependent. costly for the domestic economy, it can be highly dam- aging for the neighboring economies. Since the spillover Smaller Economies and Financially effects between sectors go through the domestic and in- Dependent Economies ternational markets rather than just the international Although set in general terms, much of the discussion markets, these effects may be large. And rules designed here is of particular importance for the smaller and de- to deal with policy assignments, short of coordination, veloping economies. Since those economies cannot, by will be accident-prone because they depend on particu- definition, do much to influence world economic condi- lar realizations of shocks. So internal coordination is re- tions or the pace of globalization, it will be especially im- ally the only option. portant for them to preserve and strengthen their policy (ii) Policy conflicts. The problem is not only that un- instruments at the local level. Hence all the arguments coordinated policies internally are inefficient and may here in favor of "localization" or "internal coordination" cause exaggerated swings in output or the exchange should be taken to refer to smaller and developing rate. There is also a danger that, in a world of compet- economies in particular, as indeed are the arguments ing policy targets and competing policymakers with dif- which say that the weaker economies are the ones to gain ferent preferences over which targets should have prior- most from the explicit coordination of economic policies ity (growth, employment, and investment vs. currency (see the next section). Stronger self-interested responses, and price stability or domestic vs. regional stability, and carefully coordinated with those of the world's larger so on), it is likely that uncoordinated policies will com- economies, will cause little disturbance to the major pete not only to reach their preferred targets, but also economies. So the smaller economies should exploit their to undo some of the effects of their rival's policies. opportunities for localization as far as possible, without The danger is that these offsetting interventions will fear of retaliation. leave the domestic economy to drift without effective A couple of specific examples make the point. Eichen- policies-at a cost to itself and to the stability of its green and Rose (1998) find that increases in OECD neighbors. country interest rates have often enlarged the probability Thus policy conflicts internally, or where domestic of a financial crisis in the developing economies. In that policies are set to undo or overcome the policies in a case localized responses in the developing countries, and rival economy, will mean that none of the policymakers a serious attempt at coordination on other issues (since will get what they want. That could easily trigger a self- the developing countries won't be able to have much im- interested political response. And if it does, both the pact on OECD interest rates), would be particularly conflicts and economic performance will get worse. In helpful. In the same way, capital flight is a perennial dan- Demertzis et al. (1999b) we find that a more conserva- ger for developing countries. It is likely to be linked to tive stance by one set of policymakers, for the sake of anticipated exchange rate movements, to changes in ex- achieving their own goals, will produce a more nation- ternal (or internal) debt, and to local investment oppor- alistic set of preferences from the voters, and hence gov- tunities vs. outside options. Again coordination will be ernments, in the rival economy or policymaking activ- PREREQUISITES FOR SUCCESSFUL INTERNATIONAL MONETARY POLICY COORDINATION 39 ity. Hence the electoral mechanism may create further prices will tend to fall further. These pressures will there- differences in priorities. And the more priorities are fore divert trade and investment from inefficient firms made to diverge by the electoral process, the greater the and countries while creating it for the efficient ones. So conflicts. Coordination therefore becomes more impor- if countries wish to protect themselves and profit from tant when we add in political responses, and the out- all these changes, they need a strong internal competi- comes are likely to get worse, if not more unstable, if tion policy to keep prices and quantities flexible and there is no coordination. to allow their other policies the scope and effectiveness (iii) An example: Policy mix and the sustainability of to create the most appropriate internal circumstances debts (see Beetsma and Bovenberg 1997; Beetsma and given the external pressures. In this sense a strong pro- Uhlig 1999). If monetary policy is increasingly dedi- competition policy is a necessary condition for dealing cated to controlling inflation in an economy with little with globalization and international shocks. internal coordination, it becomes necessary to use fiscal Trade restrictions-be they tariff barriers, quantity policy more vigorously to achieve output growth, stabi- restrictions, nontariff barriers, or restrictions on market lization, and employment. And the larger fluctuations access (including restrictions in the markets for corpo- that appear when there is little internal coordination will rate control)-will have exactly the same effect. It won't call for larger fiscal interventions to counteract them. If help much if an internally pro-competitive policy is deficits created in the downturn are not fully repaid by matched by external restrictions. Countries and blocs surpluses in the upturn, which is typically the case, then of countries must remain "open" rather than "closed" if countries will tend to accumulate debt and eventually they are to remain flexible enough to coordinate their face an unsustainable fiscal position. The outcome of internal affairs properly and reduce their vulnerability that could be price jumps (in anticipation of a bailout to the effects of globalization. A "closed" economy is one and monetization) or a default and capital rationing, in which external barriers are increased or maintained which could have a devastating effect in the form of as the internal are reduced; an "open" one is where they contagion because of the liquidity restrictions this can too are reduced. Of course trade barriers can be reduced imply both at home and for other countries borrowing as a unilateral act, but it is obviously more effective if in the same financial markets. everyone else is doing so as well. That makes low trade Indeed, trying to head off a debt problem before it barriers, with extensions into services and the financial happens can cause problems unless the policy mix is markets, a multilateral policy requirement from an in- carefully adjusted, because sharp fiscal contractions to stitutional point of view. reduce public sector debt will tend to reduce national The message here is that we need a "deepening" of income at the same time. As a result, the debt ratio will the number and type of relationships that can be used not fall by much. It may even rise because revenues will for coordination rather than just a "widening" of the ex- also fall with national income. To fix this problem, isting relationships. In the same way, exchange rate tar- countries need to match their fiscal contractions with geting schemes are helpful because they prevent coun- monetary expansions so that national income and rev- tries from maintaining over- or undervalued exchange enues do not fall with the debt. If they don't, they risk a rates for strategic purposes and to the cost of their rapid rise in real interest rates as deflation combines with neighbors. Fortunately none of these possibilities is tight money (a liquidity crunch). This will destroy in- likely to be tremendously important in practice as the vestment as well as output and hence employment and long-term benefits of strategic trade policy and exchange output capacity in the longer term. rate manipulation appear to be rather limited (Hughes Hallett 1992a). But there may be short-term opportu- Another Example: Policy Deepening Supporting nities to "kick start" changes that way or to manipulate Coordination with Policies at the National Level the monetary-fiscal mix to the same end. Hence, given The globalization of markets and investment will tend a sustained WTO program, coordination should be to reduce prices and increase output through increased used to ensure such policy changes are not introduced competition, scale economies, and a better use of com- at the wrong moment. parative advantage. Where the markets are already com- petitive and offer free access, there will be relatively lit- The Distributional Consequences of Coordination tle gain to be made. However, where the markets are One consistent feature in the results cited is that the imperfectly competitive (domestically) or fail to offer weaker party to the cooperative bargain gains most, free access (internationally), the potential gains will be which is a result that is easily explained. In any nonco- much larger. As the pressures for globalization build up, operative decision game or conflict, the stronger party 40 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION will always be able to capture the larger part of the ben- party is given a more dominant role-exactly as our ex- efits for two reasons. First, being stronger means the tar- planation would predict. Finally, and of special interest gets respond more strongly to your instruments, so you here, it also appears to be true for the less developed or can get closer to those targets than others do, and you commodity-based countries (the "South") in a game of have to expend less effort deflecting the effects of deci- growth and investment with the industrialized "North," sions made abroad. Second, being stronger or larger at least until debt repayments from South to North be- means that you can influence world monetary or trade come large enough to drain the South of resources conditions and make them more favorable to your own or until the North hits its capacity constraints (Hughes circumstances, perhaps "crowding out" the efforts of Hallett 1988). others to pull things their way. Is there anything more specific to be learned about Thus a stronger country is bound to make larger gains what drives the size of these gains from coordination? and exercise a degree of control, as long as things remain That is a question that has to be answered on a case-by- noncooperative. But to the extent that there are positive case basis, but it is quite clear that coordination always as well as negative spillovers from that better perfor- induces a degree of policy specialization. In the Hughes mance, the weaker countries will also benefit-in effect Hallett (1986a) exercise, the United States relies on free riding on the ability of the policy leader to choose monetary policy, while Europe concentrates on fiscal in- his policies correctly and improve economic performance terventions. Although this is not a policy assignment, it for all. This may suggest Stackelberg leadership; but it does agree with the relative efficiency of the policy re- must be true of Nash noncooperation too, since it de- sponses in the model: Countries use their policy instru- pends on the "locomotive" aspects of the world economy ments according to their comparative (policy) advan- being not a lot less than the "beggar-thy-neighbor" as- tage. The logic behind this is exactly the same as in pects. And if the problems of the smaller and less devel- international trade theory. By "trading" policies and re- oped economies are a matter of asymmetric dependence allocating intervention effort between the instruments through trade and capital flows, then globalization must according to their international comparative advantage, mean that the locomotive aspects are important even in countries can improve the outcomes in terms of their countries with financial markets that are incomplete or own utilities that those policies produce. These im- too small. provements could take the form of better target values Thus the weaker countries will benefit from sensibly for the same overall level of intervention, or the same chosen policies elsewhere, albeit to a smaller extent. 5 If target values with smaller interventions, or a mixture of we move now to coordination, it means the weaker coun- the two. The importance of all these results therefore de- tries get a chance to fine-tune their policies against a pol- pends on the asymmetries between economies. icy stance in the leading economies, which won't change A second result is that cooperative policies produce much through the act of cooperation. At this point the greater continuity in the instruments and greater smooth- advantage passes to the weaker countries: The stronger ness in the target paths compared to noncooperative poli- countries, having made most of their gains at the nonco- cies. Cooperation also advances the timing and increases operative stage, will have exhausted most of the opportu- the speed with which targets are achieved. This result con- nity to make further gains. However, with most of their firms Cooper's (1969) arguments: Cooperation damps chips on the table, they stand to lose more if things go out shocks more rapidly and reduces the costs of being wrong. The weaker countries can therefore exploit their out of equilibrium. chance to redesign their policies, building on the gains Third, cooperative policymaking involves smaller in- the leaders have already made and extracting some con- terventions as well as better target realizations; that cessions where advantageous-but not up to the point is, for less effort one gets greater success. Cooperation where the leader backs out or retaliates. That explains therefore reduces policy "externalities" imposed from why we consistently find that the weaker countries gain abroad. In the exercise quoted, Europe was relieved of most from coordination: See, for example, in Hughes the need to block spillovers from U.S. policy, and the Hallett (1986a, 1987a) where the United States "leads" United States was freed from the need to stimulate a the other OECD countries, and again in Hughes Hallett stagnant European economy. As a result, both countries and Ma (1996b) where the United States is stronger than can better dedicate their instruments to achieving their the rest of the G7 countries, but Germany is the hege- own targets. Coordination therefore restores policy ef- mon within Europe. Perhaps the best demonstration is fectiveness weakened by competitive policymaking. in Brandsma and Hughes Hallett (1984) where the final In summary, cooperation brings improvements be- gains to the weaker party increase steadily as the stronger cause of the following: PREREQUISITES FOR SUCCESSFUL INTERNATIONAL MONETARY POLICY COORDINATION 41 * Coordination restores policy effectiveness, which is ments cheat the private sector. This does not involve a weakened by interdependence and by expectations of bargain as such but a breach of what the private sector foreign reactions. This happens because cooperation had been led to expect. One example is that given by involves trading reductions in foreign externalities Rogoff (1985), who points out that international policy caused by domestic action against reduced externalities coordination may actually be welfare decreasing if the from foreign reactions. As a result, smaller interven- coordination process eases the constraints on govern- tions can achieve better expected outcomes on average. ments engaging in inflationary monetary expansions. * Coordination speeds up an economy's responses and Governments may conclude that it pays to join together dampens out the transitory effects or external shocks and cheat the private sector. The point here is that there more rapidly. is really more than one set of actors per economy so that s Coordination enables governments to extend the range coordination among governments may represent a coali- of comparative policy advantage, and hence the pol- tion against the private sectors, wage earners, and so on, icy specializations, which they can exploit. but not full coordination. Policy coalitions of this kind must be very common in practice, particularly between Asymmetries between countries therefore play a cru- governments or governments and institutions. But such cial role in determining the gains from cooperation. situations are not always welfare reducing. Carraro and Three kinds of asymmetry are of interest: asymmetric do- Giavazzi (1988) produce a different model with three mestic policy responses (which determine policy special- sectors (government, firms, and trade unions) to show ization), asymmetric spillover effects (which determine that cooperation will be welfare increasing in all circum- bargaining strengths), and asymmetric adjustment speeds stances. The only clear message here seems to be that co- (which determine the cost of policy responses needed). ordination with those with reputation will increase your What Are the Riskts in Coordination? credibility and hence the effectiveness of your policies. The question of cheating between governments has The previous sections have suggested that we do not received less attention. The argument here is that coor- need new forms of coordination, but "deeper" coordina- dinated policies will be sustainable against cheating if tion and perhaps more robust forms of coordination. If the losses that a country might suffer under retaliation the gains from coordination are not very large in normal or preemptive cheating by the opponent sufficiently out- circumstances, except for the weaker and more depen- weigh the gains that the first government might make dent countries, then maybe they can be made to gener- by cheating. Hughes Hallett (1 986b) finds that this con- ate more robust outcomes in an uncertain world. The dition is easily satisfied in a reflation game between the difficulty is that policymakers are not going to want to United States and the European Union; Currie et al. cooperate if (a) they feel other countries will not wish to (1987) similarly find that there is little incentive to re- stick with their part of the bargain; (b) if they think there nege on bargains struck among policymakers. are likely to be errors in the exogenous information or assumed shocks that will invalidate their decisions; (c) if -Exteral Shocks they think they may have (or may have been led to) There is some evidence that forecasting errors or exoge- wrongly estimated the preferences of others (a lack of nous shocks can cause significant disturbances and af- transparency); or (d) if they think that their partners dif- fect the incentives to coordinate. But those shocks have fer in their assessment of how their economies work or to be both large and persistent. Canzoneri and Minford that the models they use to evaluate the policy bargains (1986), in a static analysis with an empirical model, contain errors. We discuss each of these points in turn. show that the mix of shocks can certainly affect the gains from coordination. Currie and others (1987) find the Sustainability benefits of coordination to be small in the face of tem- The first issue is whether policymakers will stick to their porary disturbances, but those benefits rise steeply as the part of the bargain or cheat by redesigning their future persistence of the disturbance increases. policies once they have got the other participants locked These results, and the fact that it is not obvious why into position. There are two possibilities here. First, one noncoordinated policies should be disturbed less than party may find it advantageous to deviate from what was coordinated policies, suggest that coordination may ac- agreed once the other participants have taken up their tually increase policy robustness. Indeed, coordination agreed positions; that is, governments cheat each other. may reduce policy errors because decisionmakers try to The second is the usual time-inconsistency or credibil- share the risks rather than offload them onto their rivals. ity problem, discussed in the literature, where govern- Performance indicators are generally taken to be aggre- 42 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION gations of national targets that are affected in different subscribe to any of the 10 models, and the true state of (and often conflicting) ways by externals shocks, while the world can in turn be any of the 10 models. Coordi- coordinated policies follow by aggregating those na- nation yields benefits in just 60 percent of the combina- tional indicators into some global performance index. tions; in nearly 40 percent of the combinations, coopera- Coordinated policies will therefore normally be more tion makes at least one party to the bargain feel worse off. robust than noncooperative policies, which lack that However, this rather pessimistic view of the benefits to be extra aggregation stage and, hence, the added ability to expected from international policy coordination depends diversify the risks. on the assumption that policymakers take no account of Brandsma and others (1987) look at these robustness the different views about which model may be correct arguments explicitly using an example of United States- when computing the outcomes to be expected under each European Union coordination for the 1974-78 reces- bargain. If policymakers disagree about the choice of sion, subject to a range of "favorable" and "unfavorable" model, they are unlikely to agree to coordinate their poli- shocks to each of the target variables. They found that cies. However, Holtham and Hughes Hallett (1992) have while coordination almost always produced better out- shown that ruling out "weak" bargains in which either comes, it was not always the more robust strategy unless one party or the other expects their rival to be made worse policymakers learn to revise their policies as they dis- off by the bargain greatly improves the success rate of co- cover more about the disturbances entering their system. ordination when it occurs. They argue that risk averse But policymakers would almost certainly wish to revise governments would wish to rule out weak bargains be- their calculations in the light of past shocks and errors. cause they will be liable to reneging and would jeopardize In that case, coordination quite clearly yielded the more future coordination attempts. In other words, each gov- robust policies, and, more importantly, the degree of ernment would need to expect gains for all players ex extra robustness then increased with the size of shock. post, as well as ex ante, before entering such a bargain. Ghosh and Masson (1988a, 1988b) similarly show that Liberalization vs. Social Protection the presence of model uncertainty may raise the expected Disagreements about whether the "Washington consen- benefits of coordination if governments design their poli- sus" for a rapid liberalization of markets, but fiscal aus- cies with explicit regard to the presence of model uncer- terity and strict monetary discipline in policymaking, tainty. Governments aware of the adverse consequences would be more appropriate than an approach focusing of model disagreement would certainly take those conse- on social and economic cohesion, financial stability, and quences into account when choosing their policies, mod- arrangements to overcome market imperfections, have els, and degree of coordination. That would greatly re- done a lot to damage our ability to deal with financial duce the force of this objection to policy coordination. and economic crises. In truth, it is very hard to provide a coordinated response when there is disagreement Summary about the underlying paradigm to be followed. The recent financial and economic crisis in Asia, and the A number of papers have been written on this theme. financial difficulties in Russia and Latin America, have Pieper and Taylor (1998) and Stiglitz (1998) are two re- shown how globalization without supporting local poli- cent examples. They emphasize that the problem is not cies can bring trouble. Can local responses help "insure" so much misguided policies, but that these policies have the world against the kind of financial and trade shocks been recommended on the basis of the wrong frame- we have seen in the 1990s-or at least prevent the prop- work for understanding how emerging economies work. agation of them to the smaller and less financially devel- That means the problem is market and information fail- oped economies? ures, leading to misguided market reactions and a lack The answer seems to be "yes," but only up to a point of coordination internally or between different sectors and only if that coordination is constructed right. By of the economy. But can coordination really work when that we mean the following: policymakers disagree on the model? Frankel and Rockett (1988) examine the consequences * The forces of globalization will make greater regional of the assumption that governments differ about their or international coordination more necessary. But for view of the world. Using 10 international models in a that coordination to be successful, it must be accom- model comparison exercise, Frankel and Rockett exam- panied by explicit "localization" at the same time, that ine all 1,000 possible combinations where the two gov- is, by the more effective (or proactive and flexible) use ernments (the United States and the rest of the world) can of domestic economy instruments to create the con- PREREQUISITES FOR SUCCESSFUL INTERNATIONAL MONETARY POLICY COORDINATION 43 ditions in which international coordination efforts Consequently, the information exchanges they require can work effectively. will continue to be the most valuable element. * Under localization we have highlighted internal coor- dination (involving policy mix and market flexibility) Bibliography and "vertical" coordination. Thus wider international Beetsma, R., and L. Bovenberg. 1997. "Central Bank Indepen- coordination needs to be matched by "deepening," dence and Public Debt Policy." Journal ofEconomic Dynam- meaning a more consistent use of local policies. ics and Control 21: 873-94. * Another way of looking at it is globalization and the Beetsma, R., and H. Uhlig. 1999. "An Analysis of the Stability harmonization of policies in response to globalization and Growth Pact." Economic Journal 109: 546-71. need greater relative coordination too. Financial arrange- Brandsma, A., and A. Hughes Hallett. 1984. "Economic Con- flict and the Solution of Dynamic Games." European Eco- ments and monetary and fiscal discipline belong to the nomic Review 26: 13-32. absolute coordination category, as does the fact that re- Brandsma, Andries S., Andrew Hughes Hallett, and Joop Swank. covery in one country is much easier if others are grow- 1987. "The Robustness of Economic Policy Selections and ing than when they are not. But policy mix and ex- the Incentive to Cooperate." Journal of Economic Dynamics change ratc managemcnt belong to the latter category. and Control 11: 163-70. chaingernateimanagemenstiteongo thae lattery categor. Bryant, Ralph. 1987. "Intergovernmental Coordination of Eco- The international institutions have largely concen- nomic Policies: An Interim Stocktaking." In International trated on the former aspect, without being able to elicit Monetary Cooperation Essays in Honor of Henry C. Wallich, much response on the latter. The consequence has been Princeton Essays in International Finance. No. 169. Princeton, that the outcomes have been controversial and consid- NJ: Princeton University Press. erably less successful than they might have been. It Canzoneri, Matthew, and Patrick Minford. 1986. "When Policy institutional arrange- Coordination Matters: An Empirical Analysis." Centre for Economic Policy Research Discussion Paper No. 119. Lon- ments made the associated governments more like part- don: Centre for Economic Policy Research. ners in the coordination process and less like the recip- Carraro, Carlo, and Francesco Giavazzi. 1988. "Can Interna- ients of an imposed solution. tional Policy Coordination Really be Counterproductive?" * In any case, coordinated policies (or policy rules) must Centre for Economic Policy Research Discussion Paper No. be incentive compatible; the distribution of the gains 258. London: Centre for Economic Policy Research. crucal.Veryoftn th wekerpartes tandto ain Cooper, Richard N. 1969. "Macroeconomic Policy Adjustment is crucial. Very often the weaker parties stand to gain in Interdependent Economies." Quarterly Journal ofEconom- most over the best alternative, which makes it harder ics 83: 1-24. to persuade those who have to carry the burden of Currie, David A., and David Vines. 1988. Macroeconomic Inter- adjustment not to impose their own solutions-un- actions between North and South. Cambridge: Cambridge less, or until, a crisis breaks out. Then institutional University Press. arrangements have to be found which balance the rel- Currie, D. A., and S. Wren-Lewis. 1989. "An Appraisal of Al- ternative Blueprints for International Policy Coordination." ative coordination gains against the absolute gains or European Economic Review 33: 1769-86. losses. Currie, David A., Paul Levine, and Nic Vidalis. 1987. "Cooper- * In normal G-8 type circumstances, coordination shows ative and Noncooperative Rules for Monetary and Fiscal Pol- small but worthwhile gains. Outside that, coordina- icy in an Empirical Two-Bloc Model." In Ralph Bryant and tion, whether formal or informal and limited to ex- Richard Portes, eds., Global Macroeconomics: Policy Conflict and Cooperation. London: Macmillan. change rate targeting, has the great advantage of lim- Demertzis, M., A. Hughes Hallett, and N. Viegi. 1999a. "Can iting the damage caused by governments following the ECB be Truly Independent? Should It?" Empirica (Spe- silly policies-whether they are too much in their cial Issue for the Austrian National Bank, November) 26: own private interests, or simply misguided. This sug- 217-40. gests that even simple forms of coordination can limit Demertzis, M., A. Hughes Hallett, and N. Viegi. 1999b. "An In- the effect of large financial disturbances and inappro- dependent Central Bank Faced with Elected Governments." Centre for Economic Policy Research Discussion Paper No. priate responses to them. It's a disciplining device. 2219. London: Centre for Economic Policy Research. * Empirical studies have shown that coordination lim- Eichengreen B., and A. K. Rose. 1998. "Staying Afloat When the its the losses from unexpected shocks: The more so Wind Shifts: External Factors and Emerging Market Bank- the larger the shocks. Similarly it can help limit the ing Crises." Working Paper 6370. National Bureau of Eco- uncertainty and controversies caused by disagree- nomic Research, Cambridge, Mass. ments over teappitFrankel, J., and Katherine E. Rockett. 1988. "International ments over the appropriate policy framework. But if Mareooi'oiyCodiainWe oiyMkr . . . ~~~~~~~~Macroeconomic Policy Coordination When Policy Makers coordinated policies are to give the benefits of greater Do Not Agree on the True Model." American Economic Re- stability, it is important that they are well understood. view 78: 318-40. 44 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Ghosh, Atish R., and Paul R. Masson. 1988a. "International Pol- Hughes Hallett, A., Gerald Holtham, and Gary Hutson. 1989. icy Coordination in a World with Model Uncertainty." In- "Exchange Rate Targeting as a Surrogate for International ternational Monetary Fund, Staff Papers, Vol. 35. Washing- Policy Coordination." In Barry Eichengreen, Marcus Miller, ton, D.C.: International Monetary Fund. and Richard Portes, eds., Blueprints for Exchange Rate Man- Ghosh, A., and P. Masson. 1988b. "Model Uncertainty, Learn- agement. London and New York: Academic Press. ing and the Gains from Coordination." International Mone- Minford, Patrick, and Matthew Canzoneri. 1987. "Policy Inter- tary Fund Working Paper. Washington, D.C.: International dependence: Does Strategic Behaviour Pay?" Centre for Eco- Monetary Fund. nomic Policy Research Discussion Paper No. 201. London: Helkie, William, A. Hughes Hallett, and J. Marquez. 1989. "Pro- Centre for Economic Policy Research. tectionism and the U.S. Trade Deficit: An Empirical Analy- Oudiz, Gilles, and Jeffrey Sachs. 1984. "Macroeconomic Policy sis." Centre for Economic Policy Research Discussion Paper Coordination Among the Industrial Economies." Brookings No. 286. London: Centre for Economic Policy Research. Papers on Economic Activity 1. Washington, D.C.: The Brook- Holtham, G.H., and A. Hughes Hallett. 1992. "International ings Institution. Policy Coordination when Policy Markets do not Agree on Patrick, J.D. 1972. "Establishing Convergent Decentralised Pol- the True Model." American Economic Review 82: 1043-5 1. icy Assignments." Journal of International Economics 3: 37-52. Hughes Hallett, A. 1986a. "Autonomy and the Choice of Policy Pieper, U., and L. Taylor. 1998. "The Revival of the Liberal in Asymmetrically Dependent Economies." Oxford Economic Creed: The IMF, World Bank and Inequality in a Globalised Papers 38: 516-44. Economy." Centre for Economic Policy Research Working Hughes Hallett, A. 1986b. "International Policy Design and the Paper No. 4. New York: New School for Social Research, re- Sustainability of Policy Bargains." Journal of Economic Dy- vised January 1998. namics and Control 1O. Rogoff, Kenneth. 1985. "Can International Monetary Policy Hughes Hallett, A. 1987a. "The Impact of Interdependence on Coordination be Counterproductive?" Journal of Interna- Economic Policy Design: The Case of the U.S., E.E.C. and tional Economics 18: 1169-89. Japan." Economic Modelling 4: 377-96. Stiglitz, J. 1998. "More Instruments and Broader Goals: Mov- Hughes Hallett, A. 1987b. "How Robust Are the Gains to Pol- ing Toward the Post-Washington Consensus." Wider Annual icy Coordination to Variations in the Model and Objectives?" Lectures No. 2. United Nations University, Helsinki. Richerche Economiche (Special Issue on Game Theory and Woodford, M. 1995. "Price Level Determinancy with Control Economics) 41: 341-72. of a Monetary Aggregate." Carnegie-Rochester Conference Hughes Hallett, A. 1988. "Commodities, Debt and North- Series, 43. South Cooperation: A Cautionary Tale from the Structural- Wren-Lewis, S., P. Westaway, S. Soteri, and R. Barrel. 1991. ist Camp." In David A. Currie and David Vines, Macroeco- "Evaluating the UK's Choice of Entry Rate into the ERM." nomic Interactions between North and South. Cambridge: Manchester SchoolS 9: 1-22. Cambridge University Press. Hughes Hallett, A. 1992a. "Exchange Rates, Protectionism and Notes Commercial Policy: Alternative Strategies for Coordinating the G3 Economies." Japan and the WorldEconomy4: 215-37. 1. Woodford's (1995) "fiscal theory of the price level" shows Hughes Hallett, A. 1992b. "Target Zones and International Pol- how the experience of, or prospect of, an undisciplined fiscal pol- icy Coordination: The Contrast between the Necessary and icy can end up by undermining a carefully designed and well- Sufficient Conditions for Success." European Economic Re- conducted monetary policy and determining the effective price view 36: 893-914. level (inflation rate). This implies a need for interinstitution Hughes Hallett, A. 1993. "Exchange Rates and Asymmetric Pol- (between central banks and governments) as well as intercountry icy Regimes." Ox,ford Economic Papers 45: 191-206. coordination. Hughes Hallett, A. 1994a. "The Impact of EC-92 on Develop- 2. The distinction here is between making decisions jointly ing Countries Trade." In The World Bank Research Observer over several time periods, taking into account the predictable dy- 9: 121-46. namic responses of the policy targets, and making the same deci- Hughes Hallett A. 1994b. "The Imperfect Substitutability of sions as a sequence of individual one period decisions, in which Policy Regimes: Exchange Rate Targeting vs. Policy Coordi- case the predictable dynamic consequences of those decisions nation." In Economic Letters 44: 159-164. cannot be incorporated into each calculation. Hughes Hallett, A. 1998. "What Makes Exchange Rate Target 3. See also Hughes Hallett (1993) for more detail. Zones Work?" Open Economies Review 9: 115-39. 4. It is hard to calculate precise equilibrium exchange rate val- Hughes Hallett, A., and C.A. Primo Braga. 1994. "The New Re- ues in practice. They must depend on what you take to be "nor- gionalism and the Threat of Protectionism." Japan and the mal" capital flows, on what you define to be an equilibrium and International Economies 8: 388-421. on what model you take to calculate it, on what "fundamentals" Hughes Hallett, A. and Y. Ma. 1996a. "Changing Partners: The are assumed for productivity trends or outside conditions (for Importance of Coordinating Fiscal and Monetary Policies example, U.S. monetary policy, Japanese growth, or the Euro ex- within a Monetary Union." The Manchester School,64: 115-34. change rate), and on what fiscal and monetary policies may be Hughes Hallett, A. and Y. Ma. 1996b. "Transatlantic Policy Co- assumed for the domestic economy in the future. ordination with Sticky Labour Markets: The Reality of the 5. The only time this doesn't apply is when the stronger coun- Real Side." In M. Canzoneri, W. Ethier, and V. Grilli, eds., tries cannot choose their policies sensibly, even for themselves, The New Transatlantic Economy. Cambridge: Cambridge Uni- or when they fail to take into account large, unexpected shocks versity Press. which render those policies ineffective ex post. chapter The Impacts of Emigration on Countries of Origin Stephen Castles Institute of Social Change and Critical Inquiry University of Wollongong Vir u irtually all the world's 200-plus coun- country reaps the benefits of this investment, and the tries experience some international migrant often returns home to again become the re- migration, both arrivals and depar- sponsibility of the country of origin. tures. We speak of a "country of em- A key question for governments of emigration coun- igration" when departures of citizens tries is whether emigration contributes to development substantially exceed arrivals of nonci- or hinders it through "brain drain" and creation of sec- tizens over some years. Emigration can refer to perma- toral imbalances. The question is highly complex as em- nent departure to settle in another country or to tempo- igration exerts far-reaching and long-lasting impacts on rary migration, usually for employment. The notion of every sector of society. But overall, emigration is worth- a country of emigration also implies that the departure while for the home country only if it could not other- of many citizens has important economic-as well as so- wise employ migrants' labor power and skills, it can use cial, cultural, and political-implications.1 the human capital migrants obtain abroad upon their For many countries, emigration has become econom- return, and the transfer of income from emigration out- ically vital and a widely accepted economic strategy. weighs the costs of raising the migrant. This often leads to a "culture of migration" in which International cooperation can play a major role in en- going abroad to work becomes a rite of passage for hancing the contribution of migration to development. young men and women, conferring not only economic However, efforts and outcomes so far have been disap- well-being but also enhanced social status. Indeed, the pointing. Emigration and immigration countries often majority of emigrants are young adults at their most lack policies for maximizing the benefits of migration, productive age. In earlier times, most migrant workers exacerbated by the reluctance of immigration countries were males. But the last few decades have seen the "fem- to increase the cost of migrant labor to employers. inization of migration": Not only is a growing propor- The Extent and Causes of Emigration tion of primary migrants women, but some major mi- gratory movements today are female-dominated. Emigration statistics are incomplete and unreliable, even Migration has also become a major industry, encom- in highly developed countries and especially in develop- passing an estimated US$7 billion for air fares, interest ing countries (Boyle et al. 1998; Hoffmann and Law- on loans to migrants, and service fees to recruiters, med- rence 1996). This arises from the reluctance of govern- ical clinics giving health tests, and training centers ments to monitor departure of their citizens, the high (Abella 1995b). Emigration involves a transfer of the costs of monitoring, a lack of institutional capacity, the most valuable economic resource-human capital- difficulty of distinguishing emigrants from people leav- usually from a poor country to a rich one. The family, ing for other purposes, high levels of irregular migra- local community, and state of origin bear the costs of tion, and a lack of international uniformity in collect- raising a migrant to young adulthood. The immigration ing data. 45 46 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Estimates of annual departures or of the number of Emigration is therefore part of the social transformation nationals abroad may therefore vary widely. The Inter- linked to modernization. national Labour Office (ILO), the International Or- Migratory movements generally arise from prior links ganization for Migration (IOM), the United Nations between sending and receiving countries stemming Population Fund (UNFPA), and the Organisation for from colonization, political influence, trade, invest- Economic Co-operation and Development (OECD), as ment, or cultural ties (Kritz et al. 1992). Migrations well as regional bodies such as the European Union and from India, Pakistan, and Bangladesh to Britain were the Asia Pacific Economic Cooperation Forum are try- linked to the British colonial presence on the Indian ing to improve the reliability and comparability of the subcontinent. Caribbean migrants have similarly tended data, but international figures are only as good as the to move to their former colonial power-for example, national sources on which they are based. Despite these from Jamaica to Britain, Martinique to France, and gaps, a number of important comparative studies of em- Surinam to the Netherlands. igration dynamics have recently been done.2 However, Research on Asian migration has shown that families the level of knowledge is far lower than with regard to rather than individuals make migration decisions (Hugo immigration countries, and substantial need for docu- 1994). In situations of rapid change, a family may de- mentation and analysis remains. cide to send one or more members to work in another The most obvious cause of migration is disparities region or country to maximize income and survival. in income, employment, and social well-being among Such decisions are often made by elders (especially different areas. Demographic differences in fertility, men), with younger people and women expected to mortality, age structure, and labor-force growth are also obey patriarchal authority. Families may decide to send important (Hugo 1998).3 According to neoclassical eco- young women to the city or overseas because the labor nomic theory, one would expect individuals to maxi- of the young men is less dispensable on the farm and be- mize their income by moving from low-wage to high- cause young women are often seen as more reliable in wage economies (Borjas 1989). However, the causes and sending remittances. Such motivations parallel growing patterns of international migration are much more com- international demand for female labor in precision as- plex, owing partly to the family and community nature sembly or as domestic scrvants. of decisionmaking, and partly to state barriers to border Migratory chains typically stem from an external fac- crossing (Zolberg 1989; Portes and Boroicz 1989). Re- tor such as recruitment or military service, or initial cent research has shown the enormous influence of so- movement of young pioneers. Ensuing migrants mainly cial networks in both sending and receiving countries in follow "beaten paths" (Stahl 1993), helped by relatives encouraging and organizing migration. and friends who provide shelter, work, assistance in cop- Thus there is no simple relationship between poverty ing with bureaucratic procedures, and personal support. and emigration. Departures from the very poorest areas Migratory movements thus become self-sustaining: may be rare because people lack the economic capital Both migrants and nonmigrants become facilitators, needed to travel, the cultural capital to become aware of and a "migration industry" emerges consisting of re- opportunities elsewhere, and the social capital (or net- cruitment organizations, lawyers, agents, smugglers, and works) to find work and cope in a new environment. other middle-people. The emergence of a migration in- However, a catastrophe such as a war or environmental dustry with a strong interest in the continuation of mi- disaster may force even the poorest to migrate, usually gration has often confounded government efforts to under very bad conditions. control or stop such movements. Emigration is both a result and a cause of develop- The links between migrant community and area of ment. Development leads to migration because eco- origin may persist over generations. Remittances fall off nomic and educational improvements enable people to and visits home may decline, but familial and cultural seek better opportunities elsewhere. Research shows that links remain, and migrants may seek marriage partners middle-income groups in developing areas are most in the home country. Migration continues and may in- likely to depart. As incomes rise further, emigration then crease dramatically in time of crisis, as shown by the tends to decline (Rowlands 1998; United Nations 1998). mass movement of Yugoslav refugees to Germany dur- Migration also leads to development when remittances ing the early 1990s, where they joined compatriots who contribute to investment in the area of origin and when had migrated 20 years earlier. In the long run, migra- migrants return with skills, experience, and attitudes tions may spawn international networks that affect eco- conducive to economic, social, and cultural change. nomic relations, social and political institutions, and the THE IMPACTS OF EMIGRATION ON COUNTRIES OF ORIGIN 47 Box 5.1 The Benefits and Costs of Migration Benefits Costs Potential microlevel (individual and familial) benefits: Potential microlevel problems or costs: * Escaping from unemployment or underemployment * The loss of workers may reduce productivity and income at home * Escaping from repressive situations or threats to life and health a The high costs of migration (especially in irregular movements) * Improving individual or family income may put an individual or family into debt * Gaining capital to invest in land or in business * The risk of illness or injury (especially in irregular movements) * Gaining skills, work experience, and education may involve high costs * Improving family security through diversification of income * Remittances may be used for nonproductive purposes * Improving social status * Remittances may exacerbate existing social inequalities or cause local inflation Potential macrolevel (society-wide) benefits: * Migrants may fail to gain useful skills abroad * Reducing demographic pressures where population is outstrip- * Certain types of migration (especially trafficking of women and ping resources children for the sex industry) may lead to individual or family shame * Reducing unemployment where growth in the labor force growth exceeds demand Potential macrolevel problems or costs: * Reducing political pressures where economic and social prob- * Emigration is unlikely to reduce demographic pressures but may lems are causing discontent lead to shortages of young adults in certain regions * Using migrant remittances to improve a country's balance of pay- * Emigration of skilled workers may cause economic bottlenecks ments and fund imports in certain industries 3 Improving human capital through transfer of skills, work experi- O The import of ideas by returning migrants may increase political ence, and education pressures for change D Contributing to international trade u Remittances may occur through irregular channels : Introducing innovative attitudes toward economic development * Improvements in human capital may be minimal and outweighed through exposure to more developed societies by loss of highly skilled personnel U Loss of control of migratory movements may encourage corrup- tion and organized crime a The need for people to go abroad to seek a decent livelihood may be seen as a "national shame" culture and national identity of all countries concerned Chinese ethnic background) seek work elsewhere. But (Basch et al. 1994). Malaysia also attracts large numbers of low-skilled work- ers from Indonesia, the Philippines, and Bangladesh as The Migration Transition well as expatriate experts from developed countries. Much as the "industrial transition" refers to the shift The notion of a migration transition is helpful in eval- from agriculture to manufacturing to services and the uating the long-term results of emigration.. If the bene- "demographic transition" entails falls in both mortality fits help support the economic and demographic transi- and fertility, a "mobility or migration transition" links tions and reduce the need for future emigration, then the the preceding transitions with population movements process can be seen as conducive to development. How- (Zelinsky 1971). Emigration frequently rises at the be- ever, if emigration perpetuates economic and demo- ginning of the industrialization process owing to popu- graphic imbalances and reduces pressure for change on lation growth, a decline in rural employment, and low governments, it can hinder development. wages. This was the case in early-19th-century Britain, late-19th-century Japan, and Korea in the 1970s. As in- Weighing the Benefits and Costs of Emigration dustrialization proceeds, the labor supply declines and The potential benefits and problems arising from emi- wage levels rise. Emigration then falls, and immigration gration include microlevel effects on individual mi- occurs and then predominates (Martin et al. 1996; Skel- grants, families, and communities, and macrolevel ef- don 1997; Amjad 1996). fects on economies and societies (see Box 5.1). Most Countries often experience simultaneous emigration individuals that emigrate and governments that encour- and immigration during the transition, usually of work- age emigration obviously conclude that the benefits out- ers with different types of skills. For instance, some weigh the costs. However, microlevel and macrolevel Malaysian workers migrate to Taiwan and other coun- cost-benefit calculations may conflict. Governments tries while some highly skilled personnel (especially of often encourage emigration to reduce unemployment, 48 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION gain remittances, and transfer skills, but these expecta- and may take extreme forms such as the trafficking of tions are not always realized. In some cases, the calcula- women and children for prostitution. tion is based on short-term factors while the costs to the economy and society are long-term.4 Many governments facilitate the movement of work- ers abroad to provide employment, obtain remittances, Remittances from emigrants working abroad can consti- and mollify social discontent. Other governments are tute a major contribution to the balance of payments of skeptical of such benefits and do their best to restrict countries with severe trade deficits. For instance, Pak- emigration, with varying degrees of success. Some gov- istani workers remitted over US$2 billion in 1988 (not ernments doubt the merits of emigration but find that including unofficial transfers or consumer goods), which they cannot prevent it and therefore try to regulate covered 30 percent of the cost of imports (Abella 1995a). movements and protect their citizens abroad. Indeed, Indian workers remitted US$2.6 billion in 1988-the strict limits on emigration may be counterproductive if equivalent of 15 percent of imports (ILO 1991). Most a state lacks the capacity to enforce them: They can sim- emigration countries encourage workers to send remit- ply lead to irregular movements, which hurt both mi- tances through official channels by setting up banking grants and the sending country. facilities in countries of employment, providing special Abella (1995b) found that most Asian countries exchange rates, or by other inducements. Failure to pro- guarantee the right of their citizens to move freely out vide accessible financial services and realistic exchange of their country and back in again, but that they may rates often leads emigrants to repatriate funds through use high fees, complex procedures, or other barriers to irregular channels, which may reduce their benefits to deter such movements or select who pursues them. For the home economy. example, Pakistan, Bangladesh, India, Indonesia, and Most emigration countries maintain policies to pre- the Philippines set a minimum age for women seeking vent the abuse or exploitation of their citizens abroad employment abroad. Cultural factors can also come into and provide assistance in case of illness, accident, death, play: Women from Pakistan and Bangladesh rarely mi- legal trouble, disputes with employers, or other emer- grate for employment, while Sri Lanka, Indonesia, and gencies. Emigration countries typically station labor the Philippines encourage female labor migration. attaches and welfare workers at consulates and may Most Asian governments have set up special depart- also fund welfare centers or support nongovernmental ments or agencies to manage labor emigration, such as organizations (NGOs) that run them. Some also pro- Bangladesh's Bureau of Manpower, Employment and vide advice and counseling services before departure- Training (BMET) and India's Office of the Protector of sometimes in cooperation with immigration countries. Emigrants within the Ministry of Labour. The govern- Migrant workers must often cover the costs of such ments of China and Vietnam organize labor migration services, usually as part of the fee to obtain travel through state-owned firms that contract with foreign documentation. states or companies, and Bangladesh, Indonesia, and Measures and services for protecting emigrants are fre- Thailand have also negotiated directly with foreign gov- quently ineffective. They are not available to the large ernments. India, meanwhile, licenses private recruit- numbers of irregular migrants, and even legal migrants ment firms to hire and deploy workers while seeking to are so dependent on finding and keeping employment regulate their fees and ensure that they treat workers abroad that they hesitate to complain about abuses. fairly (Abella 1995b). In South Asia, emigration bureaus Staffing and regional coverage abroad is often inadequate review the contracts of emigrants to ensure that their to meet the need, and sending authorities hesitate to an- terms are satisfactory (Shah 1994). tagonize authorities and employers in immigration coun- However, governments find it difficult to control re- tries. In a situation of global oversupply of low-skilled cruitment firms, and cases of abuse and exploitation are migrants, market power lies with recruiting countries. widespread. For example, agents in Pakistan charge four times the permitted fee or even more (Shah 1994). In fact, the number of irregular migrants within Asia may Large numbers of migrants return to their countries of exceed the number of regular ones (Lim and Oishi origin after varying stays abroad. Some benefit from 1996), and the same is certainly true in Africa, Latin reintegration assistance from either the receiving coun- America, and even in Eastern and Southern Europe. Ir- try or the country of origin. However, most are left to regular migration gives rise to exploitative employment their own devices, and they frequently face difficulty THE IMPACTS OF EMIGRATION ON COUNTRIES OF ORIGIN 49 in finding employment commensurate with skills they lateral agreements regulated working conditions and so- have acquired abroad. Their potential input to their cial security, as many as 20 percent of employees mi- country is thus diminished, if not lost. Even though re- grated irregularly (Martin 1991). turn migration occurs in many countries on a large The Netherlands recruited Turks in much the same scale, authorities possess limited knowledge and concern way as Germany had, and migrants moved sponta- regarding these conditions. neously to France, Belgium, Austria, and other coun- The 1998 United Nations Technical Symposium on tries. The peak movement occurred from 1968 to 1973, International Migration and Development found that with 525,000 workers traveling through official chan- monitoring of return migration and research on its con- nels, and many more unofficially (Martin 1991). This sequences is insufficient to understand its contribution was guestworker migration: The workers were not meant to development. In some cases returnees appear to have to bring in dependents or stay for long periods. In the contributed little to sustainable development, while in event of an economic downturn, the system was de- others they appear to have had positive effects. However, signed to allow quick repatriation. Despite the restric- adequate counseling and information as well as help in tions, about one-quarter of Turkish emigrants were obtaining investment credits do seem to help returnees women, often traveling before their men to get jobs in reacclimate successfully and maximizes their positive ef- factories looking for light assembly workers. Once in fects on development. Returnees are more likely to be Germany or the Netherlands, they would try to persuade successful if advised and supported by governmental employers to recruit their husbands by name. This facil- agencies and NGOs, and social networks in the home itated family reunion, despite official policies (Castles country are also crucial. Cooperation between sending and Kosack 1973 and 1985; Castles et al. 1984). By and receiving countries and with international organi- 1972, half of Turkish workers had their families with zations and NGOs can play an important role, espe- them in Germany (Martin 1991). cially when significant numbers of emigrants return over The German government stopped recruiting in 1973 a short time (United Nations 1998). in response to the oil crisis, hoping that surplus workers ywould leave. In fact, many Turks stayed, and family Case Study: Turkey reunion continued. By 1974 the country housed over Turkey was a major source of labor for Western Europe's 1 million Turkish residents among a total foreign popu- postwar boom until the 1970s. Rather than undergoing lation of 4.1 million. Their number grew to 1.6 million a "migration transition," the country has since devel- by 1982 and to over 2 million by 1995. Political unrest oped new types of emigration. leading to the 1980 Turkish military coup generated Unlike the Southern European countries that sent waves of asylum seekers, who found shelter in Turkish workers to Western Europe after 1945, Turkey had no communities in Germany. Clearly, the chain migration tradition of international labor migration. Agreements process was more powerful than German policies. Mass between the German Federal Government and the Turk- deportation, though debated, was never a real option ish government in 1961 and 1964 initiated organized for a democratic state. By pursuing family reunion and movement. The latter hoped to relieve domestic unem- permanent settlement, the Turks lost their attractive- ployment and obtain foreign exchange, while the mi- ness as a flexible labor force: They became largely per- grants themselves sought to escape poverty, unemploy- manent settlers who needed housing, schools, and so- ment, and dependence on semifeudal landowners. The cial amenities. country also expected that money and skills gained Turkish expectations of the benefits of labor migra- abroad would encourage economic development and tion were also unfulfilled. Many of the migrants selected hoped to gain access to the European Community.5 by the German recruitment offices were not unem- Migration to Germany was highly organized. The ployed "surplus population" but skilled workers. Since German Federal Labour Office (Bundesanstalt fur Ar- they were generally given unskilled jobs in Germany, beit-BfA) set up recruitment offices in Turkey in col- they gained few qualifications relevant to Turkish indus- laboration with authorities. Employers requiring both trialization. Worker remittances, which ran at US$1.5- unskilled and skilled manual employees paid a fee to the 2 billion per year in the 1980s, certainly helped the BfA, which selected workers, tested their skills, provided Turkish balance of payments, but they were mainly used medical exams, and screened police records. The work- for consumption or to establish small tertiary businesses ers were brought in groups to Germany, where employ- such as taxis, cafes, and shops rather than for productive ers had to provide initial accommodation. Although bi- investment (Martin 1991). When Turkey applied for 50 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION full EC membership in 1987, there were strong reac- provide a solution unless it is part of an integrated de- tions in Germany because of fears of an uncontrolled velopment strategy. flood of workers. EC membership was refused in 1989, Case Study: The Philippines although negotiations reopened in the mid-1990s. Thus Turkish migration to Western Europe did little Just as the Mediterranean periphery fueled Western Eu- to promote economic development. Since rapid demo- ropean industrial expansion until the 1970s, China, graphic growth continued and the labor market was un- South Asian countries, and the Philippines and Indone- able to absorb all Turkish young people, pressure for em- sia have more recently become major labor providers. igration continued. At the beginning of the 1990s some The Philippines is the contemporary labor exporter par 20-40 percent of young adults professed interest in excellence (rather like Italy a generation ago), with over migrating to EC countries (Martin 1991). Because this 4 million of its people scattered across the world (Bat- possibility had largely closed, new emigration flows de- tistella and Paganoni 1992). About half are permanent veloped to the Gulf oil states, especially Saudi Arabia, settlers, residing mainly in the United States, while the as well as to Russia and other Soviet successor states. rest are overseas contract workers (OCWs)-both legal Many of these new migrants were skilled workers. By and illegal-residing in the Gulf states and Asia, includ- 1996, 3.45 million Turks were living abroad-the equiv- ing Japan, Malaysia, Singapore, and Hong Kong. In alent of 5.4 percent of Turkey's population. Yet Turkey's 1994 a total of 565,000 Filipinos were recruited to work labor market was still incapable of providing enough in other countries, and a further 154,000 took jobs as jobs: 1 million more people had applied for emigration. sailors on foreign-owned ships. Turkeys proportion of surplus labor is expected to reach Labor export is crucial to the economy of the Philip- 15 percent by 2000 (OECD 1998). pines: Unemployment levels might be 40 percent higher Why did Turkey's large-scale emigration program fail without it. Official remittances from migrants in 1994 to spark economic development? Beginning in the totaled US$2.94 billion, which financed 50 percent of 1970s, the German government and Turkish authorities the country's external trade deficit (Amjad 1996). Work- developed measures to help workers reintegrate, includ- ers remitted a total of US$23.4 billion between 1975 ing promoting new companies and establishing credit and 1995, with the largest source being the United funds. But a longitudinal study by the German Institute States (Go 1998). (These figures include only remit- for Employment Research (IAB) showed that only half tances through the formal banking system; transfers are of male respondents were still economically active in also made as cash or consumer goods.) By the early Turkey in 1988, with 90 percent self-employed. There 1990s, 16 percent of Filipino households were receiving was almost no relationship between occupations in remittances from abroad (Saith 1997).6 Turkey and in Germany: Most returnees were farmers Women have played a growing part in emigration or retailers or worked in transport occupations. The ma- from the Philippines-at least half of temporary mi- jority worked in small enterprises. Few had acquired grant workers are now female. Most migrant women qualifications in Germany, and those that had had rarely take jobs as domestic workers, entertainers (often a eu- been able to use them on return. Hardly any returnees phemism for prostitution), restaurant and hotel staff, had received counseling about return or investment, and and assembly-line workers in clothing and electronics. return seemed to have made little contribution to devel- These jobs are low in pay and status, offer poor work- opment, although there might have been exceptions at ing conditions, and are associated with patriarchal the regional level. Migrants did appear to have invested stereotypes such as docility and willingness to give per- considerably in educating their children, which might sonal service. Domestic service leads to isolation and have promoted development indirectly. The IAB's analy- vulnerability for young women, who often have little sis concludes that better measures to prepare returnees protection against the demands of their employers (Lim for economic reintegration and encouraging invest- and Oishi 1996). Female migrants from the Philippines, ments in specific regions and economic sectors are es- Thailand, and Sri Lanka also often travel as "mail-order" sential (Hbnekopp and Tayanc 1998). brides to Europe, Australia, and Japan (Cahill 1990). Apparently emigration itself does not spur economic According to Saith, both supply-side and demand-side development. However, this result must be seen in the factors underlie Philippine migration. The former in- context of a strained economy and a society deeply di- clude rapid growth of the population and the labor force vided along political, ethnic, and religious lines. Emi- amid a long period of economic stagnation, which has gration is an expression of such problems and will not led to low-quality employment and unsustainable forms THE IMPACTS OF EMIGRATION ON COUNTRIES OF ORIGIN 51 of agriculture. Such emigration pressures found an out- (Wong 1996). Public concern in the Philippines came let because of the high levels of education and occupa- to a head in 1995 when a Filipina maid, Flor Contem- tional training, familiarity with the English language, placion, was hanged in Singapore after being found and female participation among the Filipino labor force. guilty of murder. The case strained relations between the Saith's study found no sign of economic development in two countries and led to a heated debate concerning the the foreseeable future that could reduce emigration pres- 700,000 Filipinos who go to work overseas each year. sures. The study also found that migration gains accrue The Philippine government again banned migration of disproportionately to richer regions and higher income domestic workers to Singapore, but this ban, too, was groups, exacerbating existing imbalances (Saith 1997). largely circumvented. Overseas employment became an official program of However, in June 1995 the Philippines passed the the Philippine government in 1974. At first it attempted Migrant Workers and Overseas Filipinos Act to improve to follow the South Korean model by setting up an au- the welfare of OCWs. Protecting the dignity and human tonomous Overseas Employment and Development rights of Filipinos was to come before economic objec- Board with a near-monopoly on recruitment. However, tives. And rather than exporting Filipino labor, the gov- this approach was unsuccessful, and the government es- ernment adopted a policy of managing labor outflows tablished the Philippine Overseas Employment Admin- while developing a comprehensive employment strategy istration (POEA) to maintain "positive controls" over at home (Lim and Oishi 1996). Specific measures in- the private recruitment industry. The POEA has be- cluded a policy of selective deployment and measures to come a major agency, with numerous offices and a bud- improve information for prospective migrants as well as get for 1988-95 of 576 million pesos-some 7.6 per- cooperation between POEA, OWWA, the Technical cent of the budget of the Department of Labour and Skills and Development Authority, and the Department Employment (Tomas 1998). People who wish to seek of Foreign Affairs (Go 1998). work abroad must register with the POEA, and premi- It is difficult to assess the success of the new policy. gration training and certification requirements for en- Certainly dramatic cuts in emigration and substantial tertainers, domestic workers, and some other categories improvement in conditions of Filipino OCWs did not have been tightened to prevent abuse. seem to have occurred by 1997. Since then the effects The country also established the Overseas Workers' of the Asian crisis on labor migration make comparisons Welfare Administration (OWWA) to assist workers in difficult. However, it's important to note that return mi- emergencies and protect them from exploitation. Offi- gration and cuts in flows to receiving countries are likely cials at Filipino consulates in labor-importing countries to cause considerable hardship. Bohning (1998) points also help migrants, yet in 1993 only 31 labor attaches, out that women migrants originate disproportionately 20 welfare officers, and 20 coordinators were available from poorer regions of the country, and loss of their to respond to the needs of millions of migrant workers earnings can be disastrous for the families who depend in 120 countries (Lim and Oishi 1996). Philippine offi- on them. cials often find themselves powerless against unscrupu- Substantial emigration also affects Philippine na- lous agents and abusive employers who may have the tional identity. For emigration countries, admitting to backing of the police and other authorities in receiving permanent loss of substantial groups of emigrants can countries, especially the Persian Gulf be seen as a "national shame" because the nation is in- An important receiving country, especially for women, capable of providing an acceptable life for its people. is Singapore. Owing to attractive job opportunities for This has become a major theme in the Philippines, es- Singaporean women, employment of foreign domestic pecially regarding settlement in the United States. Per- servants is high: In 1993 the city-state housed some manent emigrants may be seen as members of a diaspora 81,000 foreign domestic servants, of whom 50,000 were who retain links with the homeland (even after becom- from the Philippines, 17,000 from Sri Lanka, and 10,000 ing U.S. citizens), but also as people who betray the na- from Indonesia. Fees for recruitment (mainly through tion by exporting their energy and skills (Aguilar Jr. specialized agencies) ranged from US$ 1,000 to US$2,000, 1996). usually deducted from the employee's wages. Frequent One way of dealing with this dilemma is to create a cases of abuse reported to Philippine authorities led to special status for expatriates. The Philippines established a ban on recruitment in 1987, but this was circum- the category of Balikbayans-"people coming back vented by the Singaporean authorities, who admitted home to the Philippines"-as early as the 1970s. Balik- Filipinas as tourists and then gave them work permits bayans may be overseas contract workers, permanent 52 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION U.S. residents, or even U.S. citizens of Filipino origin. Preparingfor migration. Authorities should use their Programs that facilitate their return include special consuls and labor attaches in destination countries to col- travel documents, tax privileges, and import conces- lect information on working and living conditions for mi- sions. Such official recognition of diasporas treats peo- grant workers and their families. Governments can then ple who live abroad as a part of the national community, employ counseling and other forms of information to yielding both economic and political benefits (Blanc help potential migrants make well-informed choices. Au- 1996). Indeed, a key focus of contemporary migration thorities should also provide counseling on repatriating studies is the emergence of international networks link- earnings and savings and investing them to improve eco- ing emigrants to their countries of origin. nomic well-being, health, education, and social welfare. Has large-scale emigration contributed to economic Controlling actual movement. Authorities should take development in the Philippines? At the microlevel, re- all possible steps to ensure orderly migration, including mittances have provided resources to improve agricul- organizing recruitment and deployment or regulating tural enterprises or establish other types of enterprise. At private agents. Bilateral or multilateral agreements, or the macrolevel, remittances have provided a substantial at least discussions, with receiving countries are impor- amount of foreign exchange. And returning Filipinos tant. Experience has shown that rigid regulations that have often brought new skills as well as experience in a fight market forces, such as prohibitions of cross-border wide range of economic activities. However, without ef- movement, are often counterproductive. Governments fective policies to maximize these benefits, resources have are more likely to achieve orderly movements by provid- often gone into unproductive or unsuccessful ventures. ing regulatory frameworks and services. The most common investment has been in small-scale Offering consular and social services abroad. It is vital enterprises in transport, catering, and other services, with for countries of origin to take responsibility for mi- skills gained abroad often wasted for lack of relevant op- grants. This requires provision of consular and social ser- portunities. What's more, by providing individual escape vices in the event of legal problems, disputes with em- from economic stagnation, emigration may well have re- ployers, or accident and illness. Emigration countries duced pressures for economic and social change. need to adequately fund special agencies; the Filipino In view of such problems, Go suggests major initia- experience provides a useful model. tives to achieve the benefits of migration, including ef- Transferring remittances. Orderly transfer of remit- forts to sustain Filipino economic growth, more and tances is essential for countries to realize development better-trained personnel and financial resources to pro- benefits; otherwise migrants will transfer their earnings tect the rights and welfare of Filipinos abroad, and more in the form of cash or consumer goods. Orderly trans- education and training to enable Filipinos to keep pace fer ensures direct benefits to national accounts as well as with changes in the local and global economy (Go safety for migrants. Realistic exchange rates, low bank 1998). These far-reaching suggestions clearly imply that charges, and reliable financial services in countries of efforts to manage emigration must be closely linked to both employment and origin are vital. Special invest- economic and social development. Implementing such ment schemes to encourage migrants to combine their policies requires political will as well as institutional ca- savings to establish companies or fund infrastructure pacity for effective action. projects in home communities are worthwhile. Compa- Improving the Conditions forReturnMigranies set up by Turkish migrants in the 1970s were not particularly successful, but this experience should be Despite the knowledge gaps, some policies clearly help evaluated to learn its lessons. maximize the gains from emigration and return migra- Providing education and training. Countries of origin tion. An effective strategy begins during the decision- should ensure that migrants receive education, voca- making period before departure: Governments, commu- tional training, and work experience relevant to occu- nities, and individuals need to plan to use the economic pational needs at home. This requires cooperation with and cultural capital they hope to gain through work governments and employers in the receiving country as abroad. There is considerable evidence that many mi- well as appropriate counseling for migrants. However, grants and their families do make such long-term calcu- migrants will be motivated to accumulate human capi- lations. Governments, in contrast, often concentrate tal only if they see a real chance of using it on return. mainly on short-term economic and political aspects and Such measures therefore must be closely linked to do- neglect long-term strategies. mestic economic and labor-market policies. Policies to maximize long-term gains from return mi- Maintaining links. Research has shown that the most gration might include the following elements: successful returnees maintain economic and community THE IMPACTS OF EMIGRATION ON COUNTRIES OF ORIGIN 53 links at home. Returnees who have lost local contacts (Go 1998). Japan set up the Japan International Train- and knowledge or members of the second generation ing Organization (JITCO) to work with government often have great difficulty relating to local customs and agencies, employers, unions, and international organi- participating in social networks. Governments can fos- zations on recruitment and training (Abella 1995b). But ter such networks by facilitating regular visits home and such training schemes sometimes merely undermine re- efficient and affordable telecommunications, and by strictions on the import of low-skilled workers, and the supporting social and cultural associations in receiving training is often limited. countries. Counseling and services should emphasize the International cooperation often occurs when the im- need for maintaining local business contacts while migration country wishes to encourage repatriation. Just abroad if migrants wish to establish businesses on return. as Germany tried to assist reintegration of Turkish Offering prereturn advice. Authorities from emigra- workers after 1973, France and migrant-sending coun- tion countries should provide comprehensive services to tries in North and West Africa made similar arrange- prepare migrants for return. Consular officials or social ments. The European Union has also helped facilitate workers employed by the government can provide such return migration and maximize its benefits. For exam- services, but social and cultural associations of migrants ple, in 1975 France and Senegal began encouraging the may prove more effective. Such services should include voluntary return of migrants, offering training to re- advice on social reintegration and family matters as well turnees and funding reintegration projects. However, as employment opportunities and investments at home. Senegal has apparently been more successful in ensur- Repatriating and investing savings and assets. Govern- ing transfers of migrants' savings than in utilizing their ments should encourage returning migrants to repatri- skills. One study suggests facilitating access to micro- ate their savings and assets by adopting policies such as funding and financing for small-scale enterprises-a realistic exchange rates and waiving of fees and customs "bank of the poor" based on models in Bolivia, Ban- duties. Migration agencies should also ensure that re- gladesh, and Tunisia (Diatta and Mbow 1998). turning migrants maximize the benefits of their eco- International organizations-notably the ILO and nomic and human capital to foster sustainable devel- the IOM-have played an important role in improving opment. One technique is to provide targeted credit cooperation between sending and receiving states, and mechanisms such as low-interest rates to returnees who regional organizations such as the European Union invest their savings in economically useful ways. (EU) and the Asia-Pacific Economic Council (APEC) Emphasizing nondiscrimination. Finally, a word of also seek to harmonize regulations regarding migrant warning: Services for migrants and returnees should be workers. The International Covenant on Civil and Po- designed and delivered so that they do not seem unfair litical Rights and the International Covenant on Eco- to nonmigrants. Credit mechanisms and similar mea- nomic, Social and Cultural Rights apply to migrants as sures should be built into wider strategies for economic much as to anyone else, but more specific measures are and social development for the whole population. needed. ILO conventions No. 97 of 1949 and No. 143 of 1975 do specify minimum standards of treatment for Fostering InternationaI lCooperation on lEmqigration migrant workers and their dependents, but as of 1993 Although international cooperation between sending only 40 and 17 countries, respectively, had ratified the and receiving countries could help maximize the bene- conventions (Stalker 1994). In 1990 the United Nations fits of emigration, immigration countries tend to view adopted a Convention on the Rights of Migrant Work- labor migration as a source of short-term economic ben- ers and Members of Their Families, but only a handful efit rather than as an aspect of cooperation on develop- of emigration countries and no significant immigration ment. Labor-importing countries also often see migrants countries have ratified it. as "cheap labor," using them to meet skill shortages that International organizations also offer technical assis- would otherwise lead to bottlenecks or wage inflation. tance. For example, from 1989 to 1991 the ILO, in col- Measures such as training to improve migration's con- laboration with the POEA and the Commission of the tribution to the country of origin would make it more European Communities, pursued a program called En- expensive, and therefore reduce employers' motivation. trepreneurship on Migrant Earnings in the Philippines. In any case, a large proportion of migrating labor moves The project identified training, advisory, and financial through irregular channels. services to help former migrants establish businesses. As an example of such problems, the Philippines has However, the success of would-be entrepreneurs de- concluded labor agreements with Jordan, Iraq, and Qatar pended on the local economic, political, and social cli- but not with the most important recruiting countries mate (ILO 1991). Another important scheme is the 54 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION IOM's Return of Talents Program, which encourages gration from both member states and third countries highly skilled emigrants to take temporary positions in should be essential to regional economic integration as their countries of origin to aid development. well as to international cooperation on development. Despite such efforts, relatively little bilateral and Cementing bilateral agreements. Emigration countries multilateral work is focusing on protecting migrants and frequently seek to make bilateral agreements with im- helping migration stimulate economic development in migration countries; such agreements did much to im- sending countries. The following steps may help: prove the social security and rights of migrant workers Ratifying and improving multilateral regulations. Major and their families in Western Europe in the 1 960s and elements of an international regulatory framework al- 1 970s. However, immigration countries have recently ready exist in ILO conventions and the 1990 United Na- proved reluctant to conclude such agreements. Interna- tions Convention on the Rights of Migrant Workers and tional efforts should encourage such agreements by un- Members of Their Families. Labor standards laid down derscoring their benefits for all parties. Migrants gain by the World Trade Organization also relate to the con- through greater social security-for instance, through ditions of migrant workers. These standards need to be transferability of pension entitlements gained while linked into a comprehensive framework to regulate the abroad. Emigration countries gain because such agree- rights and conditions of migrant workers. However, the ments facilitate transfer of remittances and make it main task is to persuade more countries to ratify these harder for agents and recruiters to cream off profits. Im- instruments and to take steps to implement them. A migration countries gain through improved control of world intergovernmental conference on migration could migration and the creation of a more stable and better- raise awareness of the need for such action. trained workforce. Establishing an effective international agencyfor mi- Cooperating on training and education. Such coopera- gration. An international agency, possibly within the tion should be an important part of multilateral, re- United Nations framework, could play a major role in gional, and bilateral agreements on migration, and a boosting the effectiveness of international regulation of major aspect of efforts to support international devel- migration and in developing multilateral approaches to opment. Such programs would increase the cost of mi- using migration to aid economic development. Such an grant labor somewhat, but short-term gains in produc- agency could be built on the existing International Or- tivity for host countries and long-term gains for ganization for Migration (IOM-an intergovernmental countries of origin would offset this cost. organization which has existed since the 1940s), but Linkingforeign direct investment to migration strate- would need to greatly enhance its responsibilities, fund- gies. One way to maximize gains from international mi- ing, and standing. gration is to link it to foreign direct investment. Foreign Improving monitoring and research. Most countries investment in a less-developed country can create op- lack effective systems for monitoring international mi- portunities for workers to migrate temporarily to the gration, and definitions, and statistical categories vary headquarters country to gain industrial experience and considerably. Effective monitoring and comprehensive, training. This process not only provides a source of tem- comparable statistics are essential for effective multilat- porary labor for the company but also produces well- eral action. An international agency for migration could trained workers for less-developed countries, while the assume such a task. Countries and analysts also know far workers gain through both training and enhanced earn- too little about the long-term economic and social con- ings. Many companies already follow such practices, but sequences of emigration. Greatly increased research at they need to expand and develop them systematically. the local, national, and international levels is essential. Expanding "return-of-talents" schemes. The IOM Implementing regional agreements. Regional economic has encouraged highly skilled emigrants to make their integration and political cooperation often includes a skills available for development projects in their coun- focus on labor migration. Some regional bodies have tries of origin. International institutions should develop highly developed policies on movements among mem- this approach-a way to reverse the effects of "brain ber countries-particularly the European Union, which drain"-on a much larger scale. has introduced almost totally free movement for citizens of member states. However, no regional body has intro- Bibliography duced comprehensive policies regarding migrants from Abadan-Unat, N. 1988. "The Socio-economic Aspects of Return third countries, although EU countries have begun to Migration to Turkey." Revue Europ&enne des Migrations In- harmonize national rules. Common policies on mi- ternationales (3): 29-59. THE IMPACTS OF EMIGRATION ON COUNTRIES OF ORIGIN 55 Abella, M. 1993. "Labor Mobility, Trade and Structural Change: national Migration and Development, 29 June-3 July 1998. The Philippine Experience." Asian and Pacific Migration The Hague, Netherlands: unpublished. Journal 2(3): 249-68. Go, S. 1998. "The Philippines: A Look into the Migration Sce- Abella, M. I. 1995a. "Asian Migrant and Contract Workers in nario in the Nineties." In Migration and Regional Economic the Middle East." In R. Cohen, ed., The Cambridge Survey of Integration in Asia. Paris: Organisation for Economic Co- World Migration. Cambridge: Cambridge University Press. operation and Development (OECD). Abella, M. I. 1995b. "Policies and Institutions for the Orderly Gunatilleke, G. 1991. Migration to the Arab World: Experience Movement of Labour Abroad." In M. I. Abella and K. J. Lionn- ofReturningMigrants. Tokyo: United Nations University. roth, eds., Orderly International Migration of Workers and Hoffmann, E. and S. Lawrence. 1996. "Statistics on Interna- Incentives to Stay: Options for Emigration Countries. Geneva: tional Labour Migration: A Review of Sources and Method- International Labour Office. ological Issues." Geneva: International Labour Office. Aguilar, F. V. Jr. 1996. "The Dialectics of Transnational Shame Honekopp, E. and T. Tayan,. 1998. "Releasing the Develop- and National Identity." Philippine SociologicalReview 44(1-4): ment Potential of Return Migration: The Case of Turkey." 101-36. Paper presented at United Nations Technical Symposium on Amjad, R. ed. 1989. To the Gulf and Back, Studies on the Eco- International Migration and Development, 29 June-3 July nomic Impact ofAsian Labour Migration. New Delhi: ILO/ 1998. The Hague, Netherlands: unpublished. ARTEP. Hugo, G. 1994. Migration and the Family. Vienna: United Na- Amjad, R. 1996. "Philippines and Indonesia: On the Way to a tions Occasional Papers Series for the International Year of Migration Transition." Asian and Pacific Migration Journal the Family, no. 12. 5(2-3). Hugo, G. 1998. "The Demographic Underpinnings of Current Appleyard, R. ed. 1995. International Migration: Special Issue on and Future International Migration in Asia." Asian and Pa- Emigration Dynamics in Developing Countries. Geneva: Inter- cificMigrationJournal7(1): 1-25. national Organization for Migration. ILO 1991. "Entrepreneurship on Migrant Earnings in the Appleyard, R. 1998. "Emigration Dynamics in Developing Philippines: Results and Experiences from an ILO Project." Countries: Case Study on South Asia (IOM/UNFPA Proj- Geneva: ILO. ect)." In P. Brownlee and C. Mitchell, eds., Migration Re- Kritz, M. M., L. L. Lin, and H. Zlotnik, eds. 1992. International search in the Asia Pacific: Theoretical and Empirical Issues. Migration Systems: A Global Approach. Oxford: Clarendon Wollongong, Australia: Asia Pacific Migration Research Net- Press. work Secretariat. Lim, L. L. 1996. "The Migration Transition in Malaysia." Asian Basch, L., N. Glick-Schiller, and C. S. Blanc. 1994. Nations Un- and Pacific MigrationJournal 5(2-3). bound: Transnational Projects, Post-Colonial Predicaments and Lim, L. L. and N. Oishi. 1996. "International Labor Migration Deterritorialized Nation-States. New York: Gordon and of Asian Women: Distinctive Characteristics and Policy Con- Breach. cerns." Asian and Pacific Migration Journal 5(1): 85-116. Battistella, G. and A. Paganoni. 1992. Philippine Labor Migra- Martin, P., A. Mason, and T. Nagayama. 1996. "Introduction tion: Impact and Policy. Quezon City: Scalabrini Migration to Special Issue on the Dynamics of Labor Migration in Asia." Center. Asian and Pacific Migration Journal 5(2-3). Blanc, C. S. 1996. "Balikbayan: A Filipino Extension of the Na- Martin, P. L. 1991. The Unfinished Story: Turkish Labour Migra- tional Imaginary and of State Boundaries." Philippine Socio- tion to Western Europe. Geneva: International Labour Office. logical Review 44(1-4): 178-93. Massey, D. S., J. Arango, G. Hugo, A. Kouaouci, J. E. Taylor, Bohning, W. R. 1998. "The Impact of the Asian Crisis on Fil- and A. Pellegrino. 1993. "Theories of International Migra- ipino Employment Prospects Abroad." Geneva: International tion: A Review and Appraisal." Population and Development Labour Office. Review 19(3): 431-66. Borjas, G. J. 1989. "Economic Theory and International Migra- Massey, D. S., J. Arango, G. Hugo, and J. E. Taylor. 1994. "An tion." International Migration Review Special Silver Anniver- Evaluation of International Migration Theory: The North sary Issue, 23(3): 457-85. American Case." Population and Development Review 20: Boyle, P., K. Halfacree, and V. Robinson. 1998. Exploring Con- 699-751. temporary Migration. Harlow, Essex: Longman. OECD 1997. Trends in International Migration: Annual Report Cahill, D. 1990. Intermarriages in International Contexts. Que- 1996. Paris: OECD. zon City: Scalabrini Migration Center. OECD 1998. Trends in International Migration: Annual Report Castles, S., H. Booth, and T. Wallace. 1984. Here for Good: 1998. Paris: OECD. Western Europe's New Ethnic Minorities. London: Pluto Press. Paine, S. 1974. Exporting Workers: The Turkish Case. Cambridge: Castles, S. and G. Kosack. 1973 and 1985. Immigrant Workers Cambridge University Press. and Class Structure in Western Europe. London: Oxford Uni- Portes, A. and J. Borocz. 1989. "Contemporary Immigration: versity Press. Theoretical Perspectives on Its Determinants and Modes of In- Castles, S. and M. J. Miller. 1998. The Age of Migration: Inter- corporation." International Migration Review 28(4): 606-30. national Population Movements in the Modern World. 2d ed. Rowlands, D. 1998. "Poverty and Environmental Degradation London: Macmillan. as Root Causes of International Migration: A Critical Assess- Diatta, M. A. and N. Mbow. 1998. "Releasing the Development ment." Paper presented at United Nations Technical Sympo- Potential of Return Migration: The Case of Senegal." Paper sium on International Migration and Development, 29 presented at United Nations Technical Symposium on Inter- June-3 July 1998. The Hague, Netherlands: unpublished. 56 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Saith, A. 1997. Emigration Pressures and Structural Change: Case Zolberg, A. R., A. Suhrke, and S. Aguayo. 1989. Escapefrom Vio- Study of the Philippines. Geneva: International Labour Office. lence. New York: Oxford University Press. Shah, N. M. 1994. "An Overview of Present and Future Emi- gration Dynamics in South Asia." International Migration 32(2): 217-68. 1. Countries of emigration may also be referred to as migrant- Skeldon, R. 1997. Migration and Development: A Global Perspec- sending countries or countries of origin. These terms are used tive. Harlow, Essex: Addison Wesley Longman. synonymously. Stahl, C. 1993. "Explaining International Migration." In C. 2. A comparative project of the United Nations University on Stahl, R. Ball, C. Inglis, and P. Gutman, eds., Global Popu- Asian migration to the Gulf oil states produced important findings lation Movements and their Implications for Australia. Can- (see Amjad 1989, Gunatilleke 1991). A major joint project of the In- berra: Australian Government Publishing Service. ternational Organization for Migration and the United Nations Pop- Stalker, P. 1994. The Work of Strangers: A Survey of International ulation Fund, which started in 1993, has already produced major re- Labour Migration. Geneva: ILO. ports, most of which have been reported in special issues of the IOM's Tomas, P. S. 1998. "Enhancing the Capabilities of Emigration journal International Migration (Appleyard 1995 and 1998). The Countries to Protect Men and Women Destined for Low- OECD's annual Trends in International Migration contains regular Skilled Employment: The Case of the Philippines." Paper reports, mainly from European emigration countries (OECD 1997). presented at United Nations Technical Symposium on Inter- 3. For overviews of migration research and theory see: (Massey, national Migration and Development, 29 June-3 July 1998. et al. 1993); (Massey, et al. 1994); (Skeldon 1997); (Boyle, et al. The Hague, Netherlands: unpublished. 1998), Chapter 3; (Castles and Miller 1998), Chapter 2. United Nations. 1998. "Report of the United Nations Techni- 4. A good discussion of the costs and benefits of emigration can cal Symposium on International Migration and Develop- be found in Skeldon 1997, chapter 6. Each cost and benefit is highly ment, The Hague, Netherlands 29 June-3 July 1998." New complex, and migration studies generally attempt to analyze them York: United Nations. for each migration system. Wong, D. 1996. "Foreign Domestic Workers in Singapore." 5. There is a considerable literature on Turkish emigration and Asian and PacificMigration Journal5(1): 117-38. its effects on the economy and society. An early economic study is Yukawa, J. 1996. Migration from the Philippines 1975-1995: An Paine 1974. Turkish scholars have published many works, notably Annotated Bibliography. Quezon City: Scalabrini Migration Professor Nermin Abadan-Unat (for instance, Abadan-Unat 1988). Center. Martin 1991 provides a valuable overview, which is used extensively Zelinsky, W. 1971. "The Hypothesis of the Mobility Transi- here. tion." Geographical Review 61(2): 219-49. 6. Here too, considerable literature is available. For a bibliogra- Zolberg, A. R. 1989. "The Next Waves: Migration Theory for a phy on the Philippines, see Yukawa 1996. Much material can be Changing World." International Migration Review 23(3): found in the Asian and Pacific Migration Journal, published in 403-30. Manila since 1992. See, for instance, Abella 1993. -A N>Managing Urbanization in a Globalizing X VV8 World chapter 6 Cities and Trade: External Trade and Internal Geography in Developing Economies Anthony J. Venables London School of Economics and Centre for Economic Policy Research H he evolution of the economic geogra- deconcentration of manufacturing from Seoul to outly- phy of developing countries, in par- ing cities. This research also addresses the question of ticular the process of urbanization, what functions are performed by cities; Seoul maintained is well documented (for example, a diversified manufacturing base, but as activity moved Rosen and Resnick 1980; United Na- out of Seoul to other cities, so these tended to specialize tions 1991), but perhaps less well un- in particular sectors (see Henderson, Chapter 7). derstood. The proportion of the population in urban This shifting balance between concentration and dis- centers increases with development (World Bank 1999), persion-both of activity as a whole and of particular but how is this urban population distributed between sectors-arises from the interplay of a number of differ- cities of different types and in different locations? What ent forces. Marshall, writing in 1890, listed three forces factors determine the size structure of cities in develop- that encourage concentration of activity: technological ing countries, and why does the largest city tend to be so externalities ["the mysteries of the trade become no mys- dominant? Is this concentration of activity good for tery but are, as it were, in the air .. .." (Marshall 1890)]; growth, or would countries be better served by a more benefits arising from thick labor markets, such as labor dispersed structure of economic activity? To answer these market pooling effects and externalities in training; and questions we also need to understand the functional forward and backward linkages that occur between firms structure of cities and the extent to which they specialize that are linked through the production and supply of in- in different activities. termediate goods and services.' Recent research has out- Empirical research indicates that prime cities in de- lined the benefits that come from concentration of re- veloping countries are much more dominant than simi- search activities (for example, Audretsch and Feldman lar cities in high-income countries are-or were. It also 1996), from being in close-knit production networks indicates that the dominance of the prime city tends to (for example, Porter 1990), and from locating in cities. peak during the development process (Williamson 1965; In a recent Journal ofEconomic Perspectives paper review- more recently confirmed by Wheaton and Shishido 1981 ing evidence on the subject, Quigley (1998) concludes, and Henderson 1998). Determinants of the factors that "it remains clear that the increased size of cities and their lead to dominance are researched by Ades and Glaeser diversity are strongly associated with increased output, (1997), who show that trade barriers, poor internal productivity and growth.. . " (See Mills, Chapter 8.) transport, and the concentration of political power all These forces foster spatial concentration, but the contribute to the dominance of the largest city. The world is not a single megacity. Forces for concentration deconcentration of activity that can occur in later stages are opposed by "centrifugal" forces that encourage dis- of development is nicely illustrated by the Korean exam- persion of economic activity. Chief among these are sup- ple. Henderson et al. (1998) show how increasing urban plies of immobile factors (such as land and possibly congestion and infrastructure improvements led to the labor) which lead to spatial variations in rents and 58 CITIES AND TRADE: EXTERNAL TRADE AND INTERNAL GEOGRAPHY IN DEVELOPING ECONOMIES 59 wages, the presence of external diseconomies such as particular industries in which linkages are relatively congestion costs and pollution, and the need for firms strong. to meet demand from spatially dispersed consumers. Although economists have researched these separate Analytical Framework forces, little has been done to see how they all interact My approach is based on a new economic geography to determine the actual location of economic activity.2 model which is described informally here, referring the The omission is now being rectified by recent research reader to Fujita et al. (1999) for full development. In on "new economic geography." Taking into account the order to focus on the tension between forces for agglom- components just outlined, the research shows how con- eration and dispersion the model abstracts from many centrations of economic activity can arise and evolve. It issues that are usually regarded as important in eco- shows how, even if locations are broadly similar in their nomic development. We consider just two countries and underlying characteristics, they may nevertheless de- assume that they have the same relative endowments of velop very different economic structures as concentra- two factors of production, labor and land. One country tions of activity form. This process may go on at differ- is large, and we call it the Outside country-it can be ent levels. At the urban level it can account for the thought of as the rest of the world. We focus on the formation of cities, while at the international level it other country, which we call Home, and assume to be goes some way to explain the concentration of economic divided into two regions. Each Home region has the activity in the "North" and consequent North-South in- same amount of land, and Home labor is mobile be- come inequalities. tween the two regions. Within each region there is a sin- At each of these levels there are a number of common gle city (or potential city site). features.3 First, very small initial differences between There are two production activities, agriculture and locations may become magnified through time, as manufacturing. Agriculture takes place in all locations, processes of "cumulative causation" drive growth. Sec- using land and labor to produce a perfectly tradable out- ond, spatial inequalities can be very large; furthermore, put.5 Manufacturing can operate in any of the locations; they are likely to be largest at "intermediate" stages of de- our analysis will establish the locations in which produc- velopment. Third, once established, spatial structures tion actually takes place. Manufacturing has the usual may become "locked in" and persist for long periods of characteristics of increasing returns to scale (at the firm time, although when they do start to break down, change level), product differentiation, and monopolistic com- may be quite abrupt (a phenomenon akin to the "punc- petition. Firms use labor and intermediate goods as in- tuated equilibria" of evolutionary theory). And finally, puts, and their output is used both as a consumer good economic growth and development will not take the and as an intermediate. This input-output structure form of the smooth convergence of income levels of dif- gives rise to forward and backward linkages between ferent regions. Instead, it is more likely to involve rapid firms, as each firm uses output supplied by other firms growth of a few regions, while others are left behind. Es- (a forward linkage) and sells some of its output to other sentially, the rich club expands to include more mem- firms (a backward linkage). Shipping manufacturing bers, but gaps between the rich and the poor remain.4 output incurs transport costs. On internal trade between The objective of this chapter is to show how ideas the two Home regions the transport cost per unit from this new economic geography literature can be shipped is denoted T, and on external trade between ei- brought to bear on the questions we have posed con- ther of the Home regions and Outside the cost per unit cerning the city structure of developing economies. The imported or exported is To. Notice that the two Home chapter is a speculative exploration of issues rather than regions are constructed to be identical-neither has the a definitive piece of research, but nevertheless provides benefit of proximity to the outside world.6 a number of insights. The analysis suggests that in the This structure tends to generate concentrations of early stages of economic development countries are both manufacturing activity and population. They arise likely to have a monocentric structure, with manufac- because firms want to be close to other firms supplying turing concentrated in a single dominant economic cen- intermediate goods (in order to save transport costs ter, but that this will evolve into a multicentric structure on these intermediates). They also want to be close to a as development proceeds. Furthermore, policies that in- large market, which again means being close to other hibit the development of this structure are likely be eco- firms in order to meet demands for intermediate goods. nomically damaging. Openness to international trade As firms cluster in one location, so too does labor-the brings benefits both from facilitating deconcentration firms' employees-further increasing market size. Work- of population and from allowing the agglomeration of ing against these agglomeration forces, we assume that 60 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION there are some congestion diseconomies that increase Figure 6.1 with the size of the region's population. These have to be Internal Transport Costs and Regional Industrialization offset by firms in the congested region paying higher wages, and this in turn counts against the attractiveness Share of region's labor force in industry of this region as a base for manufacturing. The actual lo- cation of manufacturing-and of manufacturing work- 0.20 - ers-is determined by the balance of these forces. We will see some cases in which all manufacturing is con- 0.16 centrated in a single location and others in which it Region 1 spreads out, giving a less concentrated spatial structure. 0.12 Our task is to identify the key determinants of this spa- ' tial structure. 0.08 - s.Total We take as a starting point in our experiments an ini- tial situation in which all manufacturing is concentrated 0.04 - Region 2 S "' in the Outside country. Outside consequently appears ISymmed \ to be a developed economy, and the demand for labor .0 L I IoI I I in manufacturing means that Outside's wage rate is 1.0 1.1 1.2 1.3 1.4 1.5 1.6 higher than Home's. However, despite this wage gap, it Internal transport costs is not profitable for any single firm to relocate from , Outside to Home because if it were to do so, it would forego the benefits of forward and backward linkages to opment. What else do we learn from the figure? The key other Outside firms. point is that, over a range of trade costs, only one of the Home country's regions industrializes, despite the fact that the regions are constructed to be symmetric.7 Why In this framework economic underdevelopment is sim- should development lead to this monocentric structure? ply a corollary of the spatial agglomeration of manufac- The answer is that a situation in which development oc- turing, and development occurs if the agglomeration curs in both regions simultaneously is unstable; if one breaks down or spreads to new locations. Let us now see of the regions got slightly ahead, then the forward and how changing trade costs can cause the Home country backward linkages from its firms would create an advan- to attract manufacturing, and see also how Home's in- tage that would cumulate over time. With the symmet- ternal economic geography evolves during this process. ric structure of this model there is nothing to determine which region industrializes, but because of the benefits Internal Trade Costs of agglomeration, it cannot be both. The region that at- Our first experiment captures the effects of Home in- tracts manufacturing also attracts population, as labor frastructure improvement, represented by a reduction in demand from industry brings in-migration until real internal transport costs between the two Home regions. wages, net of congestion diseconomies, are the same in Results are illustrated in Figure 6. 1, which has internal both regions. transport costs on the horizontal axis, and the share of As Figure 6.1 shows, the monocentric internal geog- each region's labor force employed in manufacturing on raphy occurs only for a range of internal transport costs. the vertical. (This and subsequent figures come from As these costs are reduced further, there comes a point numerical simulation with particular functional forms at which industry spreads to the second location. This and parameter values; they are illustrations of possibili- happens for two reasons. First, forward and backward ties, not general results.) linkages become less geographically concentrated as . At high levels of internal transport costs the economy transport costs fall, making it less costly for a firm to is in the initial situation-neither Home region has any move to the other location.8 And second, congestion dis- manufacturing employment. Reducing the internal economies increase the wages that must be paid by firms transport barrier has the effect of making Home a more in the industrialized location to retain workers, and attractive location, since a firm can now supply the en- thereby create an incentive for firms to leave. The pat- tire Home market more cheaply, and the figure shows tern we see then, is that at early stages of industrializa- that there is a level of transport costs below which Home tion, there is a monocentric structure, but at later stages starts to attract manufacturing industry. Reducing in- (in this example, at lower internal transport costs), a ternal transport costs therefore triggers industrial devel- duocentric structure develops. CITIES AND TRADE: EXTERNAL TRADE AND INTERNAL GEOGRAPHY IN DEVELOPING ECONOMIES 6i This example shows how the primacy of one region Figure 6.2 may occur as industrialization takes off. Does this out- External Trade Barriers and Regional Industrialization come lead to earlier development or higher income lev- els than one in which activity is more uniformly distrib- Share of region's labor force in industry uted? The dashed line in Figure 6.1 gives the share of 0 40 - the entire Home labor force in industry, and the light 0.36 - line is the share if we force the two locations to be iden- 0.32 - tical. The point at which industrialization starts is the 0.28 - same in both cases. But during the monocentric phase, 0.24 \ forcing duocentricity would reduce the overall share of o.@ A Region 1 the labor force in manufacturing and consequently re- 0.20 duce real income. The reason is that forward and back- 0.16 - ward linkages are lost, and these linkages are good for 0.12 - industrialization and real income. 0.08 - How does this story change if there are more than 0.04 - Region 2 - Symmetry two Home regions? The answer is that we see industry imposed 0.00 L I I I I I I I I I I I I I start first in one region, then spread into a second, then 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 a third, and so on, as industrialization occurs in regions External transport costs sequentially, not simultaneously. Forcing simultaneous development of all regions is even more costly, since no region has enough industry to reap the benefits of agglomeration. As this more outward-looking industrialization oc- curs it initially gives rise to a monocentric internal ge- Internal Geography and External Trade ography because (like the previous case studied) having We now consider a second experiment in which the dri- industry in both locations is unstable. However, as ex- ving force for industrialization is the Home economy's ternal trade costs are reduced and levels of industrial em- openness to international trade. Figure 6.2 illustrates ployment increase, so industry spreads to region 2 and possibilities, mapping out the share of manufacturing the economy develops a duocentric structure. The rea- employment in each region as a function of external sons for the spread are that the diseconomies of urban trade costs, To. concentration have become stronger, while more out- Several striking points emerge from this example. ward orientation reduces linkage strength within the When Home is relatively closed (high TO), it will have economy. a substantial amount of industrial employment, but this As in the previous case then, we see that industrial- will be concentrated in a single location. The reason ization, which is driven by closer integration in the Home has industry is simply the need to supply local world economy, first involves a monocentric structure consumers who, at very high trade costs, do not have ac- and then evolves into a more uniform duocentric struc- cess to imports. The reason for the monocentric struc- ture. We can study the costs and benefits of the mono- ture is the one given by Krugman and Livas (1996). centric structure by comparing it with a situation in With high external trade costs firms and consumers are which a symmetric duocentric structure is imposed. As inward looking, purchasing largely from local firms, and before, the light solid line illustrates the share of em- this makes the internal linkage effects very strong. ployment in manufacturing if the two Home regions are At intermediate trade costs Home has less industry forced to be identical, and comparison with the dashed because the economy is now more open to imports, and line indicates lower manufacturing employment and this benefits the existing agglomeration of industry in lower welfare. Outside. However, as external trade costs fall further, so Home starts to attract industry. To understand why, Regional Integration note that a firm moving into the country from the ex- An alternative to multilateral external trade liberaliza- isting Outside agglomeration pays lower wages but fore- tion (falling To) might be regional integration-prefer- goes forward and backward linkages with firms in the ential tariff reductions within a regional trading bloc. agglomeration. At low trade costs the cost of foregoing By reinterpreting the regions in our model we can offer these linkages is small-intermediate inputs can be im- some predictions about the likely effect of such a policy ported and Outside demand met by exports.9 on the location of industries in the member states. Sup- 62 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION pose that the two regions in our model are two coun- employment in each of them is fixed.'2 Although total tries in a trading bloc and that T, the internal transport employment in each industry is fixed, its location is not, cost, is now reinterpreted as the internal tariff between and we want to see how this is determined. Will each the countries. What happens if there is a preferential lib- industry operate in a different location or will they be eralization, reducing this tariff while holding external divided in some way between locations? There are three trade barriers, To, constant? The analysis is similar to main forces. First, firms will want to locate close to that illustrated in Figure 6. 1. Intrabloc liberalization other firms in the same industry in order to exploit for- promotes industrialization-firms in the bloc gain from ward and backward linkages; this is a force for the clus- better access to a larger market-but, as in Figure 6.1, tering of each industry separately. Second, firms will industrialization will occur in just one country (or want to locate in the region with the larger demand region). The reason is as before; if one country gets from consumers; this is a force for agglomeration of slightly ahead of the other, then its advantage will cu- manufacturing as a whole, since the location with the mulate as firms benefit from forward and backward more firms will have more workers and a larger con- linkages with other firms in the same location. sumer demand. Third, if there are congestion dis- Two further remarks are in order. First, the likelihood economies, then we have a force pushing in the oppo- that formation of the trading bloc will cause unequal de- site direction, encouraging dispersion of activity. velopment is larger the higher is the external tariff ( To). Figure 6.3 illustrates how the balance between these A high external tariff forces firms to be more dependent forces may depend on external trade barriers. The hori- on other local firms, so increasing the strength of for- zontal axis is this external trade cost, To, and the verti- ward and backward linkages. Second, our model as- cal is the share of manufacturing employment by region sumes that labor is mobile between regions-an assump- and industry, with subscripts referring to Home regions tion that is unlikely to be correct for countries within a (1 and 2) and superscripts referring to industries (A and trading bloc. If labor is immobile, then the likelihood of B). Thus Li is the share of the manufacturing workers agglomeration is reduced, but not removed. located in region 1 (Li + L2 = 1), and L1A is the share of The prediction that regional integration might pro- manufacturing workers located in region 1 and working mote inequalities between members of the bloc fits well in industry A. with at least some historical experience. The East The figure shows that the two regions have a hierar- African Community collapsed because of the perception chical structure that evolves as trade barriers change. At that the Nairobi region was capturing all of the benefits; high external trade costs most of the manufacturing Abidjan and Dakar are attracting a very high proportion labor force is in region 1 (Li is much larger than L2) of the new economic activity developing in the Eco- because of the tendency of inward-looking firms to nomic Community of West Africa; and Paraguay is now claiming that it is losing industry to its larger South American Common Market (MERCOSUR) partners, Figure 6.3 Amerginaand Brazil rktr.1 External Trade and Internal Economic Geography Argentina and Brazil.' v Employment share by industry and region City Specialization and Hierarchy 1.00 In the discussion so far we have assumed that linkages occur at the level of manufacturing as a whole. This L,=LA+L,8 means that growth of one industry benefits all others by 0.75 - demanding output or providing a supply of intermedi- ates, either directly or indirectly. However, some link- ages operate at a much narrower level than this, for 0.50 L A example, linkages between firms within a specialized in- dustrial sector. If we move to a multi-industry setting in L,A which the input-output linkages occur primarily within 0.25 - particular industries rather than between them, then what sort of economic geography develops?"' Can we shed light on the evolving pattern of city specialization 0.00 L , , I I I accompanying development? 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 To address these questions let us suppose that there External trade barrier are two manufacturing industries and that Home's total CITIES AND TRADE: EXTERNAL TRADE AND INTERNAL GEOGRAPHY IN DEVELOPING ECONOMIES 63 agglomerate, although full agglomeration is prevented creasing returns to scale and linkages between the loca- by congestion costs. Given that most industry is in re- tion decisions of firms and workers. This program of re- gion 1, what do we know about the industrial structures search is in its infancy, and in this chapter we have sug- of the regions? Because firms want to locate close to gested how it may be applied to analyze the effects of other firms in the same industry, one industry will be trade-internal, external, and regional-on the eco- completely concentrated and correspondingly one re- nomic geography of developing countries. Our analysis gion completely specialized. In this example region 2 is suggests the following answers to the questions posed in specialized in industry B, while region 1 has all of in- the introduction. First, economic development should dustry A (since L2A = 0) and also some of industry B. not be regarded as a smooth process of convergence, but Reducing trade costs has two interesting effects. It rather as the uneven spread of clusters of activity. It then leads to spatial deconcentration of population and con- follows that development may typically involve the con- centration of particular industries. The deconcentration centration of activity in dominant cities or regions. This of population arises for the reasons we have already concentration is beneficial insofar as it raises the level of noted (following Krugman and Livas); greater outward industrialization and real income (net of congestion orientation makes linkages less strong compared to con- costs). Second, the internal economic geography that gestion costs; it shows up as the decline in L1 and in- develops is sensitive to levels of transport costs and other crease in L2. Deconcentration of population equalizes barriers to trade; reducing transport costs promotes in- the size of the regions, and this facilitates the clustering dustrialization and also facilitates the spread of industry of each industry-region 2 gains a large enough popu- to new Home locations. Thus, the models support the lation to accommodate the whole of industry B, so we idea that concentration will increase in the early stages see L2B increasing and LIB falling. As trade costs are re- of development and then decline, as suggested by the duced the process continues until at low enough trade empirical work of Williamson and others. Finally, there costs the hierarchical regional structure breaks down are important gains from external openness. In addition completely and the regions have the same population to the usual gains from trade, we see that openness may size and complete industrial specialization. lead to a rearrangement of internal economic geography. This story is consistent with the experience of Korea It may promote deconcentration of population, while that we have previously noted (Henderson et al. 1998). at the same time facilitating clustering of particular in- The theoretical modeling enables us to infer the real in- dustries in which linkages are strong, both these changes come effects of the changes, and it turns out that there being sources of real income gain. are two important sources of gain. On top of the usual benefits of trade liberalization (the direct effects of re- Bibliography ducing trade costs and any comparative advantage or Ades, A. F., and E. L Glaeser. 1997. "Trade and Circuses: pro-competitive gains from trade), there are also gains Explaining Urban Giants." Quarterly Journal of Economics from the reorganization of the economy's internal eco- CX(M): 195-227. Audretsch D., and M. Feldman. 1996. "R&D Spillovers and the nomic geography. Deconcentrating population reduces Geography of Innovation and Production." American Eco- congestion costs, and clustering of particular industries nomic Review 86(4): 253-73. gives the benefits from intraindustry linkages. Easterly, W., and R. Levine. 1999. "It's Not Factor Accumula- tion: Stylized Facts and Growth Models." Washington, D.C.: Conclusions World Bank. Fujita, Masahisa, P. R. Krugman, and A. J. Venables. 1999. The Standard techniques of economic analysis are good at Spatial Economy: Cities, Regions and International Trade. dealing with situations in which diminishing returns to Cambridge, MA: MIT Press. activities yield "smooth" outcomes. However, the out- Glaeser, E. 1998. "Are Cities Dying?" Journal of Economic Per- standing feature of economic geography is that out- spectives 12: 139-60. comes are extremely "lumpy." This shows up in the for- Henderson, J. V. 1999. "The Effects of Urban Concentration on matio of cties,in th conrastig perormane of Economic Growth." Processed, Brown University. mation of cities, in the contrasting performance of Henderson, J. V., T. Lee, and Y.-J. Lee. 1998. "Externalities, Lo- different regions, and in spatial income inequalities in cation and Industrial Deconcentration in a Tiger Economy." the world economy. It also shows up in countries' Processed, Brown University. growth performance where the long-run picture is of di- Krugman, P. R. 1991. "Increasing Returns and Economic Ge- vergence, not convergence, of economic performance ography." Journal ofDevelopment Economics 49(1): 137-50. (Easterly and Levine 1999). Krugman, P. R., and G. Hanson. 1993. "Mexico-U.S. Free (Easterly and Levine 1999). Trade and the Location of Production." In P. Garber, ed., The "new economic geography" literature addresses The Mexico U.S. Free- Trade Agreement. Cambridge, MA: these issues and takes seriously the implications of in- MIT Press. 64 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Krugman, P. R., and R. E. Livas. 1996. "Trade Policy and the Notes Third World Metropolis." Journal of Development Economics 49(1): 137-50. 1. For a recent and more comprehensive discussion of these Krugman, P. R., and A. J. Venables. 1995. "Globalization and forces see Glaeser (1998). the Inequality of Nations." Quarterly Journal of Economics 2. Part of the reason for this is the technical difficulty of ana- CX: 857-80. lyzing situations with clustering. Often the attraction of a city Marshall, A. 1890. Principles of Economics. London: Macmillan. arises not from its exact location, but simply from the fact that it Porter, M. E. 1990. The Competitive Advantage of Nations. New is a city. Thus, there is a degree of indeterminacy in the analy- York: Macmillan. sis-many sites are perfectly suitable places to build a city, but Puga, D. 1998. "Urbanisation Patterns: European vs. Less De- once established it becomes "locked in" to the selected site. veloped Countries." Journal of Regional Science 38: 231-52. 3. See Fujita, Krugman, and Venables (1999) for a synthesis Puga, D., and A. J. Venables. 1996. "The Spread of Industry: of models in each of these contexts. Puga (1998) has analyzed an Agglomeration in Economic Development." Journal of the urban model in a developing country context. Krugman (1991) Japanese and International Economies 10: 440-64. sets out a regional model, and Krugman and Venables (1995) Puga, D., and A. J. Venables. 1998. "Trading Arrangements and and Puga and Venables (1996) develop international models. Industrial Development." World Bank Economic Review 12: 4. See Easterly and Levine (1999) for empirical support for 221-49. this position. Quigley, J. M. 1998. "Urban Diversity and Economic Growth." 5. The "agricultural" sector should be interpreted as a com- Journal of Economic Perspectives 12: 127-38. posite of the perfectly competitive "rest of the economy." Rosen, K. T., and M. Resnick. 1980. "The Size Distribution of 6. We use the usual "iceberg" transport costs; thus Tdenotes Cities: An Examination of the Pareto Law and Primacy." the number of units that have to be shipped for one unit to ar- Journal of Urban Economics 8: 165-86. rive at its destination. T= 1 is perfectly costless trade. United Nations. 1991. World Urbanization Prospects 1990. New 7. When T lies between 1.2 and 1.45, region 1 has manufac- York: United Nations. turing employment and region 2 does not. Venables, A. J. 1999. "Regional Integration Agreements: A Force 8. Other sources of agglomeration-for example, knowledge for Convergence or Divergence." World Bank Policy Re- spillovers-may be less sensitive to transport costs than are the search Paper No. 2260. Washington, D.C.: World Bank. input-output links assumed here. Wheaton, W., and H. Shishido. 1981. "Urban Concentration, 9. Quite generally in models of this type agglomeration forces Agglomeration Economies and the Level of Economic Devel- are strongest (relative to other locational forces) at intermediate opment." Economic Development and Cultural Change 30: levels of trade barriers. At very high trade barriers the need to Williamson, J. 1965. "Regional Inequality and the Process of supply immobile consumers prevents agglomeration, and as trade National Development." Economic Development and Cultural barriers go to zero the spatial dimension of linkages go to zero, Change 3-45. so the presence of any centrifugal forces prevents agglomeration. horld Bank. 1999. "Entering the 21st Century: World Develop- 10. For further development of these ideas see Puga and Ven- ment Report 1999/2000." Washington, D.C.: World Bank; ables (1998), and Venables (1999). New York: Oxford University Press. 11. This material is drawn from Chapter 18 of Fujita, Krug- man, and Venables (1999). This chapter was originally a background paper written for the 12. In other words, we abstract from the question of whether 1999 World Development Report, "Entering the 21st Century." the economy has industry or not. We assume that it has, and Thanks to Simon Evenett for valuable comments. study only where it locates. chapter _ On the Move: Industrial Deconcentration in Today's Developing Countries Vernon Henderson Eastman Professor of Political Economy Brown University ountries that have been developing rates. It also finds that many countries tend to have ex- during the past half-century have cessive concentration; many have reasonable concentra- followed an urban growth pattern tion; and few are underconcentrated. Why does exces- N._ that differs from that of earlier in- sive concentration lead to growth losses? dustrializing countries, which urban- One reason involves a country's ability to participate ized much more slowly. The recent in world markets, as well as a high overall level of indus- model consists of three stages: initial rapid population trial efficiency. Megacities become inordinately expen- growth and industrialization of a main city, such as sive locations within which to produce many products. Jakarta until the mid- 1970s or Seoul in the 1960s and Another reason is the quality of life in a country's early 1970s; decentralization or suburbanization of megacity, especially its environmental quality: In large population and manufacturing into the core city's outer congested areas, poor water quality and related diseases, rings and satellite cities, as in Jakarta today and in Seoul long commuting times, and the inequality of these ef- from the mid-1970s to the mid-1980s; and decentral- fects on the urban poor are major problems. A third ization of manufacturing-and to a lesser extent popu- concern-related to the quality of life-is the lack of lation-from the major metropolitan area into the rest technical expertise and managerial resources to effec- of the country, as in Seoul since the early 1980s or Sao tively plan, govern, and operate megacities. Fourth, and Paulo post-1970. also related to environmental quality, enormous invest- Does concentration in the core main city and its met- ment burdens need to be undertaken in providing ade- ropolitan area go too far before deconcentration sets in? quate infrastructure in megacities. Are megacities in developing countries too large or The first concern-the interaction between indus- growing too fast? Might earlier industrialization of trial competitiveness and efficiency and the urbaniza- medium-sized and smaller cities be desirable? Why have tion process-is particularly important in determining markets not produced better outcomes, if deconcentra- the economic vitality of developing countries. tion is delayed too long? Several concerns motivate these questions. First, PrOduction fic Citiency in s ejacities V5. econometric analysis suggests that an optimal degree of lodium-ized Cities urban concentration exists in countries, which first in- Studies suggest that in higher-income countries, research creases as national per capita incomes grow from low and development, corporate headquarters, financial, levels, peaks, and then declines with further income business, and professional services that are exchanged growth, as the three stages just outlined suggest (Hen- across cities, and experimental stages of manufacturing derson 1999). This analysis estimates that deviations operate most efficiently in very large metropolitan areas, from the best degree of urban concentration at any in- with their skilled labor forces and diverse economic ac- come level are reflected in significantly lowered growth tivity. Once production of a brand line becomes stan- 65 66 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION dardized, the need for a diverse, skilled force in a large arises for industry to move out of the dominant metro- metropolitan center diminishes. Producers then want to politan area. For example, Hansen (1983) specifically take advantage of the low-cost labor and land found in shows this pattern for Sao Paulo State in the early 1980s. smaller, more specialized cities. Kolko's work (1999) on Why do plants move out? Living costs, such as hous- the United States shows how the percentage of local em- ing, can be two to three times higher in megacities than ployment in manufacturing rises as city size declines, in smaller cities. Wages must therefore be at least twice being greatest in United States "rural" areas, and how the as high if workers with comparable skills are to earn sim- percentage of business and financial services rises with ilar real wages. What's more, for production to be as city size, being largest in megacities. The variation in viable in megacities as in small cities, the former must percentages is over twofold for services and almost have much higher capital-labor ratios and output per twofold for manufacturing, between rural areas or small worker-typically twice as high (Henderson 1988). For urban areas and the largest cities. most standardized production, the difference in produc- One explanation for the variation in manufacturing tivity between megacities and small cities is not great activity is that scale economies in standardized manufac- enough to support the higher wages and land rental costs. turing appear to result from localization, not urbaniza- tion-that is, from the level of activity within an indus- try, not from city size or diversity. Only new high-tech If deconcentration of manufacturing and industrial de- industries appear to rely on large, diverse urban econo- velopment of medium-sized cities is a natural process of mies (Henderson, Kuncoro, and Turner 1995). For re- a country's growth, what key factors determine the rate cently developed products, John Hekman has docu- at which deconcentration occurs? What might inhibit mented the relocation of standardized production to the process, and when might it occur later than desirable? smaller cities in the United States (1982). Other studies Investment in intercity infrastructure. For manufactur- have found similar results on scale economies and the ing in developing countries to migrate outside megaci- location of activities for countries such as Japan (Naka- ties, producers must have access to major ports-which mura 1985), Brazil (Henderson 1988), and Korea (Hen- are typically in megacities-and other transport nodes, derson, Lee, and Lee 1999). Fujita and Ishii (1994) have so they can transfer inputs and outputs and manage in- described this location pattern for Japanese electronics ventories efficiently. A world of just-in-time production companies (Table 7.1), which suggest these forces are and e-business based networks requires excellent trans- at work. port and communications. Thus only if regions invest In less-developed countries, much of the incubation in widespread transport facilities and telecommunica- stage of modern manufacturing production may entail tions can industry substantially deconcentrate. importing and adapting foreign technologies. Such de- Studies show a significant negative econometric rela- velopment may best focus on one or two metropolitan tionship between industrial concentration and crude nodes where foreign contacts and indigenous, highly measures of transportation systems such as the density skilled labor concentrate. However, as in wealthier coun- of the national rail system, road system, and waterway tries, once domestic production of a commodity is well system (Henderson, 1999; Rosen and Resnick, 1978). established, and perhaps standardized, a strong incentive Country examples also demonstrate the relationship. Table 7.1 Distribution of Plants in the Japanese Electronics Industry Japan Abroad Percentage of plants outside Japan that are in North America, Percentage of plants inside Japan in Japan's metropolitan the European Community, and newly industrialized nations, areas, for each category (total plants in each category) of all overseas plants (total overseas plants in each category) Headquarters and basic R&D 100 (70) Regional headquarters 100 (13) Development R&D 70 (54) R&D 91 (32) Trial production 91 (21) Production 59(335) Mass production 43 (333) Source: Fujita and Ishii 1994. ON THE MOVE: INDUSTRIAL DECONCENTRATION IN TODAY'S DEVELOPING COUNTRIES 67 Table 7.2 concentrated in cities in the interior. In China, strong Shares of Population and Manufacturing for beliefs regarding urban hierarchies of production have Selected Regions of the Republic of Korea, led state-owned industries-which are not very compet- 1983 and 1993 ldsaeondmutlswlhaentvr opt (percentage) itive-to locate in the largest metropolitan areas, while more private industry thrives in smaller cities. Red tape surrounding financing and production, im- Satellite cities in the provinces port and export licenses also appears to foster industrial Seoul, Pusan, containing concentration. The slow speed of deconcentration of and Taegu Seoul, Pusan, production from Seoul into Kyonggi province-its im- (Seoul) and Taegu mediate hinterland-in the 1970s appeared to be linked 1983 1993 1983 1993 to red tape. The acceleration of deconcentration within Population 37 (23) 39 (25) 14 20 Kyonggi in the early 1980s has been linked to dismant- Urban population 59 (37) 53 (34) n.a. n.a, Manufacturing Ing of these constramts. employment 44 (21) 28 (14) 30 30 However, economic liberalization does not always re- n.a., Not applicable. move the red tape joining production to megacities. In Source: Henderson, Lee, and Lee 1999. the early 1980s, Indonesia opened up banking and ex- port markets, enabling smaller producers to obtain pro- duction loans and export licenses. But firms had to apply for the major loans and licenses in Jakarta or a few Deconcentration of manufacturing in Sao Paulo since other major metropolitan areas. Thus the small firms the 1980s followed widespread infrastructure invest- could realistically obtain the loans and licenses only if ments in the late 1960s and early 1970s. In Indonesia, they maintained a strong presence in Jakarta or other rapid deconcentration of industry from Jakarta into the large urban areas. Liberalization with strings attached Botabek suburbs corresponded with construction of appears to have provoked the sharp reconcentration of highways east and west from Jakarta during the late small firms in the major metropolitan regions of Indo- 1980s and early 1990s. And in the Republic of Korea, nesia during the early and mid-1980s (Henderson and industrial deconcentration followed infrastructure in- Kuncoro, 1996). vestments in roads and in communications during the Fiscal decentralization. For manufacturers to move late-1970s and early-1980s (Table 7.2). For example, production in developing countries to smaller-sized communications infrastructure capital stocks had a 235 metropolitan areas, they must be able to find skilled percent differential between Seoul and the rest of the workers and managers there. The key to attracting country in 1972. By 1991 the differential had shrunk skilled workers to hinterland cities at reasonable wages to 25 percent. Korea blanketed the country with infra- is the public services and amenities such workers and structure in the late 1970s and early 1980s. As a result, their families demand and can get in big cities, includ- between 1983 and 1993, in rural areas and smaller met- ing high-quality schooling and medical care. Such con- ropolitan areas, the share of manufacturing employment sumer services are part of a larger package that includes outside the main three metropolitan areas and their services for the firms themselves, such as reliable local satellite cities climbed from 26 percent to 42 percent, power and local transport facilities, including container- even though population in those areas declined by ization, rail facilities, and roads for trucking. 9 percent. What is the best way to deliver such services and fa- State-owned industry, red tape, and economic liberal- cilities to firms and consumers in hinterland areas? One ization. Many central governments in developing coun- model is fiscal decentralization. The Charles Tiebout tries favor development of megacities, often the capital notion is that if localities can raise revenues and choose cities. Planners observing the higher productivity- the level of public services they provide, they will invest higher value added per worker-may assume that pro- in attributes that make their regions attractive to spe- duction is always more efficient in larger cities, without cific industries and types of workers. National govern- recognizing the disadvantages of steep labor and land ments can foster fiscal decentralization by permitting costs. In Brazil, this belief may have been partly respon- political decentralization. As countries urbanize, the sible for the high historical concentration of state-owned central government finds it increasingly difficult to industry in Sao Paulo, Rio de Janeiro, and the axis be- manage emerging large cities; those cities develop their tween them, while corresponding private production own political bases and demands for greater autonomy. 68 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION A key constraint on emerging local governments is fi- centration, especially in countries with reasonably sound nancing, particularly financing of long-term infrastruc- interregional transport and telecommunications. ture investments. The national government can help by Conclusions offering revolving-door loans for long-term investments to local governments. Revolving-door loan programs re- The form that urbanization takes in a country affects its quire localities to pay back loans to replenish the fund, development and economic growth. While some degree so loans can then be made to others (the revolving of urban concentration is essential for exploiting the door). While experience with revolving-door loans in gains from scale economies, as well as, under certain cir- some countries (Indonesia and India) has not been good cumstances, initial gains from globalization and tech- because the loans are never repaid, there have been pos- nology transfer, it appears that there can be a tendency itive results in recent years in Brazil. At the same time, towards excessive concentration in many countries. In- more sophisticated cities with strong finances may be terregional infrastructure investments, political decen- able to issue long-term bonds as a financing method. tralization, and a spatially neutral regulation of the For such bond markets to work well, the national gov- economy seem critical to ensuring a desirable level of ernment needs to be strict about not backing these debts urban concentration and urban development. in general, a policy which seems to encourage localities to effectively default. That destabilizes the national bud- Bibliography getary process, as experience in parts of Latin America Fujita, M., and T. Ishii. 1994. "Global Location Behavior and has shown. National governments can also help by de- Organizational Dynamics of Japanese Electronic Firms and veloping local technical and managerial expertise in Their Impact on Regional Economics." Paper prepared for urban planning, local taxation, and service dli Prince Bertil Symposium on the Dynamic Firm, Stockholm, ul ae dellvery. June 1994. Two major issues related to fiscal decentralization re- Hansen, E. R. 1983. "Why Do Firms Locate Where They Do?" main. First, it is unclear to what extent such decentral- International Bank for Reconstruction and Development, ization affects economic growth and income inequality. No. UDD25. Henderson (1999) finds no relationship between in- Hekman, J. 1982. "Branch Plant Location and the Product creased political decentralization and economicgrowthCycle in Computer Manufacturing." University of North creased political decentralization and economic growth. Crln origPpr ,, , . , , ~~~~~~~~~~~~Carolina Working Paper. Second, international agencies attempting to implement Henderson, J. V. 1988. Urban Development: Theory, Fact, and decentralization have often been insensitive to local po- Illusion. New York: Oxford University Press. litical realities, producing questionable results. It is un- Henderson, J. V. 1999. "The Effects of Urban Concentration on clear that half-hearted and poorly designed decentrali- Economic Growth." World Bank Working Paper. Processed. zation is a good solution. Washington, D.C. facilities . .Henderson, J. V., and A. Kuncoro. 1996. "Industrial Central- Globalization. Because megacities have the faclities ization in Indonesia." World Bank Economic Review 10: and skilled labor that enables firms to conduct inter- 513-40. national business and adopt foreign technologies, glob- Henderson, J. V., A. Kuncoro, and M. Turner. 1995. "Industrial alization may initially heighten regional inequality. To Development in Cities." Journal of Political Economy 103: operate successfully in medium-sized cities, interna- 1067-90. tional enterprises aiid joint ventures need excellent Henderson, J. V., T. Lee, and J.-Y. Lee. 1999. "Externalities and Industrial Deconcentration Under Rapid Growth." Mimeo, transport access to coastal cities, reliable global telecom- Brown University. munications, access to domestic capital markets, and Kolko, J. 1999. "Can I Get Some Service Here: Transport Costs, ready access to the domestic bureaucracy for required Cities, and the Geography of Service Industries." Mimeo licenses. (March), Harvard University. However, globalization also opens up new markets to Nakamura, R. 1985. "Agglomeration Economies in Manufac- Hinterl producers previously e d t des t turing." Journal of Urban Economics 17 1708-24. hinterland producers previously more tied to domestic Rosen, K., and M. Resnick. 1978. "The Size Distribution of markets than coastal producers. The array of new mar- Cities, the Pareto Law, and the Primate City." Mimeo, kets for hinterlands may thus promote industrial decon- Princeton University. chapter 8 The Importance of Large Urban Areas- and Governments' Roles in Fostering Them Edwin S. Mills Kellogg Graduate School of Management Northwestern University D uring the past 50 years, urban areas workers are also typically better educated than rural have become centers of growth and workers; indeed, educational opportunities are often an progress in all of the increasing num- initial reason for rural-urban migration. ber of countries that are achieving sig- nificant economic growth. No coun- Why Large Urban Areas Are More Efficient -try has achieved sustained economic Land prices in large urban areas may be 50 to 100 times growth without accompanying massive urban growth. as high as in rural areas less than 100 kilometers away. Some 75 or 85 percent of the population in the highest Why do nonagricultural producers concentrate in big income countries is urban, compared with 10-15 percent cities instead of scattering throughout the available space? in the poorest countries.' This strong correlation between The proximity of large numbers of workers, con- economic growth and urbanization is as close to a nat- sumers, and producers clearly enables people, goods, ural law as almost any social process in human history. and information to move economically. Producers com- And this correlation will continue in the future. pensate for high land prices by building upward, as of- Yet in many countries, there is strong ideological op- fices or residential complexes outbid other users for the position to large urban areas-in government, among most expensive land. Vertical expansion allows large the media, and in academic institutions. One of life's markets for many goods and services in major urban ironies is to be lectured on the evils of large urban areas areas, enabling firms to realize economies of scale and by an academic, government official, or media writer scope in many sectors. In large urban areas, many input who has lived his or her entire life in New York, Tokyo, purchases and output sales also can be made locally, and or Bombay and who could hardly imagine living any- there is a better match between workers and jobs. As a where else. Perhaps a part of the same ideology is the result, both sectoral outputs and plant sizes are greater widespread belief that poor residents should be sent in large urban areas than in small ones. back to rural areas where it is thought that they would The high density and proximity in urban areas pro- be better off. mote efficiency in nonbusiness sectors as well. Urban In fact, overwhelming evidence shows that the poor areas are centers of higher education, technology, and gain from rural-urban migration. In rapidly urbanizing innovation. Proximity permits economies of scale and middle-income countries, urban wages may be two to scope in these sectors and interaction within and be- four times rural wages (Mills and Becker 1986; Mazum- tween the sectors. For the same reasons, urban areas are dar 1987). Even after adjusting for cost-of-living differ- centers for the arts. Finally, urban areas are centers for ences, this gap remains large and is a primary motive for political reform. It is no accident that black Americans' migration. Urban workers have higher wages because demands for civil rights became effective only as large they concentrate in growing industrial and service sec- numbers of blacks migrated from southern farms to tors as opposed to stagnating agricultural sectors. Urban urban areas. Demands for political reform in the Repub- 69 70 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION lic of Korea were centered in Seoul and demands for re- countries, universities train scientists and engineers to form in China were centered in Beijing. evaluate and adapt imported technology to local mar- Similar effects apply to the movement of informa- kets and conditions. When governments locate research tion. Modern economic activity is highly specialized, re- laboratories or universities far from large urban centers, quiring a high volume of communication among busi- the distance imposes a barrier to research, and especially ness units. And face-to-face meetings are important for to translating results into industrial technology. some kinds of communication. Indeed, because mod- Schools themselves benefit from locating in large ern technology permits almost cost-free communication urban areas, just as other activities do. Student commut- over long distances, only the need for face-to-face com- ing distances are shorter, the schools can realize scale and munication justifies the extremely high office densities scope economies, and users can better exchange informa- and land values in central business districts and subur- tion concerning school quality and student achievement. ban subcenters. Urban areas may also see greater competition among Large urban areas economize not only on internal public schools and between public and private schools. movements of goods, people, and information but also Thus urban areas are centers of higher education, tech- on movements to other large urban areas. Scale econo- nology, arts, and innovation because proximity permits in- mies in producing and operating road, rail, water, and air teraction and economies of scale and scope. We can best transport make exchanges more efficient. Even small measure the difference in efficiency between urban and urban areas do not provide the same advantages, as high rural areas by comparing inputs and outputs-or what is transport costs eat up savings from large-scale production. called total factor productivity. When applied to urban Some popular writers have forecast that the spread of areas, such analyses are usually referred to as studies of ag- information technology will mean the demise of large glomeration economies. Virtually all careful estimates of urban areas. But the correlation between the expansion agglomeration economies have found that large urban of electronic communication and massive urbanization areas have higher total factor productivity than small urban undermines that theory. Indeed, the telephone, available areas (Eberts and McMillen 1999; Shukla 1996). Dou- for a century, appears to have promoted urban growth. bling the size of a sector or urban area typically produces Today the availability of faxes, word processors, photo- a 5- to 10-percent increase in total factor productivity. copiers, and e-mail in business hotels worldwide sug- (However, see Chapter 7 by Vernon Henderson.) gests that electronic communication complements Is there a limit to the size of urban areas that can show rather than substitutes for face-to-face communication greater productivity than smaller urban areas? Most au- (Gaspar and Glaeser 1998). thors assume that congestion and environmental deterio- Globalization reinforces the advantages of large urban ration will inevitably limit productivity gains. But in fact, areas. Globalization implies international specialization congestion and pollution are not inevitable; they are the and trade resulting from electronic communication and result of failed government policies. Even in large urban reduced government trade barriers. These processes will areas, appropriate investment in transportation, as well spread the benefits of large urban areas to developing as appropriate prices, can prevent congestion. Unfortu- countries. nately, government involvement in infrastructure all too Technological change originates and spreads more often fails to produce sound pricing and investment. easily in large urban areas, as Alfred Marshall and Jane Jacobs hypothesized and several scholars have confirmed Governments and Infrastructure (Glaeser 1997). The geographical concentrations of high- Categorizing a service as infrastructure sometimes im- tech firms in Silicon Valley, along Route 128 outside plies that government should supply it. However, the Boston, and in Research Triangle Park in North Carolina private sector provides virtually all kinds of infrastruc- provide informal but strong evidence of this effect. More ture somewhere in the world. Private companies effi- formal evidence comes from the fact that patent issues ciently supply almost every aspect of urban transporta- and various other proxies for technological change and tion, for example. Cars, trucks, and two-wheeled vehicles innovation tend to concentrate geographically. tend to be privately owned. Governments often own Major U.S. technology centers abut large urban cen- and operate buses, but in some countries the private sec- ters partly because they contain research-oriented uni- tor performs these functions. Similarly, governments versities with which they interact closely. Universities, commonly own and operate fixed-rail commuter sys- in turn, pursue research linked to technological change tems, but private entities build and operate a few, and and provide industrial laboratories with teachers and private companies sometimes operate government-owned students who produce new technology. In low-income systems. THE IMPORTANCE OF LARGE URBAN AREAS-AND GOVERNMENT'S ROLES IN FOSTERING THEM 71 Inadequate investment in urban transportation sys- Second, pricing infrastructure services correctly can tems and poor operation and pricing of these systems cut their effect on land values. Governments tend to un- can bring urban transportation almost to a standstill- derprice gas, electricity, clean water, and telephone ser- and has, to some degree, in cities such as Jakarta, vices. Some developing countries as well as the United Bangkok, Taipei, and Seoul. But no magic is entailed in States also price highway use by motor vehicles far below creating and maintaining a high-quality urban trans- cost. Fuel taxes are quite efficient for financing and con- portation system. Many countries, international lend- trolling road use: Many Organisation for Economic Co- ing institutions, and consulting firms maintain high- operation and Development (OECD) countries levy fuel quality planning staffs, and international institutions taxes to approximate highway user costs; $3-4 per U.S. offer loans for planning, building, and operating trans- gallon, including taxes, is probably the right range in portation systems. What's more, the returns on upgrad- most of the world's large urban areas.2 ing poor-quality systems are enormous. When governments resist charging for infrastructure But government corruption and favoritism toward services, they deprive themselves of market discipline on locally powerful groups often yield inappropriate invest- investment. Although careful benefit-cost analysis of ments. Indonesia is an example of such poor policymak- major projects could overcome this blind spot, such ing. The government spent billions of dollars planning analyses are even scarcer than adequate pricing. The domestic production of motor vehicles and aircraft at a problem is that in most countries high levels of govern- time when Jakarta was choking on traffic and had little ment finance and plan local projects. The United States traffic control, almost no public transit, and dramati- is typical: Federal or state governments plan and finance cally underpriced road use. most highway, water supply, and sewage treatment proj- Governments interfere with investment in commu- ects. If local governments were required to finance nications in a similar way because they invariably try to locally beneficial projects with locally raised taxes, resi- control the flow of information. Governments have dents would vote and lobby for projects that had sub- often maintained the telephone system as a public mo- stantial local benefits. nopoly, for example. Charges are typically low but in- Residents and businesses in each urban area invariably stallations are limited to politically favored institutions. want more infrastructure if it is financed by national Private families without political influence often wait taxes. People who favor limits on large urban areas also years to obtain telephone service, and the service that often want to locate the most heavily subsidized fixed-rail does exist is sometimes so poor that people send mes- commuter system and the best and most heavily sub- sengers instead. sidized universities in the largest urban areas. High- Several private phone companies worldwide have long quality benefit-cost analysis of alternative government in- been willing and able to install and operate phone sys- vestments could improve infrastructure allocation within tems anywhere. As with other infrastructure, the social and among urban areas. Better pricing of infrastructure returns to a high-quality phone system in large urban services-with financing based on user fees-as well as areas are enormous. Of course, the full social returns to private operation of infrastructure services would help phone and other electronic communications systems can limit demands for excessive infrastructure. Most urban be realized only when large numbers of people and busi- infrastructure has local benefits. In such cases, infrastruc- nesses are connected. But the extremely rapid spread of ture should be planned and financed at the local level. private electronic systems wherever governments permit Problems with Land Use Controls them shows that this is a minor limitation. Fortunately, governments have become more tolerant of private Government land use controls exist in all countries phone companies as private cellular phones, fax ma- where private property is permitted. An important dis- chines, e-mail, and the Internet have made it more diffi- tinction is the level of government at which land use cult for them to control information. controls are formulated and administered. In most Arthur Lewis (1978) believed that economic growth countries, provincial and local governments have little should concentrate in small urban areas because low autonomy, so land use controls are formulated at the na- land values make infrastructure much less expensive. tional level even though local governments administer But that idea is without merit. First, infrastructure is not them and may have some discretion as to their detailed necessarily more expensive in large urban areas. High- content. In a few countries, including India and the ways cost more to build per lane per kilometer in large United States, provincial or state governments have some urban areas, but not per person per kilometer traveled, constitutional autonomy and land use controls are at as the roads see heavy use. least partly delegated to local governments. 72 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION As with infrastructure, governments all too often use suburban jurisdictions some land use controls prohibit land use controls to interfere with the urbanization construction of the high-rise apartments or other multi- process. For example, many governments, especially in family dwellings that lower-income citizens can afford. Asia, place severe controls on converting land from rural Evidence from Boston, Vancouver, and some North- to urban uses. In Korea, greenbelts-privately owned ern European cities suggests that land use controls also areas where governments prohibit development-sur- slow the response of housing development to demand, round large urban areas. In high-income countries where destabilizing housing prices. For example, when exces- further urbanization is occurring slowly, the effects of sive regulations mean that developers require months land use controls may be relatively minor. But in devel- or years to obtain permission to build, construction oping countries where the urban population is growing slows and housing prices rise rapidly during periods of at two to four times the national rate, such restrictions high demand. When dwellings finally come on the result in extremely high urban land and housing prices. market, growth has often ceased, and housing prices These excessive values hamper the expansion of the large drop sharply. urban areas on which national growth depends. The economic solution to excessive government con- A comparison between Canada and the United States trols on conversion of land from rural to urban use, and is telling. The two countries have similar incomes, house on types and locations of urban housing, is easy: stop. types, and housing construction and finance sectors. The political problems of mobilizing citizens to urge Both have slowly growing urban areas and large supplies governments to adopt appropriate policies are much of land. Yet the ratio of house values to residents' in- more difficult and beyond the scope of this chapter. comes is 50 percent higher in Canada than in the United States. Canada has much more stringent land use con- Conclusion version controls than does the United States (Malpezzi Large urban areas are among the greatest social inven- and Mayo 1997). tions of all time. They are important and productive parts Why do governments restrict the conversion of land of every modern economy. Governments should stop try- from rural to urban uses? One purported rationale is ing to limit and control them and concentrate instead on that they protect people from adverse effects that private making them more productive. Toward that end, they agreements cannot mediate. Polluting discharges are the should reduce controls on the growth of large urban areas classic example. However, the appropriate policy is to and on the production, location, financing, and types of abate or treat these discharges, not to move them around housing. And they should increase private-sector partici- by controlling land use. The rationale also sometimes pation in creating and operating infrastructure. includes the need to limit pollution, congestion, the costs of infrastructure, and the need to preserve farmland. But Bibliography the growth of large urban areas actually economizes on Barro, Robert. 1996. "Determinants of Economic Growth: A rural land, since population density is greater the larger Cross-Country Empirical Study." NBER Working Paper 5698. the urban area. National Bureau of Economic Research, Cambridge, MA. Governments also justify land use controls and related Becker, Charles, and Andrew Morrison. Forthcoming. "Urban- building codes as necessary to protect people from shoddy ization in Transforming Economies." In Paul Cheshire and dy * . * .* *Edwin Mills, eds., Handbook of Regional and Urban Econom- or dangerous construction. That rationale is unassail- ics Vol. 3: Applied Urban Economics. New York: North- able-provided that the codes are fair and fairly adminis- Holland. tered. The worst abuse of building codes is undoubtedly Becker, Charles, Jeffrey Williamson, and Edwin Mills. 1992. In- the practice in some developing countries of making ille- dian Urbanization and Economic Growth Since 1960. Balti- gal the only housing that the poorest 10-30 percent of more, MD: The Johns Hopkins University Press. people can afford. This practice criminalizes poverty and Eberts, Randall, and Daniel McMillen. 1999. "Agglomeration prevents poor peopl fromlivinginurbanareEconomies and Public Infrastructure." In Paul Cheshire and prevents poor people from living in urban areas. Edwin Mills, eds., Handbook of Regional and Urban Economics One can argue persuasively that minimal land use Vol. 3: Applied Urban Economics. New York: North-Holland. controls-say, those segregating residential neighbor- Fischel, William. 1985. The Economics of Zoning Laws. Balti- hoods from factories that may generate noise and dan- more, MD: Johns Hopkins University Press. gerous road traffic-are good second-best policies. How- Gaspar, Jess, and Edward Glaeser. 1998. "Information Technol- ever, many U.S. urban areas take that claim too far: og, and the Future of Cities." Journal of Urban Economics Restrictions keep virtually all commercial and industrial 43(1): 136-56. Glaeser, Edward. 1997. "Learning in Cities." NBER Working land use apart from high-income suburbs, resulting in Paper 6271. National Bureau of Economic Research, Cam- excessive commuting and other automobile use. In small bridge, MA. THE IMPORTANCE OF LARGE URBAN AREAS-AND GOVERNMENT S ROLES IN FOSTERING THEM 73 Gyourko, Joseph, Matthew Kahn, and Joseph Tracy. 1999. Mills, Edwin. 1998. "Excess Commuting in U.S. Metropolitan "Quality of Life and Environmental Comparisons." In Paul Areas." In Larx Lindquist, Lars-Goran Mattson, and Tschargo Cheshire and Edwin Mills, eds., Handbook of Regional and Kim, eds., Network Infrastructure and the Urban Environment. Urban Economics Vol. 3: Applied Urban Economics. New Berlin: Springer-Verlag. York: North-Holland. Mills, Edwin, and Charles Becker. 1986. Studies in Indian Urban Lewis, W. Arthur. 1978. The Evolution of the International Eco- Development. New York: Oxford University Press. nomic Order. Princeton, NJ: Princeton University Press. Shukla, Vibhotti. 1996. Urbanization and Economic Growth. Malpezzi, Stephen, and Stephen Mayo. 1997. "Housing and Delhi: Oxford University Press. Urban Development Indicators: A Good Idea Whose Time Small, Kenneth. 1992. Urban Transportation Economics. Chue, Has Returned." Real Estate Economics 25(1): 1-11. Switzerland: Harwood Academic Publishers. Mazumdar, Dipak. 1987. "Rural(Urban Migration in Develop- ing Countries." In Edwin Mills, ed., Handbook of Regional Notes and Urban Economics. Vol. 2: Urban Economics. New York: North-Holland. 1. These figures exclude a few countries that are essentially Mills, Edwin. 1998. "The Economic Consequences of a Land city-states. Tax in Land Value Taxation," edited by Dick Netzer. Cam- 2. Adjustments can be made in fuel taxes for fuel burned pre- bridge, MA: Lincoln Institute of Land Policy. dominantly outside large urban areas. See Mills 1998a. chapter 9 Toward a New Model of Rural-Urban Linkages under Globalization Yujiro Hayami Foundation for Advanced Studies on International Development [ q I rbanization in developing economies suggest a new model of rural-urban linkage under has concentrated largely in megacities, globalization. * - ~~~often one per country. The result has been mushroong slm, frmid hab From Colonialism to Import Substitution - _ D~~een mushrooming slums, formidable congestion and pollution, and social, The high concentration of population and economic ac- political, and economic disruption. tivities in developing economies today stems from devel- However, contrary to widespread belief, urbanization opment policies of the 19th century. Indeed, this result is not an inevitable result of globalization. The urban- was the product of globalization. As colonial powers ization process in developing economies today is largely integrated tropical economies into a network of inter- the product of specific development strategies rather national trade, they began to specialize in cash crops than the inevitable consequence of modern technology and minerals, not only because of local comparative ad- and market forces. The integration of domestic with in- vantage in natural resources but also because the colo- ternational markets does stimulate the growth of indus- nial powers used imported capital and labor to exploit trial and service activities. Yet those activities do not them-the "vent-for-surplus" development of HIa Myint have to locate in a metropolis. (1971). In fact, Krugman and Livas Elizondo (1996) have Often colonial powers not only allocated available shown that trade liberalization can promote industrial virgin lands exclusively to colonial planters of export decentralization because it gives entrepreneurs in the crops, but they also transferred arable lands already hinterland access to international markets for products under cultivation by native peasants to white planters. and inputs. A cross-country study by Ades and Glaeser Publicly funded research on agricultural technology (1995) found a negative link between openness to in- concentrated on cash crops grown in plantations, to the ternational trade and concentration of population in a neglect of subsistence food crops. Investments in infra- metropolis. structure such as ports, railways, and roads mainly facil- However, trade liberalization alone is unlikely to pre- itated the transport of cash crops and minerals for ex- vent urban concentration, as these economists imply. port. Peasants shouldered a significant share of the cost Rural-based industrialization in Meiji Japan (1868- of such infrastructure through various taxes and levies 1912) as well as the striking success of balanced rural- such as the poll tax. This taxation forced cashless peas- urban growth in Taiwan, China, after the Second World ants to seek wage employment, expanding the labor sup- War show that globalization through liberalization can ply for colonial enclaves. All these policies increased the be a powerful means of inducing industry to locate out- comparative advantage of cash crops and minerals over side the urban center-if it is accompanied by other im- both subsistence crops and domestically manufactured portant national policies. These historical experiences commodities (Myint 1965; Lewis 1970). 74 TOWARD A NEW MODEL OF RURAL-URBAN LINKAGES UNDER GLOBALIZATION 75 This process produced large port cities, typically one Under import substitution, countries commonly in each colony, linking domestic economies with inter- used tariffs and quotas on imports to protect large-scale national markets, not only for trade but also for finan- modern industries, such as assembly of consumer dura- cial and other services as well as migrating labor. Such bles and capital equipment and manufacture of syn- cities were usually the site of colonial administration. thetic fibers and chemical fertilizers. The tariffs and Strong agglomeration economies worked among all quotas served to raise domestic prices of these products. these services, which included close connections be- Meanwhile, countries also allocated imported capital tween private firms and government offices. Consisting and intermediate goods to those industries, whose costs of colonial government officials, traders, shippers, and were low owing to the overvaluation of domestic cur- financiers, plus a large number of native employees in- rency. The victims of this policy were consumers as well cluding a swarm of coolies, these cities grew to a size in- as unprotected industries, as both were forced to pur- comparably larger than precolonial towns. chase commodities at elevated prices. Rural deindustrialization accompanied the develop- Agriculture and small- and medium-scale industries ment of megacities. Competition from commodities especially suffered (Little et al. 1970). For example, when produced in modern factories in the West typically countries protected synthetic fiber industries, high yarn destroyed production of indigenous manufactures of prices undermined garment industries usually run by farm households and cottage industries for local con- small and medium enterprises using labor-intensive tech- sumption-"Z-goods," according to Hymer and Res- nology. Note that weaving and garment making are nick (1969). The disappearance of Z-good production among the most suitable manufacturing activities for was especially pronounced in countries with vent-for- rural households and cottage industries. Protection of the surplus growth, such as Malaysia, Thailand, and the manufacture of chemical fertilizers had the same negative Philippines (Resnick 1970). Because governments kept effect on farmers. Subsidies for credit and imports such the incomes of native peasants and wage laborers low, as tractors partly compensated large estate farmers, but demand for relatively inexpensive and easy-to-make small farmers and manufacturers suffered from state trad- goods requiring no sophisticated skill to manufacture ing monopolies and prohibitions on the production and did not justify domestic production. Meanwhile, de- processing of plantation crops by smallholders (Sahn and mand from high-income planters, traders, and other Sarris 1994). Rural-based industrialization narrowed service providers in the capital city for "elite commodi- further and industrial activities concentrated among large ties" leaked out of the domestic economy to home urban-based producers. Ironically, a policy intended countries. to break the colonial economic system instead further In the later stage of colonial development, modern strengthened the inequitable and urban-biased system. manufacturing began to respond to rising domestic de- mand stemming from growing population as well as the incomes of peasants who participated in cash crop pro- As the comparison between the globalization-oriented duction despite many regulations against them. However, colonial system and the autarky-oriented import- with commercial, financial, and other services well estab- substitution system has shown, globalization per se does lished in the capital city and rural non-farm entrepreneur- not promote or prevent urban concentration, but rather ship largely destroyed, new manufacturing establishments different policies produce different outcomes. The ques- tended to locate in the metropolis. Rural deindustrializa- tion is what policies can curb pathological urban growth tion did not reverse, and the rural sector continued to and promote an equitable and efficient balance between specialize in the supply of food and export cash crops as rural and urban sectors. The colonial and import-sub- well as labor to the ever-growing urban center. stitution systems, when compared with the historical ex- After World War II, newly independent nations al- periences of Japan and Taiwan, suggest policies to cor- most unanimously made industrialization a top prior- rect their failures. ity. The most popular strategy was import substitution, The opening of Japan to international trade in the geared to protect domestic industries from competition. late 19th century under threat from Admiral Perry's fleet This doctrine was based on strong pessimism regarding produced an outcome very different from that of the the export of primary commodities because of sharp de- colonial system. In fact, nearly four decades of economic clines in their prices during the World Depression. The growth after the Meiji Restoration produced vigorous policy also received wide popular support because it was expansion of rural-based industries. This rural industrial designed to break the inequitable colonial system. growth occurred under virtual free trade and heavy re- 76 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION liance on export of primary and semi-primary com- Table 9.1 modities such as silk and tea.' The sharp contrast be- Annual Growth of Employment in Manufacturing in tween Japan and colonized tropical economies both Taiwan (China) and the Philippines under the liberal trade regime clearly shows that free (Percent) trade alone does not promote industrial decentralization outside the urban center, contrary to the predictions of Rural Urban Krugman and Livas Elizondo (1996). Taiwan In Japan, the government of the newly established 1966-80 4.99 4.94 nation-state actively attempted to boost the productiv- ity of food crops, especially rice, by investing in devel- Philippines opment and extension of new varieties and cultural 1967-75 0.57 1.42 1 975-88 ~2.03 4.81 practices as well as improvements in irrigation and 1975-88 . . . ~~~~~~~~~~~~~~Source: Ranis and Stewart (1993, p. 94). drainage facilities. Major increases in rice yields per Source:_Ranis_and_Stewart__1993__p__94__ hectare outpaced those elsewhere. The new technology was designed to save both land and capital, and it was also largely neutral with respect to operational scale-if As in the case of Meiji Japan, the key to Taiwan's suc- anything, it favored family farms (Hayami and Yamada cess with rural-based industrialization was public sup- 1991). Small farm producers spent a significant share of port in the form of agricultural research, extension, and their rising incomes on locally produced goods and farm irrigation services for smallholders who provided strong machinery and implements crafted by village black- linkages to local industries, trade, and other services (Ho smiths and local town workshops. 1977, 1979). Public support for improved rice farming Peasant agriculture not only provided strong back- began under the Japanese colonial regime and strength- ward linkages to local manufacturers but also provided ened after World War II under the Kuomintang (Nation- forward linkages by increasing the supply of cash crops alist Party). This support produced sustained increases for industrial processing, such as silk and tea. Unlike in in rice yields beginning in the mid-i 920s (except for tropical economies, smallholders instead of large plan- wartime disruption). In the Philippines, where colonial tations produced those cash crops in Japan. Hand reel- development efforts concentrated on plantation crops, ing of silk continued to be a major cottage industry even rice yields did not grow significandy until the 1970s.2 after World War I, coexisting with factories using mod- ern machinery that were also located mainly in rural tructuring Linkages between Rural Producers . . ~~~~~~~~~and Urban Trade Centers towns. Developments in commerce, transport, storage, and other services-also mostly rural-based and small- The problem with import-substitution industrialization, scale, using labor-intensive organization and technol- as practiced widely for almost three decades after World ogy-supported these intense links between agriculture War II, is that both farmers and small-scale manufactur- and manufacturing (Kawagoe 1998; Itoh and Tanimoto ers were taxed in a highly distortive manner without 1998). The simultaneous development of industrial and compensating public support for their activities. Large service activities based on rising productivity in small- implicit taxes on small producers were dissipated for bu- holder agriculture greatly increased the rural sector's reaucratic consumption as well as to support high-cost ability to hold population. protected industries, including the wages of unionized Postwar industrialization in Taiwan, China, followed labor. The policy thus choked off the possibility of na- a similar pattern with respect to parallel growth of rural tional development through labor-intensive activities by and urban enterprises. Indeed, during the extraordinar- small producers, which have comparative advantage in ily rapid industrial growth from 1956 to 1980, rural- labor-abundant and capital-scarce developing economies. and urban-based manufacturing employment grew at al- The inefficient and inequitable nature of import- most exactly the same rate. This experience represented substitution industrialization became increasingly visi- a sharp contrast with the Philippines, where the growth ble throughout the 1960s and 1970s. Then the eco- of manufacturing employment in rural areas was much nomic crisis in developing economies resulting from the less than half that in urban areas (Table 9.1). Strikingly, 1981 collapse of the second oil boom, together with Taiwan, China, achieved not only a good interregional pressure from the International Monetary Fund (IMF) balance but also much faster overall growth, suggesting and World Bank, forced them to adopt structural ad- that rural-urban balance and a high rate of economic justments designed to liberalize their markets. The suc- growth can be complementary. cess of the new policy in promoting economic growth TOWARD A NEW MODEL OF RUR-AL-URBAN LINKAGES UNDER GLOBALIZATION 77 through labor-intensive production was especially re- be punished socially through malicious gossips andos markable in economies such as Indonesia and Thailand, tracism in the small rural community.. Once community which were important components of the East Asian sanction reduced opportunism, small-scale, family- economic miracle (World Bank 1993). In Indonesia, for based rural enterprises became more efficient than large example, currency devaluations and deregulation of firms because of the farmer's advantage in monitoring trade and foreign direct investment during the 1980s labor at a lower cost in light industries with weak scale sparked rapid increases in the production and export of economies (Hemi 1997). labor-intensive manufactures-as well as sharp decreases Japanese automakers' subcontracts with parts suppli- in poverty. These effects ended only with the recent fi- ers are a prime example of the application of relational nancial crisis. contracting to capital-intensive industries. Kiichiro Toy- But although market liberalization and deregulation ota, founder of the Toyota Motor Company, intended opened the door to the effective utilization of abundant in the 1930s to build "a pastoral factory" in Korma labor in response to rising demand for labor-intensive (today's Toyota City)-then a rural town with few man- commodities from high-income economies, such pro- ufacturing activities (Wada 1998). Kiichiro's idea was duction located disproportionately within major cities not only to locate the factory in a pastoral environment or their outskirts-with severe accompanying conges- but also to surround it with trusted parts suppliers con- tion and pollution. This situation is different from the nected by community spirit. Toyota initially tried to experiences of Meiji Japan and postwar Taiwan, China, contract with manufacturers receptive to Toyota's guid- which developed rural towns and villages as the location ance, even if they had no previous experience in the pre- of industrial production for export. cise, sophisticated work of automobile manufacturing. Critical to enabling rural-based industries to partici- Kiichiro's idea apparently stemmed from his experience pate in global economic activities is the trade network as a supplier of automatic looms for rural-based weav- linking small producers to international demand, usu- ing entrepreneurs. His idea bears fruit today as Toyota's ally transmitted through urban centers. In Meiji Japan, highly efficient modern subcontracting system known if an urban trader received an order of cotton or silk as Kanab (just-in-time). cloth for export, for example, he usually made contracts The network of relational contracts between urban with local collectors in rural areas to assemble the traders, local collectors, and rural manufacturers was an needed amounts from a large number of small weavers innovation designed to organize local labor to produce living in their places. In order to meet the export de- commodities to meet large-scale national and inter- mand, a large amount of products to meet a certain national demand. Indeed, this network of long-term, quality standard specified by a foreign buyer must be multistandard contracts tied by mutual trust across urban collected by a specified delivery date. If his collectors vi- principals and rural agents substituted for intensive in- olated the contracts by mistake or opportunism, the formation flows in urban agglomeration economies. If trader was bound to lose reputation among his cus- such a system had not proved effective, traders might tomers, hence, lose future business opportunity in ex- have opted to build and manage factories for mass- port trade. Thus, he normally endeavored to establish a producing standardized commodities in an urban loca- mutual trust relationship with collectors through re- tion, even if manufacturing costs were lower in the hin- peated transactions over time as well as interlinking terla-nd. The effective use of a rural-urban trade network commodity trades with credit and other transactions so to fully exploit the opportunities for rural-based devel- as to avoid the priso ners' dilemma game solution. opment created by globalization rested upon the active The same applied for local collectors in relation with participation of rural entrepreneurs. Thus the links be- weavers operating in rural villages and towns. Typically tween industrial producers in the hinterland and interna- a local collector developed putting-out contracts with tional markets under liberalized trade assumed in the weavers by advancing yarn and collecting cloth at a pre- Krugman-Livas Elizondo model depend critically on a scribed piece rate. Long-term, repeated contracts were domestic trade network. 3 usually applied to suppress opportunistic behaviors such Such an active role of rural entrepreneurs contrasts as embezzlement of yarn by means of weaving cloth at with the traditional image of indigenous people, espe- lower density than agreed upon. For the same reason, cially peasants subsisting in rural villages, as inherently personal connections such as relatives, friends, and passive, with neither the desire nor the ability to partic- neighbors were preferred in the selection of contracting ipate in commercial or industrial activities. This view partners. Opportunism was suppressed significantly by found typical expression in Clifford Geertz's anthropo- the expectation that possible contract violation would logical study in Indonesia. He concluded that entrepre- 78 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION neurship in the nonfarm business activities that induce 15,000 spindles or more using cotton imported from social modernization cannot emerge from "the immedi- India and the United States. If Japan had been able to ate purview of village social structure" but is limited to set its own tariffs, the government would likely have populations with "extra-village status," such as ethnic protected inefficient cotton-spinning mills as well as Arab traders in East Java and traditional rulers in Bali farmers by imposing high tariffs on imported cotton (Geertz 1963). and yarn and would have lost the opportunity for cotton However, my observations of agricultural marketing spinning to become a major Japanese export industry. in Indonesia reveal local marketing systems consisting Indeed, the dramatic development of the Japanese cot- of petty traders and processors in rural villages who use ton textile industry-which soon outcompeted Britain scarce capital efficiently and local resources intensively, and India-occurred with virtually no government pro- especially labor, while developing adequate linkages tection and subsidy. This experience is clearly inconsis- with urban traders and manufacturers (Hayami and tent with the assertion that government subsidies are in- Kawagoe 1993). The entrepreneurs who organize this dispensable for late-industrializing economies hoping system use community relationships to ingeniously de- to foster export manufactures, including textiles (Ams- sign trade practices and contracts for sharing risk and den 1989). minimizing transaction costs. Rural areas of the Philip- Foreign direct investment can also be an important pines and Thailand reveal similar entrepreneurship in support for small and medium-scale rural industries. the garment and weaving industries (Kikuchi 1998; Taiwan, China, illustrates this relationship most clearly. Ohno and Jirapatpimol 1998), as do rural communities Unlike the Republic of Korea, Taiwan, China, liberal- in India (Bhatt 1998). Somewhat similar organizations ized foreign direct investment relatively early, and small are emerging in China, with "township-village enter- and medium-sized enterprises contracted with foreign prises" developing linkages with state or foreign enter- firms, which offered trade credits and technical guid- prises through trade contracts, including financial sup- ance. The former developed into the dominant export port and technical guidance (Liu and Otsuka 1998; sector, with firms with less than 300 employees account- Otsuka et al. 1998). ing for as much as 65 percent of manufactured exports The major question is thus not whether the potential during the mid-1980s (Wade 1990). Compared with for rural entrepreneurship exists, but rather how devel- Korea, Taiwan's industrialization produced different sizes oping countries can fully exploit it to achieve balanced of manufacturing activities as well as better urban-rural rural-urban growth under globalization. balance (Ho 1979, 1982). Of course, free trade and direct foreign investment alone do not assure the success of rural-based industrial- How can countries intensify farm-nonfarm linkages at ization. Otherwise, tropical economies under the colo- the local level and rural-urban linkages at the national nial system would have achieved industrial development level along the historical paths of Japan and Taiwan? similar to that of Meiji Japan. Part of the difference Liberalization of both international and internal mar- stemmed from Japanese government support for small- kets has clearly been indispensable for promoting the ac- holder agriculture, but equally significant was public tivities of small farmers, manufacturers, and traders, as support for small manufacturers and traders. Japan ac- illustrated by an episode during the early Meiji period tively pursued applied industrial research, development, when Japan was deprived of tariff autonomy as a part of and extension and provided information services such as unequal treaties with Western powers (Hayami 1997). industrial fairs, exhibitions of foreign products and ma- In the 1870s and 1880s the Japanese government chines, and newsletters from consulates on overseas mar- tried to develop rural cotton-spinning industries to res- kets. The government also encouraged small producers cue cotton producers about to succumb to imports of to form industrial associations to distribute technical cotton yarn and cloth. The government imported rel- and market information and organize product grading atively small-scale plants with 2,000 spindles for lease and quality inspection (Kiyokawa 1995; Minami et al. or sale at subsidized prices to entrepreneurs in inland 1995). Under the colonial system assistance to domestic cotton-farming areas. However, factories equipped with manufacturers was not only absent but often suppressed. these small machines failed to produce yarn at com- Public infrastructure such as roads, communication petitive prices from high-priced domestic cotton. But systems, and rural electrification is also key to widely cotton spinning soon after developed into a major ex- diffusing industrial and commercial activities in rural port industry as private entrepreneurs established large areas with strong urban links. Even more basic is pro- modern factories equipped with machines composed of viding education to rural residents, including primary TOWARD A NEW MODEL OF RURAL-URBAN LINKAGES UNDER GLOBALIZATION 79 and midlevel vocational education. Taiwan faired excep- economies achieved by producers using steam engines, tionally well in providing such supports to rural people thereby strengthening their competitive position vis-a- (Ranis and Stewart 1993). vis large-scale urban factories (Minami 1976). In terms of remarkable growth in township-village Another legitimate question is whether the simultane- enterprises, the development of postreform China may ous development of smallholder agriculture, rural manu- appear analogous to the historical experience of Japan facturing, and services in Japan and Taiwan stemmed and Taiwan. However, this development is rather spe- from relatively egalitarian distribution of farmland. After cial as it has been supported by strong regulation on all, rice-growing villages consisted of stable and relatively labor migration from rural to urban areas, in addition homogeneous family farms, providing a nursery for con- to liberalization in commodity trade and foreign direct tracting based on mutual trust and cooperation and in- investment (Byrd and Lin 1990; Lin et al. 1996). It ap- tensive local interaction. If so, developing economies pears that China will need a major expansion in rural with highly skewed land distribution, such as the Philip- infrastructure, especially for transportation, to further pines and Latin American countries, may not be able to rural industrialization at the speed to prevent the rural- follow a similar development path without drastic land urban disparity from widening. reform. However, less severe measures can help redistribute land. In Japan, unequal distribution of land in the early Rural-based industrialization in Meiji Japan depended 16th-century Tokugawa period improved with the devel- on relatively well-developed roads and irrigation systems opment of labor-intensive technology and previously inherited from the feudal Tokugawa period,4 and post- prohibited land-rental contracts (Smith 1959). This war development in Taiwan depended on public infra- process continued in the Meiji period. Large, wealthy structure built under the Japanese colonial regime farmers monopolized cocoon production until the early (Ranis and Stewart 1993). One might question whether Meiji period, for example. But with technological inno- low-income economies today, especially those in sub- vations such as hybrid silkworms and summer-fall rear- Saharan Africa, can replicate this experience without ing practices, together with institutional innovations such such a favorable initial condition. Indeed, it will be ex- as cooperative rearing of young worms, small farmers tremely difficult to mobilize enough public resources to could profitably pursue sericulture (Nghiep and Hayami support numerous small farmers, manufacturers, and 1991). traders in the vast hinterland in economies with much Developing economies today could similarly rely on lower population density than in Asia. technological and institutional innovations to correct However, one great advantage to developing econ- great inequality in land distribution. In fact, significant omies today is the availability of advanced electronic scope exists for reorganizing large plantations into small communications. The use of personal computers backed family farms. The plantation system was originally es- by satellite systems can enable small rural producers to tablished under the colonial regime to exploit tropical integrate their activities with national and international lands through export crops. Opening new lands for ex- markets, given relatively modest investment. And e-mail port production required huge capital outlays for infra- and cellular phones can facilitate formation of relational structure such as roads and docking facilities, and farms contracts linking rural manufacturers with urban traders. had to be large for investors to gain from such invest- For rural producers, the new communications technol- ments. However, as population density and infrastruc- ogy can thus substitute for urban agglomeration econ- ture grew, the scale diseconomies of centrally managed omies based on intensive information flows. estate farms began to emerge, much as with collective Technological advance can also help supply other in- farms in socialist economies. frastructure. Local power generation based on solar, Because agricultural production is performed over wind, and biomass energies can greatly facilitate rural wide space and subject to the vagaries of nature, moni- electrification, for example. Such locally generated elec- toring wage laborers is inherently difficult. This makes tricity could eliminate the need for large-scale, high-cost family farming most efficient at the field level (Bins- systems for transmitting power. The combined effect of wanger and Rosenzweig 1986; Hayami and Otsuka advances in communication and power generation tech- 1993). Today the merit of the plantation lies mainly in nologies may be similar to the situation in Meiji Japan, its ability to coordinate the supply of agricultural raw which benefited from the development of small electric materials (such as sugar cane) with large-scale process- motors and hydrobased electrification. Small rural work- ing (using a centrifugal sugar mill). However, efficiency shops relied on those technologies to offset the scale can rise if an agribusiness manages processing and mar- 8o LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION keting and provides technical guidance, credit, and trepreneurs demonstrated their latent entrepreneurship other services to smallholders in return for their pledged by pursuing market transactions. Rural villages and supply of farm products. The principle of this contract towns were progressively integrated with national and farming system is similar to that of traders (or a larger international markets, especially after the country liber- factory) who rely on relational contracts to organize a alized trade and devalued its currency in the 1980s. If network of small manufacturers. peasants in Indonesia as well as other Asian countries Production of sugar and canned pineapple in Thai- considered totally incapable of market activity only a few land based on such a system has recently been out- decades ago could follow the same path as Meiji Japan, competing Filipino production based on the plantation is there any reason their African counterparts could not system. Thailand has supported such production by in- follow a similar path, given stronger national and inter- vesting in rural infrastructure, especially roads. In fact, national support? inducing plantations to reorganize into contract farm- ing by providing public infrastructure and services, as well as by reducing favorable tax treatments and access Today globalization supported by communications tech- to public lands for plantations, will likely prove more ef- nology makes possible rural-based developments in com- ficient than coercive land redistribution (Hayami et al. merce and industry as well as agriculture in developing 1990; Hayami 1994). economies. Market liberalization has emancipated small Unlike in developing Asia, where comparative advan- farmers, manufacturers, and traders from the yoke of dis- tage stems from labor-intensive manufacturing such as criminatory regulation against their activities, so perva- textile and footwear, relatively land-abundant econo- sive under the import-substitution strategy. The ability mies in Africa and Latin Africa are likely to find com- of rural producers to organize production in response to parative advantage in high-valued agroindustry prod- international demand has recently become apparent in ucts. High-income economies have accelerated their Latin American countries such as Chile, which success- demand for such items, including exotic tropical fruits, fully produced high-value agroindustry goods for export. cut flowers, wines, and leather goods. Contract farming If the present momentum strengthens with adequate has proved effective in organizing small rural producers supports, it could reverse the growing imbalance be- to grow such high-valued commodities (Braun et al. tween rural and urban sectors. The needed supports in- 1989; Hayami and Kawagoe 1993). dude general and vocational education, applied research Transferring the Japan-Taiwan model to economies and extension for both agriculture and local manufactur- in sub-Saharan Africa may seem impossible given eco- ing, roads and communication infrastructure, and rural nomic, social, and cultural barriers (Platteau and Hayami electrification, among others. 1998). However, it was not too long ago that a distin- And the budget necessary for providing enough local guished anthropologist, Clifford Geertz, expressed dire public goods to achieve rural-urban balanced growth pessimism about the entrepreneurship of Indonesian will be large, even given significant cost savings from peasants in commerce and industry. Such a perspective technological innovation. The need may well exceed the was widespread among scholars on tropical Asia before finances of low-income developing economies where the the East Asian miracle. The pessimism regarding Asian pathology of urban growth is especially severe given high peasantry, including Indonesia, did not stand up to his- population growth and where local infrastructure has torical test. been grossly underdeveloped, such as in sub-Saharan Yet Geertz's representation of Indonesian peasants Africa. may actually have been valid in the early 1950s. At that The challenge is to make efficient use of scarce pub- time the Indonesian economy had not yet recovered. lic resources. Relying on the central government to al- from the devastation of Japanese occupation followed by locate such resources to remote hinterlands will prove a war of independence against the Dutch. The rural sec- highly inefficient. Decentralizing decisionmaking to tor was cut off from a wide market and peasants' activi- the prefecture level and further to the township and vil- ties were of necessity subsistence-oriented. In the four lage levels will be necessary. Municipal governments decades since then until the outbreak of the recent fi- and local industry associations can run experiment sta- nancial crisis, the country had established law and order, tions and extension services much more efficiently than improved transportation and communication systems, a central government. In fact, Japan has traditionally and widely diffused primary education as well as the pursued such a decentralized strategy (Itoh and Urata official Indonesian language, breaking down barriers 1994). Appropriate coordination between central and among different regions and ethnic groups. Peasant en- local governments is the major challenge facing devel- TOWARD A NEW MODEL OF RURAL-URBAN LINKAGES UNDER GLOBALIZATION 8i oping economies hoping to achieve balanced rural- Binswanger, Hans P., and Mark R. Rosenzweig. 1986. "Behav- urban growth. ioral and Material Determinants of Production Relations in In this connection, serious thought must be given to Agriculture." Journal of Development Studies 22(3): 503-39. allowing rural people, especially farmers, to share the Braun, Joachim von, David Hotchikiss, and Maarten Immink. allowing r ural people, especially farmers, to share the 1989. "Nontraditional Export Crops in Guatemala: Effects cost of developing local infrastructure and public ser- on Production, Income, and Nutrition." IFPRI Research vices. While postwar Taiwan, China, and Meiji Japan Report No. 73. Washington D.C.: International Food Policy provided public support to farmers, those in Japan Research Institute. shouldered high land taxes, and farmers in Taiwan, Byrd, William A., and Qingsong Lin, eds. 1990. China's Rural China, had to exchange rice for fertilizers under state Industry: Structure, Development, and Reform. Oxford: Oxford monopoly at a barter ratio much more unfavorable than University' Press. Geertz, Clifford. 1963. Peddlers and Princes. Chicago: The Uni- the prevailing international price. Taxation of farmers is versity of Chicago Press. not inherently antiagricultural or antirural provided that Hayami, Yujiro. 1994. "Peasant and Plantation in Asia." In it does not unduly distort market incentives and that at G.M. Meier, ed., From Classical Economics to Development least some of the taxes return to farmers in the form of Economics, pp. 121-34. New York: St. Martin's Press. infrastructure and services expected to produce high Hayami, Yujiro. 1996. "The Peasant in Economic Moderni- economic ru .zation." American Journal of Agricultural Economics 78(5): economic returns. The important considerations are de- 1157-67. signing the least distortive tax scheme and concentrat- Hayami, Yujiro. 1997. Development Economics: From the Poverty ing scarce public funds to best support rural producers. to the Wealth of Nations. Oxford: Oxford University Press. In the past, so many distortive measures have been Hayami, Yujiro, ed. 1998. Toward the Rural-Based Development used to tax rural producers, and scarce public resources of Commerce and Industry: Selected Experiencesfrom East Asia. were often used not to support but rather to suppress Washington, D.C.: World Bank. Hayami, Yujiro, and Toshihiko Kawagoe. 1993. The Agrarian their activities. Such policies have partly been the result Origins of Commerce and Industry: A Study of Peasants Mar- of the political elite's rent-seeking activities, but they keting in Indonesia. New York: St. Martin's Press. were also adopted as the result of policymakers' igno- Hayami, Yujiro, and Keijiro Otsuka. 1993. The Economics of rance and distrust of the ability of small informal pro- Contract Choice: An Agrarian Perspective. Oxford: Clarendon ducers in the countryside. In designing appropriate poli- Press. rural-urban balanced growth in the 2 st Hayami, Yujiro, and Vernon W. Ruttan. 1985. Agricultural cies to promote Development: An International Perspective, rev. ed. Baltimore: century, we must keep in mind the following remark: Johns Hopkins University. Hayami, Yujiro, Agnes R. Quisumbing, and Lourdes A. Adri- Most governments under-invest in this infrastruc- ano. 1990. Toward an Alternative Land Reform Paradigm: A ture and over-invest in their own enterprises and Philippine Perspective. Quezon City, Philippines: Ateneo de efforts to carry out agricultural marketing directly. Manila University Press. Hayami, Yujiro, and Saburo Yamada, with associates. 1991. The If policymakers trusted their peasants more and Agricultural Development ofa feared middlemen less, this government bias could Press. be reversed. Most policymakers have never met a Ho, Samuel P. S. 1977. Economic Development of Taiwan, peasant, much less one with entrepreneurial skills 1860-1970. New Haven, CT: Yale University Press. engaged in risky trading activities (Timmer 1993). Ho, Samuel P. S. 1979. "Decentralized Industrialization and Rural Development: Evidence from Taiwan." Economic De- velopment and Cultural Change 28(1): 77-96. Surely this statement applies not only to peasants but to Ho, Samuel P. S. 1982. "Economic Development and Rural In- all informal rural entrepreneurs engaging in commerce dustry in South Korea and Taiwan." World Development and industry in developing economies. 10(11): 973-90. Hymer, Stephen, and Stephen Resnick. 1969. "A Model of an Bilbliography Agrarian Economy with Nonagricultural Activities." Ameri- can Economic Review 59(4): 493-506. Ades, Alberto F., and Edward L. Glaeser. 1995. "Trade and Cir- Ishikawa, Shigeru. 1967. Economic Development in Asian Perspec- cuses: Explaining Urban Giants." Quarterly Journal of Eco- tive. Tokyo: Kinokuniya. nomics 110(1): 195-227. Itoh, Motoshige, and Masayuki Tanimoto. 1998. "Rural Entre- Amsden, Alice. 1989. Asia's Next Giant: South Korea and Late preneurs in the Cotton-Weaving Industry of Japan." In Industrialization. New York: Oxford University Press. Yujiro Hayami, ed., Toward the Rural-Based Development of Bhatt, V. V. 1998. "On the Relevance of East Asian Experience: Commerce and Industry: Selected Experiences from East Asia, A South Asian Perspective." In Yujiro Hayami, ed., Toward pp. 47-68. Washington, D.C.: World Bank. the Rural-Based Development of Commerce and Industry: Itoh, Motoshige, and Shujiro Urata. 1994. "Small and Medium- Selected Experiencesfrom East Asia, pp. 267-9 1. Washington, size Enterprise Support Policies in Japan." Policy Research D.C.: World Bank. Working Paper 1404. Washington D.C.: World Bank. 82 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Kawagoe, Toshihiko. 1998. "Technical and Institutional In- Prebisch, Raul. 1959. "Commercial Policy in the Underdevel- novations in Rice Marketing in Japan." In Yujiro Hayami, oped Countries." American Economic Review 49(Supple- ed., Toward the Rural-Based Development of Commerce and ment): 251-73. Industry: Selected Experiences from East Asia, pp. 23-46. Ranis, Gustav and Frances Stewart. 1993. "Rural Nonagricul- Washington, D.C.: World Bank. tural Activities in Development: Theory and Application." Kikuchi, Masao. 1998. "Export-Oriented Garment Industries in Journal of Development Economics 40(1): 75-10 1. the Rural Philippines." In Yujiro Hayami, ed., Toward the Resnick, Stephen. 1970. "The Decline in Rural Industry under Rural-Based Development of Commerce and Industry: Selected Export Expansion: A Comparison among Burma, Philippines Experiencesffrom EastAsia, pp. 89-129. Washington, D.C.: and Thailand, 1870-1938." Journal of Economic History World Bank. 30(1): 51-73. Kiyokawa, Yukihiko. 1995. Nihon no Keizai Hatten to Gijutsu Sahn, David E., and Alexander Sarris. 1994. "The Evolution of Fukyu (Economic Development and Technological Diffu- States, Markets, and Civil Institutions in Rural Africa." Jour- sion). Tokyo: Toyokeizai Shimposha. nal of Modern African Studies 32(2): 279-303. Krugman, Paul, and Raul Livas Elizondo. 1996. "Trade Policy Smith, Thomas C. 1956. "Landlords and Rural Capitalists in the and the Third World Metropolis." Journal of Development Modernization of Japan." Journal of Economic History 16(2): Economics 49(1): 137-50. 165-68. Lewis, W. Arthur, ed. 1970. Tropical Development, 1880-1913: Smith, Thomas C. 1959. The Agrarian Origins of Modern Japan. Studies in Economic Progress. London: Allen & Unwin. Stanford, CA: Stanford University Press. Lin, Jutin Yifu, Cai Fang, and Li Zhou. 1996. The China Mira- Smith, Thomas C. 1988. Native Sources ofJapanese Industrial- cle: Development Strategy and Economic Reform. Hong Kong: ization, 1750-1920. Berkeley and Los Angeles: University of The Chinese University Press. California Press. Little, Ian M. D., Tibor Scitovsky, and Maurice Scott. 1970. Timmer, C. Peter. 1993. In Yujiro Hayami and T. Kawagoe, Industry and Trade in Some Developing Countries. Oxford: Agrarian Origins of Commerce and Industry, foreword. New Oxford University Press. York: St. Martin's Press. Liu, Dequiang, and Keijiro Otsuka. 1998. "Township-Village Wada, Kazuo. 1998. "The Formation of Toyota's Relationship Enterprises in the Garment Sector of China." In Yujiro with Suppliers: A Modern Application of the Community Hayami, ed., Toward the Rural-Based Development of Com- Mechanism." In Yujiro Hayami, ed., Toward the Rural-Based merce and Industry: Selected Experiences from East Asia, Development of Commerce and Industry: Selected Experiences pp. 161-86. Washington, D.C.: World Bank. from EastAsia, pp. 69-86. Washington, D.C.: World Bank. Minami, Ryoshin. 1976. Doryoku Kakumai to Gijutsushimpo Wade, Robert. 1990. Governing the Market: Economic Theory (Power Revolution and Technological Progress). Tokyo: and the Role of Government in East Asian Industrialization. Toyokeizai Shimposha. Princeton: Princeton University Press. Minami, Ryoshin, Kwan S. Kim, Fumio Makino, and Joung-hae World Bank. 1993. The East Asian Miracle. Oxford: Oxford Seo. 1995. Acquiring, Adapting and Developing Technologies: University Press. Lessons from the Japanese Experience. London: Macmillan. Myint, Hla. 1965. The Economics of the Developing Countries. Notes New York: Praeger. 1. Under the unequal treaties closed between the Tokugawa Myint, Hla. 1971. Economic Theory and the Underdeveloped Shogunate and Western powers (first with the United States in Countries. New York: Oxford University Press. 1858), all Japanese tariff rates were fixed at 5 percent ad valorem. Nghiep, Le Thang, and Yujiro Hayami. 1991. "The Tradeoff be- Japan was "semicolonized" in a fashion similar to China in the tween Food and Industrial Crops: Summer-Fall Rearing of sense that it lost tariff autonomy and was forced to concede Cocoons." In Yujiro Hayami and S. Yamada, with associates. Western control on legal affairs over certain areas of port cities. The Agricultural Development ofJapan, pp. 175-98. Tokyo: From 1899 Japan was allowed to change tariff rates under the University of Tokyo Press. condition that it obtain Western nations' consent. Not until Ohno, Akihiko, and Benja Jirapatpimol. 1998. "The Rural 1911 was tariff autonomy fully restored. Thereafter, theJapanese Garment and Weaving Industries in Northern Thailand." In government began to promote heavy and chemical industries by Yujiro Hayami, ed., Toward the Rural-Based Development of means of border protection, resulting in a reversal of rural-based Commerce and Industry: Selected Experiences from East Asia, industrialization centering on light industries. pp. 131-59. Washington, D.C.: World Bank. 2. The collaborative efforts of the International Rice Research Otsuka, Keijiro, Degiang Lia, and Naoki Murakami. 1998. In- Institute and national research/extension systems resulted in the dustrial Reform in China: Past Performance and Future successful development and diffusion of modern high-yielding Prospects. Oxford: Oxford University Press. varieties of rice, first in the Philippines, followed by other Pingali, Prabhu, Mahabub Hossain, and Roberta V. Gerpacio. economies in Southeast Asia as well as in South Asia (Hayami 1997. Asian Rice Bowls: The Returning Crisis? Wallingford, and Ruttan 1985). There is little doubt that this so-called "Green U.K.: CBS International. Revolution" in the 1970s to 1980s was an important support Platteau, Jean-Philippe, and Yujiro Hayami. 1998. "Resource to the "economic miracle" in Southeast Asia. It is disquieting Endowments and Agricultural Development: Africa versus to observe that the recent economic crisis in Southeast Asian Asia." In Yujiro Hayami and Masahiko Aoki, eds., The Insti- economies was shortly preceded by significant deceleration in the tutional Foundation of East Asian Economic Development, growth in rice yield, presumably because of exhaustion of yield pp. 357-410. New York: St Martin's Press. potential opened up by the Green Revolution (Pingali et al. TOWARD A NEW MODEL OF RURAL-URBAN LINKAGES UNDER GLOBALIZATION 83 1997). One important condition for revitalizing those economies a megaport city such as Bangkok in Thailand and Manila in the appears to be further intensification of agricultural research in- Philippines. In such a case the prices of imported commodities volving not only the increased allocation of public funds to agri- are higher and those of domestic commodities for export are cultural research but also institutional innovations in the coordi- lower in the hinterland than in the center. Also, it is totally unre- nation between national and international research institutions alistic to assume that international firms can direcdy serve a large as well as between public and private research. number of small informal producers in the hinterland. 3. In the Krugman-Livas Elizondo model, international firms 4. During the three centuries of peace under the hegemony of are assumed to serve domestic producers in the urban center and Tokugawa Shogunate, which saw growing population pressures, the hinterland equally well, so that free trade establishes equality villages in Japan continued to organize communal work projects in product and input prices between the two locations. Such an and stipulated regulations for conserving and improving local assumption might be a valid representation for the relationship commons, such as forests, grazing lands, irrigation systems, and between Mexico City and the export-processing zones on the roads (Hayami 1997). The irrigation infrastructure, already well country's northern border with the United States. However, that established by the end of the Tokugawa era, was an important basis model is irrelevant and unrealistic for most developing economies for the development and diffusion of high-yielding rice varieties where domestic producers link with foreign markets only through in the Meiji period (Ishikawa 1967; Hayami and Ruttan 1985). chapter 10 Global City-Regions and the New World System Allen J. Scott Department of Policy Studies and Department of Geography University of California, Los Angeles [ q ontrary to many recent predictions, fairs are intricately bound up with far-flung and intensi- geography is not about to disappear fying extranational relationships. I shall refer to these ex- (cf. O'Brien 1992). Even in a global- tranational relationships as a symptom of "globalization," izing world, geography does not though my rather casual use of this term admittedly does become less important. It becomes injustice to many conceptual niceties and debates (cf. more important-because globaliza- Hirst and Thompson 1996; Krugman 1996). In parallel tion tends to heighten geographic differentiation and lo- with these developments, global city-regions are consoli- cational specialization. In fact, a worldwide archipelago, dating into political-geographic units as contiguous local or mosaic, of diverse city-regions has been emerging governments form coalitions to deal with both the forcefully since the end of the 1970s. These city-regions threats and the opportunities of globalization. Far from are beginning to function as the basic pillars of the new being dissolved away as definite geographic entities by world system that has been taking shape in recent decades globalization, city-regions are by and large thriving, and (Scott 1998; Veltz 1996). The internal and external rela- if anything becoming more central to the conduct and tionships of these peculiar agglomerations and their com- coordination of modern life (Taylor 2000). plex dynamics present extraordinarily perplexing chal- We can provisionally identify existing global city- lenges to researchers and policymakers alike as we enter regions by referring to a world map of large metropoli- the 21st century. tan areas (Figure 10.1). The figure plainly shows that There is an extensive literature on "world cities" and large-scale urbanization is of major importance in "global cities" by authors such as Castells (1996), Fried- today's world, and that it is characteristic of both eco- mann and Wolff (1982), Hall (1966), Knox (1995), and nomically advanced and economically developing coun- Sassen (1991), to name only a few. This literature fo- tries. In fact, contrary to numerous prognostications, cuses above all on the large cosmopolitan city as a cen- large cities all over the globe continue to grow in size. ter of advanced services and information flows and as a In 1950 there were 83 cities in the world with popula- command post for multinational corporations. I pro- tions of more than 1 million (two-thirds of them in the pose to use the term global city-regions to capture similar economically advanced countries). In 1990 there were phenomena while extending the term's meaning to 272 such cities (two-thirds of them in economically de- include a wider range of contextual characteristics and veloping countries). The United Nations' assessment of social implications. population trends in the world's 30 largest metropoli- In simple geographic terms, a global city-region can tan areas indicates that this growth will continue over at be said to comprise any major metropolitan area, or con- least the next few decades (Table 10.1). tiguous set of metropolitan areas, together with a sur- However, not all large metropolitan areas are equally rounding hinterland-itself a locus of scattered urban caught up in processes of globalization, and not all settlements-whose internal economic and political af- global city-regions can be equated with existing large 84 GLOBAL CITY-REGIONS AND THE NEW WORLD SYSTEM 85 Figure 10.1 World Distribution of Metropolitan Areas with Populations Greater than One Million w ~~~~~~~~~~~~~~~~~~~~~~1 + Source: United Nations 1995. metropolitan areas. A more effective understanding of world system and a new social grammar of space (Badie global city-regions calls for an inquiry into what we 1995). One of the outstanding features of this emerg- might call a new regionalism in contemporary capitalism. ing condition is the apparent though still inchoate for- mation of a hierarchy of economic and political institu- Gl:obalization and tbhe NVew R2egionalismz tin,fo loaolcl tions, from global to local. In the immediate post-World War II decades, almost all Four main aspects of this state of affairs call for im- the major capitalist countries were marked by strong cen- mediate attention: tral governments and relatively tightly bordered national economies. These countries constituted a political bloc 1. Huge and growing amounts of economic activity within a Pax Americana underpinned by a rudimentary occur in the form of cross-border relationships, including network of international institutions and arrangements, input-output chains, migration streams, foreign direct in- including the World Bank, the International Monetary vestment by multinationals, and monetary flows. Such ac- Fund (IMF), and the General Agreement on Tariffs and tivity is in large degree what I mean by globalization, Trade (GATT), through which they sought to regulate though it remains far from any ultimate point of fulfill- their relatively limited-but rapidly expanding-eco- ment. As globalization moves forward, it creates conflicts nomic interrelations. Over much of the postwar period and predicaments that prompt a variety of political re- the most prosperous of these countries constituted a core sponses and institution-building efforts. Practical expres- zone of the world economy, surrounded by a peripheral sions of such efforts include a complete reorganization of zone of Third World nations, with a complex set of in- international financial arrangements, along with a restruc- terdependencies running between the two (as described turing and reinforcement of international forums of collec- by world system theorists such as Wallerstein 1979). tive decision-making and action such as the G7/G8 group, Today, after much economic restructuring and tech- the Organisation for Economic Co-operation and Devel- nological change, significant transformations of this opment (OECD), the World Bank, the IMF, and a newly older order have occurred, creating the outlines of a new streamlined GATT, now known as the World Trade Orga- 86 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Table 10.1 The World's 30 Largest Urban Areas Ranked by Estimated Population in the Year 2000 Population (millions) Urban Area 1950 1970 1990 2000 (estimate) 2015 (estimate) 1 Tokyo, Japan 6.9 16.5 25.0 27.9 28.7 2 Mumbai, India 2.9 5.8 12.2 18.1 27.4 3 Sao Paulo, Brazil 2.4 8.1 14.8 17.8 20.8 4 Shanghai, China 5.3 11.2 13.5 17.2 23.4 5 New York, USA 12.3 16.2 16.1 16.6 17.6 6 Mexico City, Mexico 3.1 9.1 15.1 16.4 18.8 7 Beijing, China 3.9 8.1 10.9 14.2 19.4 8 Jakarta, Indonesia n.a. 3.9 9.3 14.1 21.2 9 Lagos, Nigeria n.a. n.a. 7.7 13.5 24.4 10 Los Angeles, USA 4.0 8.4 11.5 13.1 14.3 11 Calcutta, India 4.4 6.9 10.7 12.7 17.6 12 Tianjin, China 2.4 5.2 9.3 12.4 17.0 13 Seoul, Rep. of Korea n.a. 5.3 10.6 12.3 13.1 14 Karachi, Pakistan n.a. n.a. 8.0 12.1 20.6 15 Delhi, India n.a. 3.5 8.2 11.7 17.6 16 Buenos Aires, Argentina 5.0 8.4 10.6 11.4 12.4 17 Metro Manila, Philippines n.a. 3.5 8.0 10.8 14.7 18 Cairo, Arab Republic of Egypt 2.4 5.3 8.6 10.7 14.5 19 Osaka, Japan 4.1 9.4 10.5 10.6 10.6 20 Rio de Janeiro,. Brazil 2.9 7.0 9.5 10.2 11.6 21 Dhaka, Bangladesh n.a. n.a. 5.9 10.2 19.0 22 Paris, France 5.4 8.5 9.3 9.6 9.6 23 Istanbul, Turkey n.a. n.a. 6.5 9.3 12.3 24 Moscow, Russia 5.4 7.1 9.0 9.3 n.a. 25 Lima, Peru n.a. n.a. 6.5 8.4 10.5 26 Teheran, Iran n.a. n.a. 6.4 7.3 10.2 27 London, U.K. 8.7 8.6 7.3 7.3 n.a. 28 Bangkok, Thailand n.a. n.a. 5.9 7.3 10.6 29 Chicago, USA 4.9 6.7 6.8 7.0 n.a. 30 Hyderabad, India n.a. n.a. n.a. 6.7 10.7 Source: United Nations (1995) n.a. = data not available. nization (WTO). While these political responses to the 3. Sovereign states and national economies remain pressures of globalization remain limited in scope and se- prominent, indeed dominant elements of the global verely lacking in real authority, they are liable to grow and landscape, but they are clearly undergoing many sea consolidate as world capitalism continues its expansion. changes. Individual states no longer enjoy the same de- 2. As a corollary of these pressures, multination blocs gree of political autonomy they once possessed, and such as the European Union (EU), North American Free under intensifying globalization find themselves less able Trade Agreement (NAFTA), South American Common or willing to safeguard all the regional and sectional in- Market (MERCOSUR), Association of South-East Asian terests within their jurisdictions than they once were. Nations (ASEAN), Asian-Pacific Economic Cooperation Also, national economies have been subject to massive (APEC), Caribbean Community and Common Market debordering, so it is increasingly difficult, if not impos- (CARICOM), and many others have proliferated. These sible, to say precisely where, for example, the American blocs, too, can be seen as institutional efforts to capture economy ends and the German or Japanese economies the benefits and control the negative effects of the steady begin. As a result, some of the regulatory functions for- spilling over of national capitalisms beyond their tra- merly carried out by the central state have been drifting ditional political boundaries. The blocs remain in vari- to higher levels of spatial resolution, while other func- ous stages of formation, with the EU in the vanguard. tions have been drifting downward (Swyngedouw 1997). Because such blocs involve only small numbers of par- 4. Accordingly, and most importantly in the present ticipants, they are more manageable as political units context, there has been a resurgence of region-based (transactions-costs problems are relatively restrained) forms of economic and political organization, with the compared with actual or putative global organizations. most overt expression being the formation of large city- GLOBAL CITY-REGIONS AND THE NEW WORLD SYSTEM 87 regions. These constitute a global mosaic that is begin- returns effects that typically appear when firms work in ning to override the core-periphery relationships that collaborative association or the knowledge spillovers that have hitherto characterized much of the macrogeogra- occur in day-to-day business dealings. phy of capitalist development. Consider two opposing hypothetical situations. First, The propensity of many types of economic activity- in an economy marked by uniformly high spatial trans- manufacturing and service sectors alike-to gather into actions costs and simple organizational bonds, where all dense regional clusters or agglomerations appears to be exchanges are purely additive and lacking in synergy, we intensifying. This renewed quest for collective propin- would expect to observe limited forms of urban and re- quity on the part of all manner of economic agents gional development. The kinds of transport-cost mini- can be interpreted as a strategic response to heightened mizing outcomes described by classical Weberian or (global) economic competition, in the context of a turn L6schian location theory would be the main kind of to post-Fordism in modern capitalism. Propinquity is outcome here. In another and purely fictional economy especially important because it is a source of enhanced where the time and money costs of spatial transacting competitive advantage for many types of firms (Porter are zero, we would expect to observe a state of geo- 2000; Scott 1988; Storper 1997). As a corollary, large graphic entropy or randomness no matter what forms regional production complexes are starting to function of organizational interaction may be present. as platforms for contesting global markets. In contrast to these two cases, the world that we ac- At the same time, the diminishing capacity of cen- tually inhabit is one in which the spatial costs of trans- tral governments to deal with the nuanced policy acting are sometimes extremely high (for example, many needs of each region within their borders means that kinds of face-to-face exchanges), and sometimes ex- many regions are now faced with the choice of either tremely low (for example, international monetary flows). passive subjection to cross-border pressures or active Ours is a world, too, in which the organizational bonds institution-building, policymaking, and outreach to try between different economic agents are frequently im- to turn globalization to their advantage. Regions that bued with multiple synergies. Moreover, as post-Fordist take the latter course are likely to face many new tasks forms of productive organization take deeper hold in of political coordination and representation. These tasks economic and social affairs, new synergistic relations are of special urgency at a time when large city-regions seem to be multiplying (Cooke and Morgan 1998; Scott function more and more as poles of attraction for low- 1998). I shall now attempt to show that this state of af- wage migrants from all over the world, so that their pop- fairs leads to the formation of new urban superclusters ulations are heavily interspersed with polyglot and often or city-regions. disinherited social groups. Hence, many city-regions The leading edges of the post-Fordist economy are today are confronting pressing issues related to political represented by sectors like high-technology production, participation and the reconstruction of local political neoartisanal manufacturing, cultural-products indus- identity and citizenship. tries, the media, and business and financial services- The Economic Order of Global C:ity-Regions that is, by sectors that persistently take the form of intricate networks or complexes of producers. These sec- One of the seeming paradoxes of contemporary geogra- tors are important sources of increasing-returns effects, phy is that whereas dramatic improvements in trans- and hence of competitive advantage, in that they exhibit portation and communication are bringing all parts of high levels of specialization and complementarity so that the world into ever-closer contact, dense urban agglom- firms tend to interlock organizationally with one an- erations continue to grow in size and importance. These other and to develop relational asset specificities. Spe- apparendy incompatible trends turn out on closer scrutiny cialization and complementarity also provide insurance to be two mutually reinforcing faces of the network against critical supply failures due to sudden or unpre- structures (that is, transactional interdependencies) that dictable input needs-an important feature in flexible constitute the basic scaffolding of economic and social economies, where long-run planning of production life. These economic and social networks are character- schedules is extremely difficult. The many traded and ized by an intrinsic duality: they possess a definite spa- untraded interdependencies among producers in these tial structure, signifying that any bilateral or multilateral networks continually generate new information and transaction will exact some locationally dependent cost; practical know-how, leading in turn to learning effects they also represent social organizations marked by bond- and to high levels of innovation. These effects are com- ing and interaction effects that often produce synergistic plemented by the network-specific skills and habits that outcomes. The latter are exemplified by the increasing- workers acquire, which become part of a common pool 88 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION of human resources upon which employers can draw. As large-scale agglomeration occurs, diverse other out- Relationships like these are the lifeblood of dynamic comes result, underpinning the dynamics of convergence economic systems, enabling them to achieve capacities and growth. Among the more important are the forma- as a whole that are much greater than the sum of their tion of dense physical infrastructures typically supplied component parts. In this situation, there is a strong like- out of public funds as cities expand, the development of lihood at the outset that the geography of production dense local labor markets and the concomitant emergence will be resolved as a series of sizable agglomerations. of extended webs of residential services, and the gradual Because traded and untraded interdependencies are consolidation of conventions and cultures that enhance always geographically extensive, the transactions costs the capacities of individuals to perform in the local eco- that they incur have a spatial dimension, and in the con- nomic environment (Storper 1997). Above all, agglomer- temporary world, these costs tend to vary enormously, ation has many positive effects on the ability of cities to depending on what is being transacted. For many inter- function as centers of learning, creativity, and innovation. related groups of producers (especially in post-Fordist Precisely because cities are transactions-intensive foci of sectors), some transactions depend on locational prox- many interdependent activities, they are also places in imity of all participants if they are to be successfully car- which new social encounters and experiences endlessly ried out (notably where transacting involves frequent, occur, and in which enormous quantities of information unpredictable, and constantly shifting face-to-face en- are daily created and circulated. These processes unfold counters). Other transactions incur such low unit costs informally in many small, unrecorded events and encoun- that they can effectively span an unlimited geographic ters, but they cumulatively function as important foun- range. Additionally, the rich synergies and increasing- dations of local innovative energy and successful entre- returns effects typical of producer networks tend to preneurial effort. They are all the more pervasive in large magnify any basic proclivity to physical agglomeration cities because of the countless kinds of interpersonal en- due to high transactions costs. This state of affairs is ex- counters that can occur, and out of which may flow emplified by the motion picture industry of Hollywood, serendipitous forms of creative action (Scott 1999). Large where actual production occurs mostly within a geo- cities, as a result, are invariably important centers of re- graphically circumscribed network of transactions that sourcefulness and invention for all sectors of production is also the site of many interfirm synergies, whereas final (especially, perhaps, for the neoartisanal, fashion, and products circulate with ease around the globe (Storper cultural-products industries that are to be found within and Christopherson 1987). The net result will be a ten- them in increasingly significant concentrations). dency for urban superclusters to spring forth. Because of these processes, large cities or city-regions In brief, the clustering of economic activities will usu- have become a more insistent element of the geographic ally be especially pronounced where three sets of relation- landscape than at any previous time in history. Over the ships come into mutual interaction. First, where there past few decades and across the globe, numerous urban are selectively high transactions costs on the production centers have been transformed into superclusters whose side, interlinked firms will tend to converge toward their extent and growth are due above all to the condition own center of gravity. Second, if spatially dependent that many of the leading sectors of capitalism today are transaction costs on certain outputs are low and markets organized as intensely localized networks of producers are expanding (due to globalization, for example), the ag- with powerful endogenous growth mechanisms and an glomeration will tend to grow and differentiate internally increasingly extended market reach. via the division of labor. Third, increasing-returns effects The Political Order of Global City-Regions embedded in traded and untraded interdependencies among producers will reinforce agglomeration and en- The world system is thus in a state of rapid economic flux, sure that growth leads to more growth (cf Romer 1986). leading to many adjustments in political geography. On The superclusters that result from these processes are the one side, profound economic changes are prompting the core functional units of global city-regions. To be diverse responses and experiments in regulatory coordi- sure, large-scale urban or regional growth also brings a nation, from the global to the local. On the other side, variety of negative externalities that (in the absence of the new regulatory institutions that are now beginning to remedial action) set in motion centrifugal tendencies. assume clearer outline on the world map help channel What we normally observe in response to this situation, economic development into spatial structures running however, is regulatory action on the part of local author- parallel to the fourfold political hierarchy described ear- ities to bring such externalities under at least approxi- lier. While the political shifts occurring at each level in mate control. this hierarchy pose perplexing problems, the new global GLOBAL CITY-REGIONS AND THE NEW WORLD SYSTEM 89 mosaic of city-regions is certainly one of the least well un- such as agencies of local government, private-public derstood. Moreover, precisely because individual regional partnerships, and civil associations. But it derives its units are the basic motors of a rapidly globalizing produc- force and legitimacy from the positive role that coordi- tion system, much is at stake as they sharpen their politi- nating agencies can play by promoting and shaping crit- cal identities and institutional foundations. ical increasing-returns effects that would otherwise fail We may well ask how these regions are to be defined to materialize or that would be susceptible to severe mis- as political-cum-territorial units with greater or lesser allocation. The kinds of tasks that agencies such as these powers of coordinated action. In many instances, of might perform include the fostering of agglomeration- course, the boundaries of city-regions will tend to coin- specific technological research, providing high-risk capi- cide with a preexisting metropolitan area. But how will tal to small start-up firms, protecting infant industries, these boundaries be drawn when several metropolitan upgrading workers' competencies, cultivating collabora- areas lie in close proximity-as, for example, on the tive interfirm relationships, and promoting export mar- northeast seaboard of the United States? And how far kets for local products. There is also, of course, a contin- into its hinterland will the political mandate of any city- ued urgent need for more traditional types of urban region range? In fact, the final geographic shape of any planning to ensure that periodic land use and transporta- given global city-region is apt to be indeterminate in a tion breakdowns do not cut too deeply into local eco- priori terms because so much depends on how adjacent nomic performance and social life. territorial units interact politically, and on their ability The prospect of a mosaic of global city-regions, each or willingness to enter into durable political coalitions. characterized by an activist collectivity seeking to rein- That said, we can already see some of the outlines of force local competitive advantage, brings up additional things to come in the new regional government systems questions and problems. Rising levels of regional ac- that are being put into place in a number of European tivism can destabilize and politicize interregional rela- countries and in the maneuvering (some of which may tions, both within and across national boundaries. Such bear fruit, some of which will lead nowhere) that is cur- predicaments are already manifest in the formation of rently gathering steam around prospective municipal al- regional alliances (such as the Four Motors of Europe liances such as San Diego-Tijuana in the United States Program and the recent linking of the London and and Mexico, Cascadia in the northwestern United States Frankfurt stock exchanges), leading to complaints from and western Canada, the Trans-Manche region in north- outsiders about unfair competition. east France, western Belgium and southeast England, Another predicament resides in attempts by represen- Padania in northern Italy, Singapore-Johore-Batam, and tatives of some regions to lure assets of others into their Hong Kong-Shenzen. Some of these alliances involve own geographic orbit, often at heavy social cost. Similar cross-border agreements. problems ensue when different regions push to secure a Much of the political change occurring in the world's decisive lead as the dominant center of some budding large city-regions represents a search for structures of industry. These interregional collisions also afford multi- governance capable of securing and enhancing competi- national corporations expanding opportunities to play tive advantage in a globalizing economic order. Agglom- one region off against another in competitive bidding erated production systems are the arenas of region wide wars for new direct investments, a phenomenon partic- synergies, but because these synergies so frequently as- ularly pronounced in contemporary Brazil (Rodriguez- sume the guise of externalities, they will always exist in Pose and Tomaney 1999). some suboptimal configuration as long as individual de- As the new regionalism takes deeper hold, stresses cisionmaking and action alone prevail in the economic and strains of these types are likely to be magnified, and sphere. These synergies have enormous relevance to the a need for action at the national, plurinational, and destinies of all the firms and workers in the immediate eventually global levels of political coordination is fore- locality, and they assume dramatic importance in a world seeable to establish ground rules for interregional rela- where the continued spatial extension of markets gives tions (including aid to failing regions) and to provide each city-region vastly expanded economic opportuni- forums for interregional problem solving. The European ties but where the same extension also greatly heightens Committee of the Regions, established under the terms outside economic threats. The economies of large city- of the Maastricht Treaty, may represent an early expres- regions are thus intrinsically overlain by a domain of col- sion of this dawning imperative. lective order defined by these synergies, and this consti- A further question arises as to what macropolitical or tutes a crucial domain of social management. Such ideological formations will provide a framework for the management may assume various institutional forms institution-building and policymaking projects that can 90 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION now be ever more strongly envisioned at different spatial tant first step in constructively treating their internal levels. Giddens (1998) has forcefully argued that two dysfunctionalities. Large city-regions, with their rising main political principles appear to be fighting for posi- levels of social distress resulting from globalization, con- tion, certainly in the more economically advanced parts front urgent political challenges in this regard, not only of the globe. One is the dominant neoliberal view pre- because their internal conviviality is in jeopardy, but also scribing minimum government interference in, and because any failure to act is likely to undermine the ef- maximum market organization of, economic activity-a fectiveness of purely economic strategies. view that is sometimes erroneously considered to be an From all of this, it follows that some reconsideration of inescapable counterpart of globalization. In light of the the everyday notion of citizenship is long overdue. An pressing need for collective action in global city-regions alternative definition of citizenship, one more fully in har- (not to mention other domains of economic activity), mony with the unfolding new world system, would pre- neoliberalism, certainly in the version that crudely advo- sumably assign basic political entitlements and obliga- cates laissez-faire as a universal panacea, seems to offer a tions to individuals not so much as an absolute birthright, seriously deficient political vision. The other approach is but as a function of their changing allegiances in different a renascent social democracy, which, especially in West- geographic contexts. In fact, traditional conceptions of ern Europe, has enjoyed notable electoral success of late. the citizen and citizenship are vigorously in question On the economic front, social democracy is prepared to at each geographic level of the world system, for we all acknowledge and work with the efficiency-seeking prop- simultaneously participate in local, national, plurina- erties of markets where these are consistent with stan- tional, and global communities. Perhaps nowhere is the dards of social fairness and long-term economic well- urge to participate more immediate or pressing than in the being, but to intervene selectively where they are not. A case of large global city-regions (cf. Holston 2000; Keat- social democratic politics would seem to be well armed ing 2000). Even though only a few tentative and pioneer- to build the social infrastructures and enabling condi- ing instances of pertinent reforms are as yet in evidence in tions at every geographic level, which are becoming more such regions (the case of Amsterdam being perhaps the critical to high levels of economic performance. At the most advanced), increasing numbers of experiments in city-region level, in particular, these tasks can be identi- local political enfranchisement will no doubt proceed as fied with the compelling need to promote local levels of city-regions start to deal seriously with the new economic efficiency, productivity, and competitiveness that mar- and political realities that they face. In a world where mo- kets alone can never fully secure. bility is continually increasing, individuals may one day There is a further argument in favor of a social demo- freely acquire citizenship in large city-regions many times cratic approach to the governance of global city-regions, over, in conjunction with their movements from place to one that is associated with the need for remedial collec- place over the course of their lifetimes. tive action in local economic affairs, but that also goes Coda well beyond this particular objective. Issues of represen- tation and distributional impact are always in play in any Globalization has potentially both a dark, regressive side political community, especially where social manage- and a more hopeful, progressive side. If the analysis pre- ment of the local economy is under way (Mouffe 1992). sented here turns out to be broadly correct, then the The question of local democratic practice and how to es- views expressed in some quarters that deepening global- tablish effective forums for popular participation is in- ization is a retrograde step for the masses of humanity can escapably linked to the more technocratic issues raised be taken as a warning about a possible future world-but by the challenges of economic governance in global city- by no means as a representation of all possible future regions. worlds. Insistent globalization under the aegis of a tri- This question takes on special urgency in view of the umphant neoliberalism would no doubt constitute some- role of large global city-regions as magnets for low-wage thing close to a worst-case scenario, leading to rising so- immigrants, many undocumented, with significant seg- cial inequalities and tensions within city-regions and ments of their populations often composed of margin- greater discrepancies in growth rates and development alized and politically dispossessed individuals. Beyond potentials among them. Alternative and realistic possibil- considerations of equity and social justice, enlarging the ities can be advanced, however, and I have tried to sketch sphere of democratic practice is an important practical out some of them here. Globalization, indeed, is the po- means of registering and dealing with the social tensions tential bearer of significant social benefits. At this stage that invariably occur in dense social communities. The in history its future course is open-ended, and it will cer- mobilization of voice in such communities is an impor- tainly be subject in the future to many different kinds of GLOBAL CITY-REGIONS AND THE NEW WORLD SYSTEM 9I political contestation, some of which will mold it in de- Geographies of Global Change: Remapping the World in the cisive ways. In particular, globalization raises important Late Twentieth Century, pp. 232-47. Oxford: Blackwell. new questions about economic governance and regula- Krugman, P. 1996. Pop Internationalism. Cambridge, MA: MIT tion at all spatial levels, and some form of social market Press. Mouffe, C. 1992. "Democratic Citizenship and Political Com- politics seems to offer a viable, fair, and persuasive way munity." In C. Mouffe, ed., Dimensions ofRadical Democracy, of facing these questions. pp. 225-39. London: Verso. I have said little or nothing in this account about the O'Brien, R. 1992. Global Financial Integration: The End of less-developed countries, but I see no reason-with due Geography. London: Pinter. acknowledgment of the enormous difficulties posed by Porter, M. 2000. "Regions and the New Economics of Compe- tition." In A. J. Scott, ed., Global City-Regions. Oxford: Ox- the vicious circles in which they are often caught-why for Unvrst Prss ford University Press. at least some of them cannot benefit from the processes Rodriguez-Pose, A., and J. Tomaney. 1999. "Industrial Crisis in of urbanization and economic growth brought into the Centre of the Periphery: Stabilisation, Economic Restruc- focus here. These processes suggest that some of the turing and Policy Responses in the Sao Paulo Metropolitan more urbanized regions in these countries will eventu- Region." Urban Studies 36: 479-98. ally join the expanding mosaic of global city-regions, Romer, P. M. 1986. "Increasing Returns and Long-run Growth." just as Seoul, Taipei, Hong Kong, Singapore, Mexico Journal of Political Economy 94: 1002-37. Sassen, S. 1991. The Global City: New York, London, Tokyo. City, Sao Paulo, and others have done, and are doing, Princeton: Princeton University Press. before them. Scott, A. J. 1988. New Industrial Spaces: Flexible Production Or- ganization and Regional Development in North America and Bibliography Western Europe. London: Pion. Badie, B. 1995. La Fin des Territoires: Essai sur le De'sordre Inter- Scott, A. J. 1998. Regions and the World Economy: The Coming national et sur l'Utilite'Sociale du Respect. Paris: Fayard. Shape of Global Production, Competition, and Political Order. Castells, M. 1996. The Rise of the Network Society. Oxford: Black- Oxford: Oxford University Press. well. Scott, A. J. 1999. "The Cultural Economy: Geography and the Cooke, P. N., and K. Morgan. 1998. The Associational Economy: Creative Field." Media, Culture and Society 21: 807-17. Firms, Regions, and Innovation. Oxford: Oxford University Storper, M. 1997. The Regional World: Territorial Development Press. in a Global Economy. New York: Guilford Press. Friedmann, J., and G. Wolff. 1982. "World City Formation: An Storper, M., and S. Christopherson. 1987. "Flexible Specializa- Agenda for Research and Action." International Journal of tion and Regional Industrial Agglomerations: The Case of the Urban and Regional Research 6: 309-44. U.S. Motion-Picture Industry." Annals of the Association of Giddens, A. 1998. The Third Way: The RenewalofSocialDemoc- American Geographers77: 260-82. racy. Cambridge: Polity Press. Swyngedouw, E. 1997. "Neither Global nor Local: Globaliza- Hall, P. G. 1966. The World Cities. London: Weidenfeld and tion and the Politics of Scale." In K. R. Cox, ed., Spaces of Nicolson. Globalization: Reasserting the Power of the Local, pp. 137-66. Hirst, P., and G. Thompson. 1996. Globalization in Question. New York: The Guilford Press. Cambridge: Polity Press. Taylor, P. J. 2000. "World Cities and Territorial States under Holston, J. 2000. "Urban Citizenship and Globalization." In A. Conditions of Contemporary Globalization." Political Geog- J. Scott, ed., Global City-Regions. Oxford: Oxford University raphy 19: 5-32. Press. United Nations. 1995. World Urbanization Prospects: The 1994 Keating, M. 2000. "Governing Cities and Regions: Territorial Revision. New York: United Nations. Reconstruction in a Global Age." In A. J. Scott, ed., Global Veltz, P. 1996. Mondialisation, Villes et Territoires: L'Economie City-Regions. Oxford: Oxford University Press. dArchipel. Paris: Presses Universitaires de France. Knox, P. L. 1995. "World Cities and the Organization of Global Wallerstein, I. 1979. The Capitalist World Economy. Cambridge: Space." In R. J. Johnston, P. J. Taylor, and M. J. Watts, eds., Cambridge University Press. ~~ -,',N a i t> Making Decentralization s W VVork in a Global 1 L% Economy _ F _LI chapterl | How to Design a Fiscal Decentralization Program Roy Bahl Dean of the Andrew Young School of Policy Studies Georgia State University H - he past decade has been replete with tional governments, in the context of the new global rhetoric about endowing subnational economy. How would it design such a program to have governments with more autonomy maximum chances for success? This chapter offers a set to spend as well as collect revenues. of normative guidelines about the design for a compre- Although decentralization has pro- hensive fiscal decentralization program.' ceeded slowly, more and more gov- Guideline 1. Consider all three levels of ernments around the world are coming to power on a Governent. platform of citizen participation, and decentralization is government. becoming difficult to resist. One important lesson of the A key issue is whether a central government's decentral- past two decades, learned especially well in East Asia, is ization program will include local as well as state gov- that strong central governments cannot sustain them- ernments.2 There are good arguments that it should. In selves without granting appropriate decisionmaking au- some countries provincial governments are too large to tonomy to their subnational governments. Economic allow citizen participation and ensure that voters will development has also dramatically improved the capa- hold government officials accountable, so fiscal decen- bilities of local governments, and recognition is grow- tralization must include a lower level of government. ing that granting local governments some autonomy is For example, eight Chinese provinces have populations preferable to breaking up nation-states into smaller en- of more than 50 million, which would rank them among tities. Yet even as signs appear that some countries are the 20 largest countries in the world. ready to move forward, poorly conceived decentraliza- While it is essential that the lower tiers of govern- tion policies could prevent progress. ment be explicitly recognized as a component of the fis- The major problems with most decentralization pro- cal decentralization program, there are choices about grams are, first, that the government is often commit- how this should be done. The most common arrange- ted more in rhetoric than in fact, and second, that the ment is for each province to define the intergovernmen- design is usually too narrow and does not focus on all of tal fiscal relations with its local governments (as in Rus- the components. These are of course related issues be- sia and the United States). A second approach, preferred cause to do fiscal decentralization "right," a major shift by centralists, would require all provincial governments in the balance of fiscal power is involved. The tendency to mandate a nationally uniform set of relations with in too many countries is to take a much narrower view their local governments. A third approach, growing out in order to escape most of the hard decisions about re- of some suspicion that provincial or state governments assigning tax, expenditure, borrowing, and budgetary will not "do the right thing," involves the central gov- powers. ernment prescribing safeguards to protect the rights of But suppose a country did want to implement a pro- local governments (for example, in India, Nigeria, and gram that would pass significant fiscal powers to subna- the United States in the 1970s). 94 HOW TO DESIGN A FISCAL DECENTRALIZATION PROGRAM 95 Guideline 2. Match the fiscal system to the technical assistance to local governments in accounting, administrative capacity of local governments. treasury, tax administration, data processing, and proj- Many countries believe that effective decentralization ect evaluation. A corollary to the guideline just outlined requires giving uniform powers and responsibilities to is that the central government will make the rules of de- all subnational governments. However, local and even centralization, and they must also monitor the subna- provincial governments vary widely in their capacity to tional governments to make sure the rules are obeyed. deliver and finance services, and thus may not fit well Because developing and transition countries are by into a uniform system of taxing, spending, and borrow- definition changing rapidly, the intergovernmental sys- ing powers. In fact, the decentralization process may tem will also change. A central fiscal monitoring system work best if larger local governments are assigned signif- must be in place to guide restructuring of the grant sys- icant fiscal powers at the outset while smaller ones grow tem, detect early warning signs of cities in financial into it. Kenya, for example, classifies municipal govern- trouble, monitor borrowing limits, decide when a local ments and gives them differential fiscal powers, and government is ready to graduate to the next level of au- many U.S. states distinguish among local governments tonomy, and so on. There is also a need to set up special in terms of the taxing powers they assign to them. provisions to assist subnational governments facing fis- cal distress (for example, financial control boards as are Guideline 3. Ensure that central governments being set up in South Africa) or mechanisms of inter- keep the fiscal decentralization rules they set. governmental assistance to cover debt service costs, as Provincial and local officials in many countries believe are being used in some Latin American countries. that the central government will not hand over funds Guideline 5. Impose a hard budget constraint on when times are hard, nor will it honor its pledge to per- local governments. mit local autonomy. In fact, central governments often- times do violate or compromise the principles they es- Provincial and local governments must balance their tablish. There are few developing or transition countries budgets without any end-of-year assistance from the that cannot recite instances of "broken promises" by cen- central government. This guideline is particularly im- tral governments. Examples of such behavior include: portant because fiscal discipline will occur only if sub- national governments believe they are on their own. En- * Imposing unfunded mandates on subnational govern- emies of this rule include year-end grants to cover revenue ments shortfalls or project overruns, and bailouts of subna- * Underfunding promised revenue transfers to subna- tional governments that are delinquent on debt pay- tional governments ments. A good case in point is the Russian regional gov- * Reassigning expenditures without reassigning revenues ernments, where some intergovernmental transfers from * Abolishing or restricting subnational government taxes. the center are not known until the end of the fiscal year, and even then are determined partly on an ad hoc basis. If decentralization is to have a chance, the central This lack of transparency leads to a dependence of the government must adhere to the rules it sets. The Presi- lower level governments on the center to provide the dent and the Parliament must guarantee that fiscal regular "bailout" (Bahl et al. 1999). promises are kept. This means that central governments The absence of a hard budget constraint on local gov- must take care to design programs that include only ernments has important consequences. First, the consol- those fiscal promises that can be kept. Otherwise, sub- idated government deficit will be higher because subna- national governments will build this uncertainty into tional governments will have an incentive to overspend. their fiscal decisionmaking process, and the result may Second, increased tax effort by subnational governments be inefficient budgetary planning and actions based on will be discouraged. Finally, subnational government of- the perception of a soft budget constraint. ficials will be held less accountable by their constituents for imprudent fiscal behavior because the national gov- Guideline 4. Maintain a strong central ability to monitor and evaluate decentralization. ement will also be seen as culpable. Despite the need for flexibility in adapting plans to local Guideline 6. Assign local governments significant conditions, central governments will have to impose a taxing powers. uniform system of financial accounts and audit rules, Voters will hold their elected officials more accountable and at least set disclosure requirements for governments if local governments finance services largely through that borrow. Central governments must also provide local taxes rather than transfers of funds from the central 96 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION government. It is important that these local taxes be (a) fuels-are an excellent revenue source for subnational visible to local voters, (b) large enough to impose a no- governments because they fall on people who reap the ticeable burden, and (c) not easily exported to residents benefits, and they are relatively easy to administer (Bahl of other jurisdictions. Nuisance taxes and minor levies, and Linn, 1992). A local tax on motor fuel, for exam- the preference of many centralists for assignment to sub- ple, especially in urban areas, can charge motorists for national governments, will not lead to adequate account- the marginal costs of the congestion and pollution they ability of government officials. Those who say that the generate and the road services they use. The property tax, feasible set of provincial and local government taxes is on which local governments in most countries rely, is very limited are mistaken, as is suggested by the follow- also a quasibenefit tax in that it is used to provide ser- ing list and discussion of possible subnational govern- vices to property owners and occupants. What's more, ment revenue sources (Bahl and Linn 1992; Bird 1999). the tax is highly visible in the local area, its burden is not Individual income taxes are a good choice for sub- easily exported to outside residents, and it allows local national governments, either as a piggyback on the assessors to identify and tax local wealth. national income tax or as a standalone tax on local pay- Finally, user charges for local services are especially ap- rolls. Local income taxes fulfill the criteria for an appro- propriate because governments can determine the costs priate tax: Local residents shoulder the burden and ad- of many such services and fully recover them-or priva- ministration is feasible. tize the services. Taxes onforeign trade and taxes on corporate income do To further enhance the ability of state and local gov- not meet the exportability and "easy administration" ernments to raise revenues, central governments should tests, and thus are probably a poor choice for subna- provide incentives for them to borrow funds to finance tional governments. Few developing countries allow major infrastructure projects and other long-lived capi- subnational governments much discretion in setting the tal assets. The duration of bonds used to finance these rate and base of the company income tax. projects should match the expected life of the assets. General sales taxes have always been thought a diffi- However, if provincial and local governments are to be cult choice for subnational governments, either because given borrowing powers and some degree of autonomy, of administrative difficulties or because the tax burden then central governments need to lay down strict rules would be exported. However, there is now a flurry of in- about the conditions under which such borrowing can terest in subnational value-added taxes, for example, at take place, including disclosure requirements, senior the state level in India (Bird and Gendron, 1998). McLure claims, and actions that may be taken in the event of (1999) reviews the administrative arrangements under nonpayment.3 which a broad-based VAT could work at the subnational level (a nationally uniform tax base and a well-function- Guideline 7. Match the system for transferring ing VAT operating at the central government level), funds between central and local governments to Many developing and transition countries also find re- the goals of decentralization. tail sales taxes impractical because of the difficulty of col- Despite the theoretical appeal of provincial and local tax- lecting taxes from small vendors. However, in Russia, ation, revenue transfers from national to state and local for example, some regional governments impose retail governments are often the most important decentraliza- sales taxes on big-ticket luxury items. Other countries, tion tools. Each system for transferring funds has differ- including the Philippines and South Africa, target the ent impacts on local governments, yet countries often de- retail sector by taxing businesses' gross sales. sign their grants without fully exploring those impacts.4 Excise taxes can be an appropriate state and local rev- The impacts of such transfers depend on two factors: enue source-but not on goods for which a natural mo- how the central government determines the total re- nopoly exists, for example, petroleum production, pos- source pool, and how it distributes this pool among sibly beer and cigarette producers, and some natural local governments. International practice suggests three resources. That's because taxing monopoly goods would basic approaches to the former (see the column head- impose a burden on people who do not gain the bene- ings in Table 11.1). Once the size of the pool is deter- fits derived from the public services supported by the mined, developing countries and transition economies tax. Revenues from monopolies that happen to locate in typically use four methods to allocate the funds (see the their districts thus encourage those provincial or local row headings in Table 11.1). This taxonomy yields 12 governments to overspend. potential grant types (Bahl and Linn 1992). Taxes related to the use of motor vehicles-such as Consider the shared tax column in Table 11.1. Under license fees, tolls, parking taxes, and taxes on motor a type A grant, the subnational government is allowed to HOW TO DESIGN A FISCAL DECENTRALIZATION PROGRAM 97 Table 11.1 Forms of Intergovernmental Grant Programs Method of determining the total divisible pooi8 Specified share of Reimbursement of Method of allocating the pool national or state tax Ad hoc decision approved expenditures Regional or local origin of collections A E n.a. Formula B F n.a. Total or partial reimbursement of costs C G K Ad hoc D H n.a. n.a. Not applicable. aGrant types are described in the text. Source: Bahl and Linn 11992). keep a share of national taxes collected within its bound- national government maximum flexibility in deciding aries. Most transition economies and many developing how much to distribute to the local public sector each countries use this approach.5 Under a type B grant, a for- year. The type C, G, and Kgrants are also centralizing in mula governs the distribution of a defined amount of that they give central government ministries significant national taxes among local governments. For example, control over how money will be spent and allow them the Philippines distributes 40 percent of nationally col- to establish standards for construction projects, salary lected revenue among local governments on the basis of rates, and delivery of services. population, land area, and an equal share for each recip- Countries often mix these grant types, with the result ient. Under type C grants, a fixed percentage of a na- that some grants in the system are decentralizing and tional tax reimburses local governments for the costs of some are centralizing, some are equalizing and some are public projects or needs such as teachers' salaries. counterequalizing, some encourage increased tax effort The second column in Table 11.1 lists types of grants while others discourage it, and so forth. The potential awarded after an ad hoc political decision-such as a for such offsetting effects is a very important problem yearly budget process-establishes the total pool of in grant design and one that few countries have been funds. Under a type H grant, the central government able to avoid. makes all decisions concerning which entities receive the funds and how much. Under type E, E and G, the cen- tral government distributes grants according to objec- Provincial and local administrators often cannot handle tive criteria. The type G grant is conditional, while the complicated arrangements for intergovernmental trans- others are not. fers that impose significant compliance requirements Under a type Kgrant in the third column, a central on them, nor can they meet the requirements of com- ministry decides the amount of funding local govern- plicated expenditure mandates or collect taxes with ments will need to pursue specific projects and whether complicated definitions of the tax base. It is well to re- they meet prescribed eligibility requirements. Examples member that central governments may add to their own indude grants for infrastructure projects, teachers' salaries, problems by designing an intergovernmental fiscal and transfers to individuals to cover income mainte- system that they cannot monitor or support. Well- nance costs. meaning policy analysts often introduce complexities These very different types of grants exert very differ- such as: ent effects. Type A and B grants are arguably the most decentralizing because they guarantee local governments a Complicated formulas and conditions for allocating a specified share of national revenue and usually carry grants that require provincial and local governments no conditions. However, even these two transfers have to supply demographic and economic information very different impacts. The type A grant favors subna- that they do not have or that impose heavy costs on tional governments with strong tax bases, whereas the the central agency that must allocate the funds and type B grant can redistribute funds to subnational gov- monitor the compliance. ernments according to the local characteristics that the * Local taxes structured to accomplish goals other than central government most wants to reward. The ad hoc revenue raising, for example, property taxes that in- transfers are the most centralizing because they allow the clude special provisions to encourage homeownership 98 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Qr suburban locations, or to subsidize the construc- sessments. Central governments should encourage tion of certain types of buildings. such improvements in tax structure. * Expenditure mandates that are difficult to monitor * Providing a specified timetable for review to deter- and with which are difficult to comply. For example, mine whether a subnational government can gradu- requirements that x rubles per patient be spent on ate to the next level of fiscal autonomy. hospital meals or that classrooms be built to certain * Minimizing the degree to which the detail of the specifications and be used in only certain ways. fiscal decentralization program is laid down in the constitution. This is not to say that simplicity alone should drive Guideline 10. Approach fiscal decentralization as a intergovernmental reform. Some complicated rules are essential, such as establishing disclosure requirements for comprehensive system. governments that borrow funds, and establishing uni- Piecemeal reform encompassing only one element of a form accounting systems and auditing procedures. But decentralized system, such as revenue sharing between the objective should be to limit the goals that each pol- central and subnational governments, is not likely to icy instrument is designed to accomplish and to consider succeed. As is noted in Table 11.2, there are many di- the administrative capacity of both the central and the mensions to a fiscal decentralization plan, beginning subnational governments. Simpler arrangements will re- with the need for a popularly elected local government duce the resources that provincial and local governments and a locally appointed administration. Without this must allocate to administration as well as lower moni- feature, accountability will be "up" to the higher-level toring and evaluation costs for the central government. government rather than "down" to the voters, and the Guideline 9. Recognize and plan for the fact that potential gains from fiscal decentralization are not likely intergovernmental systems are alwa ys for ie traniti. to be realized. It is essential that those making the plan for fiscal decentralization have all of the necessary com- Regional disparities, infrastructure quality, investment ponents involved in the discussion. priorities, and local capacity all change over time. Fiscal Guideline 11. Build an internal champion. decentralization programs must be designed to accom- modate these changes. A major factor that will affect de- Fiscal decentralization often has involved more rhetoric centralization is the global economy. Provincial and local than action in developing and transition countries be- governments, especially those in large urban areas, will cause some interest groups must give up power. How- increasingly be able to tax a base of industry that is eas- ever, successful decentralization requires a strong domes- ier to assess and collect, but is more footloose. It also will tic champion who understands the costs and benefits face more demands for workforce development and for for those who must approve and implement the system. developing an IT infrastructure. Giving subnational gov- A "scorecard" that might help in identifying the poten- ernments the ability to respond to these new needs is a tial champions is presented in Table 11.3. major challenge in designing a fiscal decentralization policy. Central governments must therefore keep their decentralization plans flexible while also transparent- Table 11.2 The Components of a System of Fiscal that is, the process by which the system changes should Decentralization be open and obvious. Possible approaches include: * Establishing a commission that reviews the allocation Necessary conditions Desirable condifions of intergovernmental transfers every few years and * Elected local council * Freedom from excessive . . . . . ~~~~~~~~~~~~~~~~~~~central expenditure recommends changes. This approach is flexible yet mandates enables local governments to plan their finances over * Locally appointed chief * Unconditional transfers a multiyear period. Australia and India follow such an officers from higher-level governments approach. * Significant local government * Local borrowing powers * Allowing changes in the local tax structure to reflect discretion to raise revenue economic changes. As some areas develop and urban- * Significant local government ize, for example, they may piggyback new local taxes expenditure responsibilities onto central revenue collections, tax sectors such as . A hard budget constraint the self-employed and small shops, or impose dedi- a Transparency cated revenue raisers such as tolls or special land as- HOW TO DESIGN A FISCAL DECENTRALIZATION PROGRAM 99 Table 11.3 The Champions of Fiscal Decentralization Potentially Strong Supporters Comments The people Often demand more participation in governance at the local level. The president Because decentralization is popular the president often supports it, but must be very mindful of national stability as that affects his or her political standing. The Parliament or Congress Members would like to claim credit for specific local projects, so they will favor a less transparent, more ad hoc system. Urban local governments Urban local governments want to expand their autonomy and gain access to their tax base. External donors They provide encouragement and some technical assistance but are no substitute for an in- country champion. Potentially Weak Supporters Ministry of finance To stabilize a country's economy, these officials want to continue the ministry's hold on fiscal tools and strictly limit decentralization. Ministry of economy These officials want to control how and where a country invests its revenues. They are typically interested in programs with national payoff rather than local benefits. Line ministries These officials would like to control standards for delivery of public services, and often like to approve transfers of funds to local governments. Ambivalent Supporters Local government These officials favor a guaranteed share of revenues for local governments and want to control their distribution. Weaker local governments These officials would like a guaranteed transfer of resources from urban and wealthier local governments. More interested in a transfer system than in local taxation. Conclusion Bahl, Roy W. 1999b. "Fiscal Decentralization as Development Policy." In Public BudgetingandFinance, pp. 59-75, Summer. Fiscal decentralization has not yet become the center- Bahl, Roy W., Galina Kouriandskaia, John Mikesell, Sally Wal- piece of economic development programs that many ex- lace, Natalia Golovanova, Dmitry Shiskin, Andre Timofeev, pected. The advantages of centralization and the politi- Alexander Derugin, Yelena Nikolayenko, Inna Verbina, and cal power of the centralists have been too strong. But as Natasha Minkova. 1999. "Intergovernmental Fiscal Relations the world changes, the case for decentralization becomes in Leningrad Region." International Studies Program, Andrew more and more irresistible. It may'be slowed by an un- Young School of Policy Studies, Georgia State University. Bahl, Roy W., and Johannes Linn. 1983. "The Assignment of stable world economy, as the East Asian crisis demon- Local Government Revenues in Developing Countries." In strated, but its time may have come. Governments Charles E. McLure, Jr., ed., TaxAssignment in Federal Coun- around the world are increasingly elected, and increas- tries. Canberra: Centre for Research on Federal Financial Rela- ingly on a platform of citizen participation in gover- tions, Australian National University. nance, economic development has erodedsomeoft Bahl, Roy W., and Johannes Linn. 1992. Urban Public Finance in Developing Countries. New York: Oxford University Press. arguments in favor of fiscal centralization, and the ser- Bird, Richard M. 1999. "Rethinking Subnational Taxes: A New vice delivery capabilities of local governments have im- Look at Tax Assignment." Working Paper, International proved dramatically. Moreover, much of the world has Monetary Fund (IMF) Fiscal Affairs Department. come to see that granting some form of local autonomy Bird, Richard, and Pierre-Pascal Gendron. 1998. "Dual VATS is better than separatism as a policy direction. The and Cross-Border Trade: Two Problems, One Solution?" In- is better than separatism as a policy direction. Te ,ternational Tax and Public Finance 5: 429-42. greater enemy of progress now is poorly conceived de- Bird, Richard M., Robert D. Ebel, and Christine I. Wallich, eds. centralization policies. Design must match objectives, 1995. Decentralization of the Socialist State. A Regional and and implementation must face up to the many dimen- Sectoral Study, Washington, D.C.: World Bank. sions of decentralization. This chapter has been an at- Bird, Richard M., and Francois Vaillancourt, eds. 1998. Fiscal to stimulate that discussion. Decentralization in Developing Countries. Cambridge: Cam- bridge University Press. Bibliography Dillinger, William. 1994. Decentralization and Its Implications for Urban Service Delivery. Urban Management Program Bahl, Roy W. 1999. Taxation and Intergovernmental Fiscal Rela- Series 16, Washington, D.C.: World Bank. tions in China. Ann Arbor, MI: The 1990 Institute and the Dillinger, William. 1991. Urban Property Tax Reform: Guidelines University of Michigan Press. and Recommendations. Urban Management Program Discus- Bahl, Roy W. 1999a. Implementation Rules for Fiscal Decentral- sion Paper Series 1. Washington, D.C.: World Bank. ization, Working Paper No. 10. International Studies Pro- IADB (Inter-American Development Bank). 1997. Latin Amer- gram, Andrew Young School of Policy Studies, Georgia State ica After A Decade of Reforms: Economic and Social Progress. University. Baltimore, MD: Johns Hopkins University Press. IOO LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Litvack, Jennie, Junaid Ahmad, and Richard Bird. 1998. Re- Wong, Christine. 1995. "Overview of Issues in Local Public Fi- thinking Decentralization in Developing Countries. A Sector nance in the PRC." In Christine Wong, ed., Financing Local Studies Series. Washington, D.C.: World Bank. Government in the People's Republic of China, Manila: Asian Martinez-Vazquez, Jorge, and Robert McNab. Forthcoming. Development Bank. "The Tax Reform Experiment in Transitional Economies." National Tax Journal. Notes McLure, Charles. 1999. "Implementing Subnational Value 1. A much longer discussion of these rules appears in two re- Added Taxes on Internal Trade: The Compensating VAT." cent papers, Bahl, 1999b. Unpublished paper, The Hoover Institution. 2. Some countries transfer funds among more than three lev- McLure, Charles E., Jr., Jorge Martinez-Vazquez, and Sally els. In the Philippines, the central government explicitly includes Wallace. 1999. Fiscal Transition in Kazakhstan. Manila: Asian submunicipal Baranguays in the grant distribution system. Development Bank. 3. For a thorough discussion of borrowing by local govern- Peterson, George. 1999. "Building Local Credit Systems." Urban ments, see Peterson (1999). Management and Municipal Finance Program Paper 25. 4. The definition of an intergovernmental transfer is often de- Washington D.C.: World Bank. bated. While grants to lower-level governments are clearly inter- Prud'homme, Remy. 1995. "The Dangers of Decentralization." governmental transfers, confusion arises in the case of taxes and The World Bank Research Observer 10(2): 201-20. tax expenditures. If the local government can control either the Spahn, P. Bernd. 1997. "Decentralized Government and Macro- rate or base of a levy, it is a tax. If the rate and base are deter- economic Control." Paper prepared for the International Insti- mined by the higher-level government and revenue collections tute of Public Finance 53rd Congress (August, 1998), Kyoto, are assigned to the local government, it is a transfer. If the cen- Japan. tral government allows deductibility of property taxes from cen- Tanzi, Vito. 1996. "Fiscal Federalism and Decentralization: A tral income tax liability, this is also a form of transfer. Review of Some Efficiency and Macroeconomic Aspects." In 5. For a discussion of shared taxes in transition countries, see Michael Bruno and Boris Pleskovic, eds., Annual World Bank Bahl (1999), Martinez-Vazquez and McNab (forthcoming), and Conference on Development Economics 1995. Washington, Bird, Ebel, and Wallich (1995). The sharing of revenues derived D.C.: World Bank. from the taxation of natural resources is a special case of shared taxes. chapter 12 Strategic Fiscal Decentralization in Developing Countries: Learning from Recent Innovations Paul Smoke International Development and Regional Planning Program Department of Urban Studies and Planning Massachusetts Institute of Technology iscal decentralization emerged as one Reform Origins: Many recent fiscal decentralization of the most pervasive global develop- programs have been undertaken in response to economic ment trends of the 1 990s. Rapid eco- and political difficulties, commonly under pressure from nomic, political, and technological international development agencies. The perceived ur- changes-along with disappointing gency results in programs that are hastily defined and and uneven progress in meeting key initiated without developing a minimum enabling frame- public sector objectives-have forced many countries to work or carefully thinking through implementation re- think beyond traditional top-down approaches to de- quirements. Such programs tend to be superficial, fo- velopment and to rely more heavily on lower levels of cusing on textbook solutions derived from the fiscal government.1 federalism literature.5 This literature provides a useful Conventional theory and some limited empirical evi- analytical entry point, but its development in the con- dence suggest various potential benefits of fiscal decen- text of industrialized countries suggests a need for care- tralization.2 Some analysts, however, caution that decen- ful reflection in applying it to developing countries.6 tralization may result in negative macroeconomic effects, Reformers preoccupied with normative models and diminished opportunities for interregional redistribution, rapid results fail to recognize that attitudes concerning and other possible problems.3 In fact, fiscal decentraliza- the way the public sector works change slowly. Central tion is neither inherently desirable nor undesirable-the officials are commonly used to making major fiscal de- appropriate degree and form depends on the context of a cisions unilaterally and controlling local authorities ex- particular country. A common failure to recognize this tensively. Local governments are often comfortable with reality has resulted in many unsuccessful or disappoint- central subsidization, and they may not feel particularly ing reform efforts. With few exceptions, the academic lit- accountable to the electorate. Finally, local government erature and development agency publications are domi- constituents are often not accustomed to paying for ser- nated by reports of weak local governments and costly vices or to expecting much from local governments. fiscal decentralization programs that have made only Such mindsets and resulting patterns of behavior require modest progress toward meeting their goals.4 Even cases years and careful planning to change. cited as successful typically focus on the importance of Institutional Realities: The institutional environment "good leadership" or "effective stakeholder participation" in which decentralization and local government reform without providing much insight about how these mech- programs must be implemented suffers from a number anisms work or how other countries can replicate them. of weaknesses that hinder success.7 First, systems, pro- Barriers to Reform cedures, incentives, and capacity tend to be inadequate at all levels of government. Good performance is often Several key factors underlie the modest performance of not well defined or rewarded, poor performance is rarely fiscal decentralization programs in developing countries. punished, and job satisfaction is often low.8 IOI 102 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Second, the institutional framework in which decen- of rigid technocratic steps. The circumstances under tralization must be designed and implemented is com- which local staff function, however, can differ widely, so plex and poorly coordinated. Most developing countries that standardization may prevent them from achieving appoint a central agency to oversee local governments, key objectives rather than helping them to do so. but powerful coordinating ministries, such as Planning A final structural weakness of many recent decentral- and Finance, often have greater control over decentral- ization programs is their excessive focus on greater con- ization. Key sectoral ministries-public works, trans- sultation with the consumers of public services. Few de- port, health, education, energy, and so on-also typi- velopment analysts would deny the need to increase cally have some jurisdiction over local services under citizen participation and to improve accountability, but their general expertise. The situation is particularly a demand-side emphasis without adequate considera- problematic when responsibilities are dispersed across tion of supply-side deficiencies in public service provi- several powerful agencies with different views of what sion is unlikely to improve performance. decentralization means. International Agency Objectives: The interests of in- Third, whatever the economic and political impetus ternational donors are not necessarily well served by de- to decentralize, there are rarely strong bureaucratic in- centralization. For reasons already outlined, such efforts centives to do so.9 Central ministries generally do not are time-consuming and difficult, and, therefore, likely want to decentralize major responsibilities, thereby un- to cause substantial delays in donor efforts to start new dermining their own control over substantial resources. projects and to move funds. Given spending pressures, They also tend to protect their own institutional inter- substantial donor funds continue to flow for decentral- ests as reforms proceed, complicating coordination. ization, even if recipient governments do not have the Fourth, well-developed local government account- capacity to ensure that the funds will yield sustainable ability to a constituency is rare in developing countries, benefits. although generally better in Latin America.10 Even A second concern is that donors often try to avoid the where local councils are elected, they may not be gen- complexities of cross-donor and sectoral coordination by uinely autonomous or adequately empowered, and they limiting their decentralization support to particular sec- may be dominated by local elites. In any case, local ac- tors or local units. Since donors tend to adopt their own countability means little if the center can override the individualized systems and procedures, the process of results of local democratic processes, which often occurs building a consistent intergovernmental system may be in developing countries. hindered by such fragmentation. Moreover, because Structure of Decentralization Initiatives: Decentraliza- donors often develop individual client relationships with tion initiatives can generally be classified as two types. rival ministries or local governments, they may exacer- The first is unworkably comprehensive, overwhelming bate competition. Donor agencies recognize the need to technical capacity at all levels and threatening bureau- focus more on institution building than on moving cratic and political tolerance at the center.11 The second money, and they also support coordination. To date, is unduly limited, focusing on narrow technical activi- however, there has been only limited concrete action on ties that are not conceived as part of a broader fiscal re- these fronts. form agenda."2 Neither comprehensive nor limited re- forms take into adequate account the wider scope of Recent Cases of Improved Performance bottlenecks previously discussed that make institutional and Innovation change so difficult. To complicate matters, many initia- Although the barriers outlined here may seem over- tives are led by a single central agency perceived as a rival whelming, a number of countries have been managed by other agencies whose cooperation is required for suc- to make progress in recent years.13 Their experiences cessful decentralization. provide ideas for other countries about how to approach Another serious design concern is that decentraliza- decentralization. tion programs tend to treat all local governments (or Ethiopia and Uganda: Building a Solid Foundation classes-large urban, small urban, rural, and so on) as for Refoun similar in capacity and staffing. Treating those with weak for Reform capacity as if they can handle new responsibilities invites In contrast to many developing countries, Ethiopia and failure. Providing technical assistance to those that do Uganda have established strong enabling environments not need it wastes resources. Standardized decentraliza- for decentralization."4 They have been able to do so be- tion programs that have similar expectations of all local cause changing political conditions created a strong in- governments also tend to define performance in terms ternal incentive for reform.15 Donor agencies were in- STRATEGIC FISCAL DECENTRALIZATION IN DEVELOPING COUNTRIES I03 vited to assist only after the respective governments de- tant to decentralize or to work cooperatively with other termined the general direction of reform. ministries.17 The Ethiopian and Ugandan experiences suggest a Second, neither country's decentralization program number of critical components of a sustainable en- adequately recognizes the severe capacity constraints abling environment for fiscal decentralization. First, that must be overcome. In both Ethiopia and Uganda, both countries have developed robust constitutional at least some functions and resources are being decen- and legal provisions for decentralization and the reform tralized to regional and local governments with little ca- of lower levels of government. The 1995 Ethiopian pacity to manage them. Moreover, the need for techni- Constitution and various laws provide for a federal cal assistance to local governments as they assume new system, with clear service provision and revenue re- duties far exceeds the capacity of the relevant central sponsibilities for federal and regional governments. In agencies to deliver it in both cases. Uganda, the Local Governments Statute of 1993, the Third, more emphasis has been given in both 1995 Constitution, and the Local Governments Act of Ethiopia and Uganda to the development of intergov- 1997 provide considerable detail about local service and ernmental transfer programs than to improving local revenue powers. revenue sources, and in neither country are there ade- Second, both Ethiopia and Uganda have reduced the quate incentives in the transfer formulas to encourage coordination problems previously discussed by develop- local revenue generation. Local revenue yields have gen- ing special-purpose, somewhat more "neutral," and high- erally increased under the decentralization programs in level mechanisms to oversee and coordinate decentral- both countries, but much less than expected, and grants ization. In Ethiopia, the Regional Affairs Department in still account for the bulk of local revenues and increas- the Office of the Prime Minister has substantial respon- ingly substantial portions of the central government sibility in this regard, and a Decentralization Secretariat budgets. was set up for similar purposes in Uganda. Although the Fourth, although both countries are moving toward Secretariat reports to the Permanent Secretary for Local greater local participation and/or democratically elected Government, staff members are not civil servants, and councils, there is a lack of explicit linkage between the they are in principle detached from the normal govern- development of local administration and the develop- ment bureaucracy. In addition, the Local Government ment of local democracy. In both Ethiopia and Uganda, Finance Commission, a constitutionally mandated inde- civil society has been historically weak, and there is lit- pendent body that reports directly to the President, has tle recognition that people must learn over time how to considerable influence over local revenue and intergov- be good citizens and how to interact with their local ernmental transfer design. governments. 18 Third, rather than thinking of decentralization as a In summary, both Ethiopia and Uganda have estab- way to reduce pressures on the central budget, both lished a constitutional, legal, and fiscal basis for decen- Ethiopia and Uganda recognize that effective local gov- tralization programs that could serve as models for other ernments must have adequate resources to meet their re- countries. Their perceived political imperative to decen- sponsibilities under decentralization. Accordingly, both tralize, however, has led to attempts to move forward countries have developed significant transfer programs more rapidly and less strategically than a complex process that account for substantial proportions of total central like decentralization merits. revenues (around 30 percent in the Uganda case and more than 40 percent in the Ethiopia case). Indonesia: Building a Decentralization Process in a While these positive features of the decentralization Highly Centralized System frameworks in Ethiopia and Uganda legitimately distin- Indonesia, a Southeast Asian nation of more than 14,000 guish them from many other countries, implementation dispersed islands with the world's fourth largest popula- has been less impressive. First, although there is a strong tion, has long suffered from many of the decentralization legal and constitutional basis for decentralized govern- obstacles we have outlined. Numerous central agencies ment rights and responsibilities in both countries, nei- with conflicting visions about local government roles are ther has a clear process or pragmatic timetable for the involved. They have generally been reluctant to decen- gradual transferring of functions.16 This is partly due to tralize, allegedly because of weak local capacity, but no an initial lack of appreciation for the complexity of de- doubt equally because they control a substantial volume centralization. Equally important is the weakness of the of "decentralization" program resources, which have been coordination bodies, which apparently have inadequate largely provided through heavily fragmented donor ini- authority to mandate cooperation from ministries reluc- tiatives. The real challenge in this case was how to begin 104 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION the contentious process of loosening up top-heavy con- trolled by the center after independence, primarily in trol in a system with several powerful central actors.19 the name of national unity. There were many efforts in In response to disappointing progress of their decen- the 1 980s and 1 990s to reinvigorate local governments, tralization efforts, the Indonesian authorities initiated but most were donor-driven and not a priority for the an experimental government-led effort in 1993 to de- national authorities. As a result, local revenue yields and centralize strategically, based on local government ca- services deteriorated substantially. In recent years, how- pacity. The process of evaluating local capacity brought ever, fueled by political and economic pressures, some together key central ministries under a coordinating genuine attempts to strengthen local governments are committee chaired by a nonthreatening "neutral" agency under way.20 with no direct stake in decentralization. The committee Given the number of local authorities and the mag- intended to repeat the evaluation process regularly. Al- nitude of their problems, the government recognized though the process was not institutionalized because of that not all of the necessary reforms could be under- interministerial disagreements and the onset of political taken at the same time or in all places. Accordingly, the and economic crisis, the fall of Suharto has opened new Small Towns Development Program (STDP), the first opportunities for reform based on the principles of this of a new generation of reform efforts based in the Min- experiment. istry of Local Government (MLG), involved experimen- The Indonesia approach, which was piloted in three tation with about 15 local authorities. This project, provinces before being abandoned, has several positive which was jointly funded by the Kenyan and German features. First, it exposed an underlying agreement (GTZ) governments, defined several broad common re- among bureaucrats in competing agencies on the ability form goals for the participating local authorities, but the of some local governments to manage their fiscal affairs exact steps each took and the pace at which they took more independently. Second, it stimulated extensive di- them were negotiated. During implementation, a mo- alogue on the meaning, scope, and pace of decentraliza- bile technical assistance team worked periodically on site tion. Third, the process was structured to reduce control with local officials to implement gradually, and, as nec- of more capable local governments, freeing up resources essary, to modify, the negotiated program of reforms. to target technical assistance to weaker ones. Fourth, the Local citizen participation mechanisms were required as implementation approach was pragmatic, such that tasks part of the reform process for certain types of activities, were to be decentralized gradually to more capable local and some training opportunities were provided for local governments. Fifth, the evaluation process provides in- officials. Local governments that met their agreed goals centives for good performance if repeated periodically. were rewarded in various ways, such as greater access to Finally, the experiment allowed the government to take funding and reduced MLG interference in their deci- fuller control of decentralization agenda from donors, sions. Most participating local governments improved who were excluded from the process. their fiscal performance considerably. Given the severity of the current economic and po- The STDP has several advantageous features. First, litical crisis, it is hard to know what the future will hold because it built on earlier efforts and slowly evolved over for Indonesia. Several developments, including the loss several years in close consultation with MLG officials, of East Timor as a province and the increasing uncer- STDP developed strong credibility with the local coun- tainty of the situation in Aceh (North Sumatra) and terparts, who increasingly accepted an approach based other areas, have placed increasing pressure on the new on limited, gradual, individually tailored reforms. Sec- Wahid government to move forward with economic and ond, the negotiation, rather than imposition, of reforms political reforms. Decentralization has clearly emerged placed some responsibility on local governments to meet as a genuine high-priority element of the reform pro- the goals they agreed to. Third, virtually all technical as- gram, and in 1999 the government laid the basic legisla- sistance personnel were local rather than foreign consul- tive groundwork for a new attempt at decentralization. tants, raising acceptance and credibility of the initiative There can be little doubt that a long-term, gradual, and among Kenyans. Fourth, STDP efforts generated infor- strategic process that builds on the experiment discussed mation that assisted the MLG to create procedural man- here is the likely way forward. uals and training materials for different types of local governments. These helped convince central and local Kenya: Tehnicaltingistcentroralization Reforofficials to accept the use of standards in making basic fiscal decisions. Finally, the ongoing efforts of the STDP Kenya is a country with a long history of local govern- over more than a decade raised the visibility of local gov- ment, but the system was weakened and heavily con- ernments and demonstrated that, with adequate assis- STRATEGIC FISCAL DECENTRALIZATION IN DEVELOPING COUNTRIES 105 tance and incentives, they could improve their long- collection and service delivery. Since the mid-1980s, unacceptable performance. Ceara has completely turned its negative reputation The Kenyan government, with substantial World around.21 It has been heralded as a success in the inter- Bank funding and technical assistance, has recently em- national press and development literature because of its barked on a broader program of reform-the Kenya recent efforts to reform certain decentralized services, Local Government Reform Program (KLGRP)-largely including health care, agricultural extension, business based on STDP experiences. The premise of KLGRP is extension, and employment-creating public works. that successful fiscal decentralization requires phased Many of the major factors underlying this impressive and integrated fiscal, legal, institutional, technical, and reversal are related to the altering of state-local relations procedural reforms at central and local levels. An Inter- and improving the environment in which public ser- ministerial Task Force on Local Government Reform vants function. First, the state played a significant role, manages KLGRP, building it gradually and systemati- rather than the typically prescribed reduced role, in pro- cally in a series of manageable, mutually reinforcing viding the "decentralized" services. This included sub- steps. Most KLGRP functions are the responsibility of stantial control over hiring and management of local civil servants, who also continue to perform many of government employees in key sectors, the creation of their regular duties. This establishes KLGRP as an inte- strong performance incentives, and the strengthening of gral and mutually reinforcing part of government func- local civil society in a way that led to a substantial in- tions rather than a separate activity. crease in public pressure for improved local government The configuration of KLGRP involves three major performance. phases. The first phase, which is essentially completed, Second, in contrast to the conventional emphasis of solidified agreement on key reforms, including local decentralization reformers on simple, standardized pro- managerial and procedural reforms, local tax reforms, cedures, local employees were given considerable discre- intergovernmental tax harmonization, the evaluation of tion in how to perform their responsibilities, and this options for intergovernmental transfers, and the resolu- often resulted in undertaking service delivery activities in tion of a serious intergovernmental debt problem. The more expansive, complex, and individualized ways. This second phase, which started in 1999, involves imple- was possible because workers were judged on the end re- menting basic initial reforms defined in the first phase, sults of their efforts-that is, services actually being de- including basic organizational and managerial reforms, livered to and well received by local residents-rather harmonization of key central and local revenues, and es- than on the specific means they used to provide services. tablishment of a simple intergovernmental grants pro- Third, the increased pressures from the state govern- gram. The third phase will involve implementation of a ment and from civil society-partially fueled by agents physical infrastructure program. Funding for basic in- of the state government-instilled a greater sense of job frastructure will begin after minimal reform require- commitment and engendered greater personal job satis- ments defined in the previous phases are met. As a pre- faction for local government employees. Local staff who condition to more substantial funding, an increasingly performed well were publicly recognized and rewarded, comprehensive set of reforms at both the central and enhancing their prestige.22 The efforts made by the state local level must be satisfactorily undertaken. to create greater visibility and legitimacy for the services Although the Kenyan government's relationship with being decentralized and the civil servants responsible for the international community has deteriorated in 1999- them also resulted in more informed and careful moni- 2000, this program continues to move forward slowly toring by local government constituents. with donor support. Kenya's experience illustrates that a This brief exposition greatly simplifies the story of gradual approach and negotiated packages of reform, Ceara's success, but it does highlight a situation in which supported by carefully targeted technical assistance, can some of the common prescriptions of "mainstream" de- have substantial benefits. This approach also recognizes centralization reform were absent, but good results were that decentralization involves slowly changing attitudes achieved. A higher level of government (the state) played about what central and local governments should do, a strong role in controlling local behavior and educat- and building their capacity to do it. ing citizens. Local governments were given discretion Brazil: Altering Intergovernmental Relations about how to undertake responsibilities because the and the Work Environment state judged performance on results rather than on adoption of specific procedures. Finally, the state en- The State of Ceara, located in Brazil's poor northeast re- sured that local civil servants were given recognition for gion, had a long history of clientelism and poor revenue good performance. Together these factors created an en- io6 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION vironment that was conducive to better performance NGOs. Thus, the LDF approach focuses on the devel- and enhanced local capacity over time. opment of local governments, which are given some choices, but they are subject to a degree of quality con- Cemamodi and Vietnam:MBalancing Public Service trol from higher levels on technical, financial, and gov- Demand and Supply Mechanisms ernance matters, as in the Kenyan and Brazilian cases. As noted here, many recent decentralization efforts have This basic approach is followed in both cases, but the focused on the development of demand-driven mecha- specific design accounts for cross-country differences. nisms, forcing participation to extremes and all but For example, the LDF in centrally planned Vietnam is ignoring the role of the supply side in providing infor- under the Ministry of Planning and Investment, and mation, standards, and technical assistance for local gov- commune-level governments, which have a long history ernments. In contrast, Cambodia and Vietnam have of administrative responsibility, serve as the focal insti- been experimenting with a decentralization mechanism tution locally. The LDF in Cambodia, which has a re- that tries to balance the demand and supply sides while cent history of ineffective central government that has incorporating elements of the gradual, strategic ap- at times been hostile to the people, was initially set up proach embodied in the Indonesian and Kenyan cases. under an interdepartmental management structure at The Cambodia and Vietnam experiences are based on the provincial level, essentially bypassing the center.24 the "Local Development Fund" (LDF) approach devel- In addition, the lowest tier of government, villages, oped by the United Nations Capital Development Fund which are closest to the people and were allowed to elect (UNCDF).23 The LDF shares some characteristics with representative committees before communes, initially a number of other types of programs used to support lo- played the dominant local role in Cambodia's LDF.25 cally generated and managed investments, including Thus, basic LDF principles can guide decentralization small town/rural development programs, the older fam- in countries where resources are limited and local gov- ily of Community Development Funds, and the more ernments are poorly developed, but the design must be recent Social Investment Funds and Micro-projects pro- tailored to local context. grams. These programs typically have a poverty allevia- tion focus, share a "demand-driven" approach, and tar- Tentative Lessons and Conclusions get local investments (health and education facilities, There is obviously no single approach to effective fiscal roads, water supply, income-generating community pro- decentralization that applies to all countries. Reforms jects, and so on) of a similar type and scale. must be determined in a particular case, given available Under most of these programs, various types of local resources and capacity, as well as the institutional, so- groups [local governments, nongovernmental organiza- cial, and political realities that govern an appropriate tions (NGOs), community groups, and so on] can ac- and realistic scope and pace for decentralization. Most cess resources for projects (typically though a program- countries, however, face some similar problems and con- specific planning process) by applying to a fund, usually straints in moving forward. A number of countries based in a national or a regional capital, that is often around the world have managed to develop in recent largely independent from the government structure. In years a variety of mechanisms that help to overcome, contrast, LDFs provide resources only to local govern- even if modestly, some of the formidable obstacles out- ments, "entitling" them with an objectively defined lined earlier. A number of broad tentative lessons about block allocation of resources (rather than a project- approaching decentralization, many of which are inter- specific grant) from a fund managed by a higher level of related, can be drawn from these experiences. government, with technical assistance. The LDF also in- Policy Framework: A solid policy framework that de- troduces a simple participatory local planning process, fines and guarantees the rights and responsibilities of but local governments use it to plan beyond the use of local governments can be an important starting point LDF resources, making requests for larger-scale projects for an effective decentralization program. The Ethiopian to higher levels of government and applying for other and Ugandan cases demonstrated, however, that such a available sources of funding. This process is demand framework and the commitment to implement it are driven, but technical, institutional, and political stan- not enough to ensure that decentralization will progress dards must be met before funds are released. Local gov- smoothly in practice. ernments have access to resources to obtain technical as- Process and Consensus Building: Developing a process sistance to meet these standards. They decide whether to build consensus about the appropriate extent, form, to obtain assistance through conventional channels and pace of fiscal decentralization is at least as impor- (agencies of higher level governments), private firms, or tant as-and initially often more imporrant than- STRATEGIC FISCAL DECENTRALIZATION IN DEVELOPING COUNTRIES 107 defining an "end-product" system that conforms to the an integrated approach that involves a mutually rein- normative decentralization models. This is particularly forcing set of reforms, as illustrated in various ways by critical in highly centralized countries where there is no the Kenyan, Brazilian, Cambodian, and Vietnamese unified vision of decentralization, and responsibilities cases, is more likely to yield better results. for local development are fragmented across central Balancing Roles among Levels of Government: Fiscal agencies, as in the Indonesian case. Such a process, if decentralization means more than just shifting power properly structured, representative of the key stakehold- from central to local governments. In fact, decentraliza- ers, led by a trusted or neutral agency, broadly perceived tion typically involves an increased initial role for cen- as fair, and adequately empowered, can help to define a tral or regional agencies. This reflects the need for a realistic starting point for fiscal decentralization. A do- greater balance between reforms on the supply and de- mestically led process can also reduce the risk that the mand side of service provision, as in the Brazilian, Cam- decentralization agenda will be dominated by outside bodian and Vietnamese cases. Consulting the consumer experts bearing substantial funds. on public expenditures can be critical, but reforming Coordinating Reforms: Given the complexity of de- procedures for delivering particular types of services in centralization, there are many opportunities for its op- an appropriate and cost-effective way is equally impor- ponents to undermine progress or its proponents to tant for success. move too quickly. Coordination is essential to ensure Building Capacity and Governance: A decentralization that decentralization proceeds in a rational and con- program must build the commitment and capacity of trolled manner. Responsibility for coordination can be local governments to assume new responsibility on be- under the same type of interinstitutional mechanism half of their constituents. Disillusioned local residents that was recommended above to negotiate the initial de- have to be gradually convinced that their local govern- centralization reforms. Such a mechanism would need ments can and will respond to them, and this requires to develop checks and balances among the various in- some concrete results in the form of improved services, terested organizations to reduce power imbalances that as in the Brazilian and Cambodian cases. As already can undermine decentralization. Once agreed upon, the noted, such improvements are more likely to be attain- responsibilities of all relevant actors should be formal- able if decentralization reforms are phased in with ca- ized, and an effective system of monitoring and enforc- pacity building and incentives, making it clear to local ing compliance is required. governments what they must accomplish before receiv- Strategic Implementation: An effective decentraliza- ing additional responsibility or resources, as in the tion program requires a strategic implementation ap- Kenya case. Without increased local accountability and proach, as in all of the cases discussed here. Even capacity through political and procedural reform, de- though a clear vision of longer-term reform is needed, centralization is ultimately a meaningless exercise. initial steps should be modest and logically phased. Re- EnsuringAdequate Flexibility: A pragmatic fiscal de- forms that have the greatest possibility of relatively centralization program should judge service providers rapid success should be undertaken first, primarily sim- and employees on the basis of results, not on slavish ad- ple tasks that don't immediately threaten in a signifi- herence to rigid bureaucratic procedures. This requires cant way the central power base or overwhelm local ca- a process that provides incentives for good performance, pacity. In addition, it can be useful to differentiate but that also allows flexibility, so that local governments among local governments, a highlight of the Indone- can operate in a customized way if this is appropriate. sian and Kenyan approaches. Some are more capable The potential success of such a strategy was particularly and can be given greater responsibility, while others will highlighted by the Brazilian case. require substantial technical assistance and oversight. As economic and political pressures for decentraliza- Starting carefully and modestly should raise the pros- tion and local government reform continue to escalate pects for initial success, creating a stronger base on which and forces driving democratization continue to develop, to build. many countries will feel an increasing urgency to for- Integrating Reforms: Effective fiscal decentralization mulate decentralization policies that produce good re- normally requires a focus broader than conventional fis- sults. There is much to learn from the cases and princi- cal concerns. Increasing local spending responsibilities ples discussed here, but there are still many gaps in our and access to revenues will be unlikely to result in im- understanding of how to design and implement effec- proved performance unless supporting institutional and tive decentralization policies. The numerous decentral- political reforms are also undertaken. Many countries ization activities currently under way around the world fragment different types of decentralization reform, but provide a significant opportunity for further experimen- io8 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION tation and research to improve our knowledge of how Smoke, P., and B. Lewis. 1996. "Fiscal Decentralization in In- to approach this challenging and potentially productive donesia: A New Approach to an Old Idea." World Develop- type of public-sectorreform. ment24(8), pp. 1281-1299. Tanzi, V. 1995. "Fiscal Federalism and Decentralization: A Re- Bibliography view of Some Efficiency and Macroeconomics Aspects." Pro- ceedings of the Annual World Bank Conference on Development Bahl, R. 1997. "Fiscal Federalism in Uganda." World Bank, Economics, pp. 295-316. Washington, D.C. Processed. Tendler, J. 1997. Good Government in the Tropics. Baltimore, Bahl, R., and J. Linn. 1992. Urban Public Finance in Developing MD: Johns Hopkins University Press. Countries. Baltimore, MD: Johns Hopkins University Press. Tiebout, C. M. 1956. "A Pure Theory of Local Expenditure, Barnes, N. 1998. "How Local Can We Go? Lessons from Fiscal Journal of Political Economy 64(5), pp. 416-424. Decentralization in Uganda." Cambridge, MA: Department Raimondo, H. J. 1992. Economics of State and Local Government. of Urban Studies and Planning, MIT. New York: Praeger. Bird, R 1993. "Threading the Fiscal Labyrinth: Some Issues in Fis- UNCDF (United Nations Capital Development Fund). (1996) cal Decentralization." National TaxJournal46(2), pp. 207-227. Local Development Funds. New York: United Nations Devel- Bird, R., and F. Vaillancourt 1998. "Fiscal Decentralization in opment Programme. Developing Countries: An Overview." Discussion Papers, No. UNCDF 1999. Taking Risks. New York: United Nations Devel- 11. Toronto, ON: International Center for Tax Studies, Uni- opment Programme. versity of Toronto. Wilson, R. H., and R. Cramer. 1995. International Workshop on Cohen, J. 1995. "Ethnic Federalism in Ethiopia." Development Good Local Government. Austin, TX: Johnson School of Pub- Discussion Paper. Cambridge, MA: Harvard Institute for In- lic Affairs, University of Texas. ternational Development. Cohen, J., and S. Peterson 1999. Beyond Administrative Decen- Notes tralization: Strategies for Developing Countries. West Hartford, 1. The cases discussed here are based on programs financed by CT: Kumarian Press. the World Bank, the United Nations Capital Development Fund Dillinger, W. 1995. Better Urban Services: Finding the Right In- (UNCDF), the United Nations Development Program (UNDP), centives. Washington, D.C.: World Bank. the Deutsche Gesellschaft fui Technische Zusammenarbeit Fisher, R. C. 1996. State and Local Public Finance. Chicago, IL: the uts. Gency forITernaioaDevelopmenaIDei Irwin. (GTZ), the U.S. Agency for Internatonal Development (USAID), Khellaf, A. 1992. "Decentralization and Centralization of Local and various governments. Some findings are based on work con- Khellaf, Services 1992. "Deentaizatn Camidg : C zartment of Lo ducted for the Management Development and Governance Divi- Public Services in Tunisia." Cambridge, MA: Department of sion of UNDP. The perspectives expressed here are those of the UJrban Studies and Planning, MIT. author and should not be attributed to any of these organizations. Leonard, D.K. 1987. "The Political Realities of African Manage- 2. A substantial literature on the benefits of decentralization ment." World Development 15(7):899-910. dates back to the 1960s. Much of it is reviewed in Smoke (1994). Litvack, J., J. Ahmad, and R. Bird. 1998. Rethinking Decentraliza- More recent literature is discussed in Smoke (1999). tion in Developing Countries. Washington, D.C.: World Bank. 3. See, for example, Bird (1993), Prud'homme (1995), And Manor, J. 1998. The PoliticalEconomy ofDecentralization. Wash- Tanzi (1995). ington, D.C.: World Bank. 4. Much of this literature is reviewed in Smoke (1994). Other Ndegwa, P., et al. 1987. Managementfor Development: Priority relevant literature includes Dillinger (1995); Tendler (1997); ThemesforAfrica Today. Nairobi: Oxford University Press. Litvack, J., J. Ahmad, and R. Bird (1998); and Cohen and Peter- Oates, W. 1972. Fiscal Federalism. New York: Harcourt, Brace, son (1999). and Jovanovich. 5. These models originated with a few seminal works, includ- Prud'homme, R. 1995. "The Dangers of Decentralization." The ing Tiebout (1956) and Oates (1972). Reviews of how these World Bank Research Observer 10(2), pp. 201-220. basic models have been applied, expanded, and empirically tested Rasheed, S., and D. F. Luke, eds. 1995. Development Manage- are found in Bahl and Linn (1992), Raimondo (1992), and ment in Africa: Toward Dynamism, Empowerment and En- Fisher (1996). trepreneurship. Boulder, CO: Westview Press. 6. See Smoke (1989 and 1994), Linn (1992), Bird (1993), Smoke, P. 1989. "Is Local Public Finance Theory Relevant for Prud'homme (1995), Tanzi (1995), and Bird and Vaillancourt Developing Countries?" Development Discussion Papers, No. (1998). 316. Cambridge, MA: Harvard Institute for International 7. Many of these weaknesses are elaborated in Smoke (1994); Development. Litvack, Ahmad, and Bird (1998); Manor (1998); and Cohen Smoke, P. 1994. Local Government Finance in Developing Coun- and Peterson (1999). tries: The Case of Kenya. New York: Oxford University Press. 8. See Leonard (1987), Ndegwa (1987), and Rasheed and Smoke, P. 1999. "Fiscal Decentralization in Developing Coun- Duke (1995); tries: A Review of Current Concepts and Practice." Geneva: 9. There are some cases in which significant political motives United Nations Research Institute for Social Development. for decentralization have developed, including Ethiopia, Indone- Smoke, P. Forthcoming. "Rebuilding Local Government in sia, South Africa, and Uganda. Kenya." In J. Wunsch and D. Olowu, eds., Building Democ- 10. See Manor (1998). racy in Africa: Polycentric Strategies and Experiences. Boulder, 11. See Khellaf (1992); Smoke (1994); and Lirvack, Ahmad, CO: Westview Press. and Bird (1998). STRATEGIC FISCAL DECENTRALIZATION IN DEVELOPING COUNTRIES lO9 12. An example would be the reform of local government decentralization was technically occurring, but without devel- accounting practices. oping the links between local governments and their con- 13. Some of the cases discussed here are more detailed than stituents that underlie the public choice models used to justify others are, either because more information was available or be- decentralization. The situation is almost certainly worse in cause certain aspects of their performance were considered par- Ethiopia. ticularly relevant. 19. The Indonesia experience is examined in Smoke and 14. There are several good references on Ethiopian and Ugan- Lewis (1996). dan decentralization policy, including Cohen (1995); Bahl (1997); 20. See Smoke (1994) and Smoke. (forthcoming). World Bank public expenditure reviews for Ethiopia and Uganda, 21. The Brazilian experience is examined in Tendler (1997). 1994-98; and a 1999 World Bank study on regionalization in Tendler links her major findings closely to the literature on in- Ethiopia. dustrial performance and workplace transformation. 15. Both countries independently adopted decentralization 22. A key element in recognizing good local performance to hold their ethnically fragmented nations together during a pe- was a system of competitive prizes. Similar mechanisms are used riod of crisis. This central government attitude toward local gov- in a number of countries. See, for example, Wilson and Cramer ernments is particularly interesting because the historical (post- (1995). independence) trend in many postcolonial developing countries, 23. The LDF approach is discussed in considerable detail in particularly in Africa, has been to increase state control in order UJNCDF's 1996 publication Local DevelopmentFunds. UNCDF's to build national unity. 1999 publication, Taking Risks, reviews experiences with the 16. In the Ethiopian case, there is also inadequate specificity LDFs to date. As of mid-1999, LDF programs have been or are in the constitution and relevant legislation about the functions being designed in more than a dozen countries (Vietnam, Cam- to be decentralized to tiers below the regions (the zones and bodia, Bangladesh, and Nepal in Asia; Palestine in the Middle woredas). There is no reference to the municipalities, which are East; Ethiopia, Uganda, Tanzania, Malawi, Zambia, Equatorial important in the urban areas. Guinea, Mozambique, and Senegal in Africa). Because UNCDF 17. In some sectors, constitutionally mandated decentraliza- can focus only on the least developed countries, the LDF is de- tion occurs slowly because a ministry delays moving forward, signed primarily to promote decentralization of rural and small- such as the case of public works in Ethiopia. In other sectors, town infrastructure-primarily through grant mechanisms, but responsibility has been devolved too rapidly by ministries at- the basic model for building capacity and accountability has tempting to comply with the letter of the law, overwhelming the broader relevance. capacity of local governments, such as the case of education in 24. The Cambodian LDF is now being "mainstreamed" Uganda. under an interministerial mechanism at the central level that is 18. A recent study of "successful" local governments in to be funded by multiple international agencies and the Cambo- Uganda, for example, demonstrated that local government dian government. revenue yields and local service expenditures increased after 25. This will change when commune governments are elected decentralization. See Barnes (1998). Further analysis, however, under a decentralization law that is expected to take effect in the revealed that people paid local taxes primarily because they year 2000. As of May 2000, the Law on Commune Administra- were being forced to do so by aggressive enforcement officers tive Management had passed the Council of Ministers and was rather than because they were satisfied with the types and lev- being referred to the Parliament. The Law on Commune Elec- els of services being provided. Thus, in this anecdotal example, tions was still being finalized by the Ministry of Interior. chapter Decentralization Comes to Japan Kengo Akizuki Kyoto University ~ lobalization tends to weaken the no longer be that of master and servant. Local govern- credibility of national governments ments will now perform "legally entrusted functions" and their willingness to intervene in (hoteijutakujimu), such as administering national elec- local government. Japan, which has tions, collecting statistical data, and implementing na- maintained one of the world's most tional welfare programs. In cases of conflict the CPD centralized governmental systems, at proposes a system of mediation, and parties not content least on paper, is no exception. with the results may initiate lawsuits. The commission The Japanese intergovernmental system has under- has been strongly criticized for not addressing the issue gone fundamental reform since the mid- 1990s, with the of financial dependence of local government on the cen- basic trend to loosen central control. The country's pro- tral government-the most difficult challenge of all- longed economic difficulties have undermined the cen- but some members say the commission will do so by the tral bureaucracy's credibility, and the public has become end of its tenure in 2000. more aware of bureaucratic corruption. Most important, In response to such changes, many localities are pur- globalization has limited the ability of the national gov- suing new approaches to governance that tap the world- ernment to control movement of the country's popula- wide exchange of information but that promise to pro- tion, information, capital, services, and goods. Accord- duce winners and losers, threatening Japan's postwar ing to a common slogan, it is high time for Japan to shift commitment to economic equality. Japan's experience "from the days of uniformity under centralization to va- reflects its own strengths and idiosyncrasies as well as the riety under decentralization." challenges that all governments, central and local alike, In May 1998, the cabinet formally authorized a De- confront under globalization. centralization Promotion Plan, proposed by the Com- mittee for Promotion of Decentralization (CPD), which National Party Conflict Drives Decentralization affects virtually all ministries and agencies. This scheme The reform process resulted from political turmoil entails amending at least 500 laws, decentralizing au- in the central government, beginning when the Liberal thority, and deregulating functions from the public to Democratic Party slipped from the ruling position in the private sector. The most symbolic yet best known re- 1993 for the first time since its birth in 1955. This tur- form entails abolishing "agency-delegated functions" moil brought several political aspirants to the top before (kikan i'fninjimu), a legal concept that enables the na- they had become completely "nationalized." Until then, tional government to require local governments to do a candidate for a cabinet post had to serve at least six its bidding outside the purview of local assemblies. In- terms in the Lower House of the Diet; a candidate for stead, the national government will give technical ad- prime minister had to hold office for 25 years and as- vice, ask for reports, and if necessary, even provide di- sume important posts such as minister of finance or rection, but the relationship between the two levels will international trade and industry. The new prime minis- IIO DECENTRALIZATION COMES TO JAPAN III ter, Morihiro Hosokawa, was young and lacked legisla- (The legal status of the Tokyo metropolitan government tive experience-he had served as governor of Kuma- has fluctuated, and the U.S. military once controlled moto Prefecture-and made clear that decentralization Okinawa; both are now prefectures.) However, postwar would be his top priority. When the LDP regained con- reforms driven by U.S. occupying forces did bring sig- trol in 1995 by forming an unlikely coalition with the nificant changes such as the direct election of prefecture Socialist Party, Tomiichi Murayama, a veteran Socialist governors. politician originally from a local government union, be- The performance of the prefectures has often drawn came prime minister. criticism. Some doubt that dividing Japan into 47 pieces Murayama established the CPD and appointed its can yield administrative efficiency; the business sector members, who largely represent business, academia, and has been the most eloquent critic in preferring regional the press and favor decentralization. Six nationwide government. Some experts in forestry conservation, groups of local leaders have voiced support for the com- river improvement, and other environmental issues also mission and maintain pressure for reform. The commis- maintain that convincing prefectures to address such sion has concentrated on refuting the "who-would-fill- concerns is difficult. Critics further maintain that pre- the-vacuum" argument advanced by skeptics of local fectures play a double-edged role, representing and sup- autonomy. Decentralization sounds fine, these skeptics porting cities and towns but also controlling them-a say, but localities need guidance and protection from the plausible criticism given the exchanges of personnel be- national government to get the job done. tween national and prefectural governments. Because of such doubts, former would-be reformers Despite these criticisms, prefectures have played an often felt they had to revamp the three-tiered Japanese important role in Japanese democracy. One dramatic system of government-which includes national, pre- change occurred when leftists took over virtually all im- fectural, and local levels-into two or four tiers and portant prefectural governorships in the 1960s and hope that somebody or something would fill the vac- 1970s. These so-called progressives could not be consid- uum. Those ideas never won acceptance. But the CPD ered mere national agents, and kept the intergovern- maintains that there is no need to change the number mental games going through careful negotiation-and of tiers, rewrite the Constitution to reflect a federal sys- sometimes confrontation. They also responded to citi- tem, or combine prefectures into regional governments. zens' demands for consumer and environmental protec- The current actors, says the CPD, can fill the vacuum. tion, partly by retaining access to the national govern- For example, the CPD calls for more active local assem- ment through the bureaucrats dispatched to them. blies in formulating policy and overseeing the local bu- The profile of prefecture governors is rising these reaucracy and chief executives. And instead of relying days, and they are now acting even more boldly. For ex- on the national government to supervise the activities ample, the governor of Saitama, formerly the speaker of of local governments, citizens will do so through infor- the Upper House of the Diet, sent back the personnel mation disclosure and referendum and recall. dispatched from the national government. Some gover- nors are pursuing an alternative strategy of recruiting national bureaucrats to serve the prefecture directly. The CPD focuses on prefectural governments as key to Thus, prefectures have benefited from evolution as coordinating the activities of cities and towns and dele- well as consistency. To increase their administrative ca- gating authority from the national to the local level, and pacity, villages amalgamate to become towns, and towns from the public to the private domain. This represents become cities, constantly changing names, size, and legal the first time that postwar governmental reform does not status. Citizens usually identify themselves with prefec- propose abolishing or merging prefectures. Two preemp- tures. "Character of prefecture" (Kenminsei) is a com- tive moves have convinced the CPD and the national mon notion: Men from Kochi prefecture drink a lot, government to make prefectures a partner of the center. people of Shiga are stingy, and the wives of Gunma rule Two prominent ex-governors, one an urban liberal with their families. Doctors' organizations, bar associations, leftist leanings, and another a former rural police bu- and farmers' unions also organize along prefectural reaucrat, sit on the commission. And prefectures have boundaries. The CPD's decision to put prefectures in initiated studies and issued proposals on how to cope charge of decentralization formally confirms this status. with the coming decentralization since the mid-i 980s. Japan has 47 prefectures. Their number and borders The Changing Style of Japanese Governance have remained largely unchanged since Aritomo Yama- Although Japan clearly maintains more central control gata formulated the prefecture system in the Meiji era. and overlapping jurisdictions than many countries, 112 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Table 13.1 megaprojects, and they oppose not just the specific Percentage of Public Revenue after Fiscal Transfer schemes but also the fact that once initiated, these schemes are almost impossible to stop. Even if some state/ county/ projects are rational responses to credible needs, critics National local province city/town say, these needs may evaporate during the decade or Federal more needed to complete these projects. Canada 38 62 43 19 In response, Hokkaido prefecture has initiated a new FRG 39 61 32 29 approach called "assessment by time," as well-known Unitary writer Soh Kuramoto nicely named it. A committee UK 71 29 headed by the governor evaluates projects according to Francen 72 28four standards: necessity, validity, priority, and effect. cPrefecture village The committee releases the results and posts them on Japan 38 62 30 32 the Internet. The Hokkaido government has so far se- Source: Mikiko Iwasaki, Politics of Subsidies in Japan (1990). lected only seven big projects for review but pledges Notes: Fiscal year 1985 for Japan, 1986 for other countries. eventually to apply this scheme to all projects. In another approach, Mie and Shizuoka prefectures have formulated an evaluation system that attempts to transfers of funds from central to local governments weigh the purpose and progress of virtually all govern- are also higher than in most countries, and the system ment tasks. Every agency must make an inventory three clearly allows for local initiative. In fact, Japan's fiscal re- times a year of the goal of every task, the time period and sponsibilities clearly follow the pattern of countries with person-hours required to complete it, and the funds en- a federal system (Table 13.1). However, that record leads tailed. Since Mie Governor Masayasu Kitagawa began to another difficult challenge: Each level of government this system in 1996, the prefecture has eliminated 740 must cut the significant public-sector deficit. tasks. Shin-ichi Ueyama, a consultant with the interna- All levels of government are, therefore, now attempt- tional firm McKinsey & Company and the leading pro- ing to restructure the bureaucracy, privatize some func- ponent of such schemes, criticizes these two efforts be- tions, and rely on partnerships between citizens and cause they exclude private citizens and public interest nonprofit groups for providing services and collecting groups from directly participating. However, he applauds information. Prefectures may well serve as coordinators the prefectures' willingness to disclose the information. and shock absorbers during this process, as some Amer- Local Japanese government is considered severely lim- ican states did during the federal budget cuts of the Rea- ited in its ability to revamp itself, as laws require locali- gan era. ties to have at least one vice-mayor, for example. How- However, new styles of governance are rapidly emerg- ever, in 1997, Shiga prefecture upset the conventional ing among Japanese localities which go well beyond the wisdom that localities cannot reorganize without help or need for budget cuts. Because the national government a nod from the central government by creating a new di- will no longer protect localities, these are trying to sat- vision called the Lake Biwa Environment Bureau. En- isfy their other bosses-their inhabitants. Because local- compassing 674 square kilometers, Lake Biwa is the ities do not possess enough financial resources to sim- largest lake in Japan and supports the daily life of more ply expand services, they are experimenting with new than 13 million people in seven prefectures. Shiga had tools, many invented in Anglo-Saxon countries. been seeking more comprehensive policymaking author- Two approaches to administrative evaluation (gyosei ity over the lake because it occupies one-third of the pre- hyoka) are becoming popular among local officials. fecture, yet four different bureaus, all associated with na- Today large construction projects are encountering pub- tional agencies, controlled it. The new bureau combines lic resistance. The Kobe municipal offshore airport, for sections from those four agencies and asserts Shiga's pri- instance, a longtime pet project of city government, is ority in protecting the lake and its surroundings. Al- under serious attack from citizens as well as outsiders. though Shiga expected strong opposition to this initia- More than 30 percent of the city's voting population has tive from central bureaucracies, they remained silent. signed a petition favoring a public referendum on the project. Questions posed by local governments, commu- Mechanisms for Change nity leaders, and citizens and landowners have also A prefecture's ability to retrain its staff is essential to halted construction of several dams planned by the cen- such administrative reforms. All prefectures and some tral Ministry of Construction. Critics maintain that fis- cities maintain independent training programs and fa- cal difficulties and environmental issues preclude such cilities-particularly important for rural areas. The pro- DECENTRALIZATION COMES TO JAPAN I13 grams have also strengthened connections between local The nation is already talking about pioneers and late- governments and universities and colleges. These uni- comers in improving governance. Pioneers tend to be versities provide teachers and policy consultants and suburban/semirural prefectures. Where is Tokyo metro- help develop new ideas or governance. A typical case is politan government-the largest in population and fi- Shizuoka prefecture, which tapped a university profes- nances? Where is Osaka City, known for its ambitious sor to organize its evaluation scheme. prewar development policy and postwar high-speed Leadership by governors is another key to reform, as growth? These big urban localities grew by initiating in Shizuoka, where the governor urged bureaucrats to new programs, expanding their functions, and employ- publicize their assessments. Mie Governor Kitagawa, ing more officials. They have strong unions, formidable formerly a successful Diet member, presented a clear and contentious assemblies, a critical local press, many agenda of administrative reform and is now considered local actors, and proud governmental staff. This accu- a champion of such efforts. In fact, driven by top lead- mulated tradition could work against them. ership, some local governments seem to be giving birth In the past, local competition did occur, such as when to a new management class composed of specialists from localities fought to win the designation of "new indus- consulting firms and universities. Whether this coalition trial city" in the 1960s, but this race occurred in an era will supplant older coalitions of labor, bureaucrats, and of increasing expansion. Today market forces will choose service providers depends on circumstances unique to which localities can deliver effective policies, and decen- each locality-the front lines of this significant political tralization will determine outcomes in arenas ranging battle. from environmental policy and consumer protection to Many Japanese localities study local experiments in taxes and finance. Localities can expect more autonomy, foreign countries, from benchmarking in Oregon, which but they will receive much more scrutiny. projects the future of the state by designating 92 indica- Local governments, especially prefectures, occupy a tors and setting numerical targets for each indicator, to strategic position in Japan, yet converting that position administrative reform in New Zealand. Books, confer- into effective policymaking presents a formidable chal- ences, the Internet, and informal networks of public of- lenge. No one can ensure competent political leadership. ficials allow the exchange of ideas as well as results. Some Democracy, after all, does not provide answers but skeptics say it is absurd to follow in the wake of coun- rather demands ongoing efforts from the governing tries with different cultures, histories, and institutions. elite, midlevel officials, as well as the voting public. They are usually right. However, total quality manage- ment (TQM), which local Japanese governments are Bibliography now discussing, is proving worthwhile investigating. As Akizuki, Kengo. 1995. "Institutionalizing the Local System: The is widely known, TQM was conceived of in the United Ministry of Home Affairs and Intergovernmental Relations States but further developed in Japan. In today's hyper- in Japan." In Hyung-ki Kim, T. J. Pempel, Kozo Yamamura, globalized world of ideas, even overly cautious public and Michio Muramatsu, eds., The Japanese Civil Service and g ~~~~~~~~~~~~~~~~~Economic Development. Oxford: Clarendon Press. officials in Japan are looking into this new venture. Akizuki, Kengo. Forthcoming. "Partnership in Controlled De- Some localities are also aggressively taking up the centralization: Local Governments, The Ministry of Home globalization challenge. Kanazawa and several other Affairs, and Intergovernmental Linkages in Postwar Japan." cities have established closer ties with Russia (Siberia), Washington, D.C.: World Bank. for example, while Osaka, Fukuoka, and other western Omori, Wataru. 1998. Bunken Kaikai to Chiho Gikai (Decentral- have invited investment from China. This kind of ization Reform and Local Assembly). Tokyo: Gyosei Uapanese). cities nave mvltea mvestment trom tuna. l ms tna ot Osborne D., and T. Gaebler. 1992. Reinventing Government. strategy is not confined to industrial areas: Shiga spon- Reading, MA: Addison-Wesley. sored the World Lake Conference while the city of Peters, B. Guy, and D. J. Savoie, eds. 1997. Governance in a Kyoto organized the Historical Cities Roundtable- Changing Environment. Montreal: Canadian Centre for Man- both designed to enable localities rather than central agement Development. governments to confer on common problems. Reed, Steven R. 1986. Japanese Prefectures and Policymaking. Pittsburgh, PA: University of Pittsburgh Press. Winners and Losers Rhodes, R. A. W. 1997. Understanding Governance: Policy Net- works, Governance, Flexibility and Accountability. London: Japan's postwar intergovernmental system has commit- Open University Press. ted itself to the idea that no regions or communities Samuels, Richard J. 1983. The Politics of Regional Policy in Japan. should be losers, but decentralization emphasizes com- Princeton, NJ: Princeton University Press. petiton an tend to poducewinnes andloser. Is Stoker, Gerry. 1998. "Governance as Theory: Five Propositions." petition and tends to produce winners and losers. Is UNESCO paper, Oxford. Japan ready for intercity competition and regional eco- Ueyama, Shin-ich. 1997. Gyosei Hyoka no Jidai (The Age of Pub- nomic disparity? lic Evaluation). Tokyo: NTT Press Uapanese). chapter 14 Why Are Social Funds So Popular? Judith Tendler' Department of Urban Studies and Planning Massachusetts Institute of Technology -. hy are Social Funds (SFs) so popular Since 1987, the World Bank (WB) has committed among international donors, who US$1.4 billion worldwide to SFs, the Inter-American have committed more than US$3 bil- Development Bank (IDB) has committed almost the lion to their creation and perpetua- same in Latin America (US$1.3 billion), and in Latin tion in 40 countries since the late America alone, the European donors have committed 1980s?2 Ostensibly, SFs seem to do US$570 million. The pace of commitment shows no sign several things that are dear to our hearts. They are said of slackening, with a new round of SFs cropping up in to reduce poverty and unemployment and to bring ser- reform packages for crisis-afflicted Asian economies. vices and small works projects to myriad poor commu- Roughly one-third of SFs go to economic infrastructure; nities in a way that is decentralized, demand-driven, another third to education, health, nutrition, and popu- participatory, low in cost, and fast-disbursing. By the lation activities; and another third to miscellaneous activ- donors' own accounts, however, SFs do not live up to the ities such as microfinance, training, and environmental faith placed in them. This chapter explores this conun- interventions. Social Funds are administered by a sepa- drum. More important than any particular judgment on rate agency or unit within a department of the central SFs,3 it seeks to assess the SF experience in a way that government. They allocate grant funds directly to myriad contributes to the larger debates about improving the communities for projects often chosen by them; the fund- quality of public-service delivery in developing countries. ing passes through local governments in only a minority Social Funds started in Latin America in the mid- of cases. Some of the acclaimed strengths of SFs are based 1980s as a temporary antidote, according to the lore, to on a general set of arguments about the superiority of more the adverse impact of structural adjustment programs and decentralized and "demand-driven" approaches in con- other reforms on the poor.4 Originally, the programs were trast to traditional public-service supply, with its "supply- meant to provide quick employment through public- driven" set of problems-overcentralization, rigid and works projects and emergency social services, particularly top-down bureaucracy, and insensitivity to service users. in rural areas and towns, and partly in lieu of the increas- At the start, donors and the borrowing governments ingly faltering presence of fiscally strapped line ministries. viewed SFs as a safety net for ameliorating the harsh ef- Some were designed explicitly to compensate for layoffs fects of structural adjustment on the poor. More recendy, caused by downsizing of the public sector and its state en- SFs have also come to be seen as a refreshing new model terprises. By the early 1990s, donors judged the SFs to be of service delivery for poor communities, with their more so effective at temporary relief, and so appealing as a dif- independent project agencies or units and their involve- ferent and low-cost model of public-sector service deliv- ment of beneficiary communities in choosing, monitor- ery, that they provided follow-on funding to several SFs ing, and financing projects. In typical descriptions, donor and started new ones, expanding first to Africa, and then narratives commend the SFs as "an imaginative effort to to Eastern Europe and Southeast Asia. make government actions and resources more beneficial "4 WHY ARE SOCIAL FUNDS SO POPULAR? 115 to the poor"; as having "considerable potential ... for sus- ery-just the opposite of the stereotypical government tainable service delivery ..." (italics mine); and as suc- agency.1 Several organizational traits are considered key ceeding, often, "in targeting the poor and in providing to this success. As noted previously, SFs are often run by basic services more cheaply5 and speedily than public sec- semiautonomous units or agencies operating outside tor agencies that have traditionally been charged with line agencies, sometimes newly created, and often close these functions."6 and accessible to the office of the country's president. This chapter draws mainly on evidence about SFs They also work outside civil service regulations, partic- from the donors themselves,7 in addition to findings ularly in hiring, firing, and setting salaries. They have from outside researchers8 and from my own research ex- therefore often been able to recruit excellent managers perience with four SFs in northeast Brazil.9 The evi- with experience in the private sector or to lure to their dence raises questions about the effectiveness of SFs as a staffs the best of the public or NGO sector. They have safety net for the poor and, more significantly, about the also succeeded in operating outside government pro- presumed greater desirability of SFs as an alternative to curement regulations, simplifying and speeding up the traditional government supply, or reformed versions of construction of small works projects. it. I therefore question the large amounts of funding and Social Funds vary considerably from one country to enthusiasm dedicated to SFs, as well as the distraction the next, sometimes even on the aforementioned key of attention and the diversion of resources from other traits.12 Although the variety among SFs sometimes paths toward these ends. makes it difficult to describe this group of programs in Many SF goals and shortcomings are not peculiar to uniform terms, I follow the convention of the donors in them. Some of their main problems also plagued earlier, treating and praising these projects as a single category more supply-driven and centralized public programs, as called Social Funds. To the extent that this chapter well as recent attempts to decentralize power and re- dwells on issues relating to sustainability, participation, sponsibility to local governments, to rely more on mar- and demand-driven decisionmaking, it relates more ket mechanisms and nongovernmental providers at the to the ongoing or more permanent funds than to the local level, and to involve citizens in user choice and earlier temporary and emergency funds. (Many of the monitoring. I hope, then, that this chapter will interest latter evolved into the former.) In the same vein, I pay the development community in expanding its thinking relatively more attention to the claims about SFs as beyond SF-type models to other ways of improving gov- a service delivery model-sustainable, decentralized, ernment performance. These ways may not be as new demand-driven-than to their effectiveness as a safety and exciting as SFs but, at the least, they would seem to net for the poor. have no more frailties than SFs do. For all the talk of SFs as decentralized, they are in some ways the opposite of real decentralization, even T he Claims - when they are demand-driven. They are run by entities Although SFs vary widely across countries, they tend to of the central government (state government in my have several common components. First, they make Brazilian cases), which are often newly empowered and grant funds available to enable communities or muni- transformed by their association with international cipal councils to choose among a menu of possible proj- donors and by the corresponding direct support they re- ects (such as a well, a health center, a school, a grain mill, ceive from the president of the borrower country. The or road repair). Second, project design and construction majority of SFs do not formally devolve power and re- are partly outsourced to local actors ("partially priva- sponsibilities to local governments. As agencies of the tized")-private firms, nongovernment organizations central government, they try to reduce their overhead (NGOs), community associations, and sometimes local and personnel costs by deploying some of their staff out- governments; community groups themselves find and side the capital and devolving some responsibility to contract a design or construction firm or equipment sup- community groups and local providers. But this is de- plier, monitor project execution, and/or take some re- scribed more accurately as "deconcentrating" rather than sponsibility for operating and maintaining the project. as "decentralizing" responsibilities and finances. Decon- Third, a local contribution is often required, roughly centration, of course, shares some of the traits and pur- 10-15 percent of project costs. Together, these features ported advantages of decentralization. But "deconcen- add up to the "demand-driven" moniker by which these tration" does not have the cachet that decentralization programs are often described.10 has: It does not fit the image, now so popular, of "tak- Donor evaluations characterize the SF success as a ing power away" from central government, of moving fast, flexible, and low-overhead model of service deliv- from a "top-down" to a "bottom-up" style. ii6 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION In the small number of cases where SFs do work employment-creating programs deliberately set wages so through local governments, they often do so ad hoc, not low as to not draw labor away from private-sector em- as part of a larger decentralizing effort nor following the ployers and to keep the nonpoor from applying for these fixed criteria of geographical allocation typical of systems jobs. At the same time, however, the lower-than-subsis- of revenue sharing with local government. In the former tence level plus the short-term nature of the job add up cases, SFs channel resources only through some local to a weak instrument for reducing poverty and unem- governments, while avoiding others. This happens ac- ployment in a more sustained fashion. In the same vein, cording to one or two informally held sets of criteria; one the voluntary labor often required of communities for involves the perceived capability and "seriousness" of the SF projects, although meant to serve the goal of reduc- particular government, and the other with whether it is ing costs and eliciting community "ownership" of the run by the same party of the government or the opposi- project, represents a regressive tax on the poor.19 tion. Where a local government is run by the opposition The donor evaluations do not really bear out the claim party, the political calculus of the central government that SFs are particularly good at reaching poor commu- may dictate that the SF circumvent it and work directly nities, or the poor within them.20 Higher per capita SF with communities-or not operate there at all-while expenditures often go to communities or provinces that working through local governments in the districts are better off than to the poorer or the poorest.21 In its headed by the party faithful.13 It is for this reason, among 1997 study of four countries with SFs (Bolivia, Egypt, others, that some view SFs as working at cross-purposes Sri Lanka, and Zambia), the World Bank found that "the to decentralization reforms, or at least not advancing the higher poverty headcount index of the province, the cause of decentralization more generally. lower was the actual per capita SF expenditure it received; or the actual expenditures lagged behind allocations in The Evidence: Reducing Poverty the areas with the highest poverty index while they far ex- and Unemployment ceeded allocations in areas with low poverty indices."22 By the donors' own accounts, SFs do not live up to the Even in the "star" Bolivian SF, the richest of five income faith placed in them. They have contributed only in- areas received two-and-a-half times as much SFs funds significantly to the reduction of unemployment and per capita as the poorest five (US$25 vs. US$10).23 There poverty in the countries where they operate, even when are some exceptions,24 though most of the data cannot compared to other programs with the same goal. The distinguish between poor and rich within municipalities Latin American SFs are reported to have "created rela- or other administrative units. tively few jobs" and reached only a small fraction of the Some researchers comparing SFs to other programs labor force (in the Latin American case, less than 1 per- have found that certain more traditional targeted pro- cent at best).14 They also devoted only 30 percent of grams have actually had more identifiable impacts in re- their expenditures to labor costs, a low share for pro- ducing poverty. They refer to employment-guarantee grams dedicated to employment creation.15 Jobs offered schemes, food stamp programs, food commodity pro- by the SFs were temporary, of low quality, and provided grams, and school feeding programs in particular coun- little or no training. Most of the better jobs went to tries.25 Others have pointed to broader entitlement or skilled laborers brought in from elsewhere by outside transfer programs, such as the extension of social secu- contractors; skilled labor, for example, accounted for 42 rity to poorer persons and rural areas, as having had sig- percent of labor expenditures in the Nicaraguan SF.16 nificantly greater impacts in reducing poverty. Indeed, Several employment-creation programs in Latin Amer- the magnitude of these latter impacts simply dwarfs the ica, Africa, and Asia, which antedated the SFs, created achievements of SFs in this area.26 Although SFs are not significantly more jobs, employed a larger share of the direct-transfer programs like those just mentioned, these labor force,17 and elicited much greater budgetary re- researchers nevertheless warn that the resources and con- sources from their respective governments. In comparison fidence lavished by donors on SFs as a safety net have to the SFs, most of these programs were supply-driven, diverted attention, funding, and research to measures run by traditional public agencies, and funded by their with only a fraction of the payoff of other measures af- governments rather than donors (at least initially). fecting poverty and unemployment. Wages paid by SFs, although often set at the legal Despite the continued characterization of SFs as safety minimum, have nevertheless been typically lower than nets in donor documents, their evaluations report that subsistence, and sometimes significantly so.18 The wage the available data and its quality do not permit judg- in the Nicaraguan SF, for example, represented only ments as to whether SFs contributed to reducing 57 percent of a basic family food basket. Granted, most poverty.27 In most cases, ir is not possible to determine WHY ARE SOCIAL FUNDS SO POPULAR? 117 whether poverty has been reduced or income increased ties even knew what the operating and maintenance in the regions served by SFs, or, even when such changes costs and responsibilities would be, according to the are detected, it is not possible to determine whether they evaluators, before they chose their project. And only a are attributable to the program.28 "[WI e have no way," a small percentage of the SFs turned out to have actually WB study concludes, "of comparing how well [SFs] tar- required community contributions, even though the get poverty compared with other programs."29 SF projects presented for approval to the WB Board of With respect to SFs' reputation for combating unem- Directors (the "appraisal reports") always included an ployment and poverty, then, they have "created rela- estimate for upfront contributions from communities. tively few jobs and generated little additional income for Little ex-post information on such contributions was the poor,"30 even though many of them included in- available.40 come and employment generation among their stated Social Funds financed many activities-such as objectives. They were not "effective safety nets in any schools, clinics, and wells-that needed sustained sup- significant scale," and many countries therefore did not port from line ministries or other agencies of govern- have "an effective mechanism to protect the poor from ment, once completed. But either no formal arrange- output, employment, and price risks."31 "The message ments were made or those that were made were not is clear," the IDB evaluators conclude, "that if these respected.41 In many cases, no operating funds came funds had an impact on poverty, it cannot have been through for staff and maintenance, particularly for from employment creation and income generation for schools and health.42 According to the logic of decen- the poor."32 Despite these findings, all Latin American tralization, however, this should not have been a serious countries with SFs gave them a "high profile and a cen- problem, at least for some types of projects: The demand- tral role in the campaign to reduce poverty."33 driven features of the SF should have led inexorably to The Evidence: Service Delivery ownership of the new projects by communities, who would have willingly taken responsibility for operations The SF track record as a model of service delivery is, at and maintenance themselves or have successfully pres- best, mixed. A World Bank tracking of the performance sured local governments to do so. But this kind of own- of SF projects in relation to more traditional supply- ership was not forthcoming.43 If, as the evaluators re- driven programs found no clear superiority for SFs, port, neither the donors nor the recipients designed though it nevertheless concluded on a positive note many of these programs with sustainability in mind- about the SF as a model.34 The IDB evaluators report especially the earlier ones-then it is not fair to judge that evidence was not sufficient to form a judgment as them by these criteria. But the donors themselves have to whether SFs have actually made a difference in the made strong claims for these programs as successful, ex- availability of basic economic and social services in the post, on the grounds of community involvement. In- various communities where they operate.35 In addition, deed, they have hailed the SFs as models of "sustainable" they found that the most successful and innovative of service delivery, as attested to by the aforementioned the SFs were those conceived without donor input and quotations. financing (Chile, Costa Rica, and Guatemala), and were With respect to sustainability and ownership at the different from the typical SF in other important ways.36 national level, finally, both donors lament the fact that (More on the Chilean case appears later in this chapter.) most SF programs, 10 years after they were started, con- Both major donors themselves gave distinctly low tinue to be dependent for most of their financing on out- marks to the SFs for "sustainability" and for "ownership" side donors.44 After noting that most Latin American of the projects by the communities in which they were governments with SFs have financed less than 20 percent placed.37 The WB evaluators could find no data on the of their SF operations, the IDB evaluators warn that extent to which SF projects were being operated and "[d]onors cannot claim that the funds are successful and maintained.38 At the same time, frequent reports ap- sustainable" until countries make a greater contribution. peared about health clinics without refrigerators for vac- " [D] onors cannot be expected to provide 80-90 percent cines, school buildings without textbooks, and wells that of the cost of fund operations indefinitely."45 were not maintained. What's more, the evaluators noted The disappointing record of SFs in targeting poor that ownership of these projects by communities often communities may result from the demand-driven dy- required distinctly different technical designs, at least namics unleashed by the SF model itself. Poor commu- for economic infrastructure. But a large number of the nities are handicapped in responding to SF-like initia- SFs were found to have been designed without issues of tives in that they require prior organizing, preparation sustainability in mind.39 It was "not clear" if communi- of project proposals, and choosing and monitoring of ii8 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION outside contractors.46 These communities are often less more responsive to consumer needs and preferences, and organized, less linked to official networks, and are spo- more tailored to local conditions. ken for more by single clientelistic figures than by inclu- The evidence from donor evaluations and other stud- sive community groups. The communities are often dif- ies, including my own, does not support these assertions. ficult to access, and they tend therefore to be less visited First, it revealed that private providers can be as stan- by private providers and NGOs. Whether for reasons of dardizing and insensitive to user needs or local condi- necessity or of maximizing profits and efficiency, these tions as considered typical of the public sector.49 Second, providers concern themselves with keeping down the NGOs were barely present on the scene, accounting for costs of transport, salaries, per diems for travel, and time no more than 15 percent of expenditures by most Latin spent preparing proposals. In some cases, and with the American SFs.50 When they were involved, their projects best of intentions, the SF agency painstakingly mapped were among the least sustainable,51 and often suffered poverty and deficiencies of social services in the region from incompetence and politicization.52 This raises served by the SF. But this still did not counteract the questions about leaning so heavily on NGOs for decen- comparative advantage of communities that were better tralization's benefits to materialize. off-within the "poor-designated" municipalities or sub- Another troubling assertion is that many SFs are regions-in competing for funds.47 All this explains why demand-driven. The evidence suggests that many of these it is often not the communities that choose project particular SFs can be more accurately described as supply- choices and designs, but firms, politicians, or SF staffs. driven in that they substitute a new cast of "supply- Some of the SFs' very strengths seem to be the sources driving" characters for the traditional bureaucrats of far- of their weaknesses, which does not augur well for cor- away agencies. These new characters include building recting the latter. For example, the flexibility and speed contractors, equipment suppliers, and project design of SF disbursement frequently causes communities to firms, together with the more familiar political person- have less choice, not more-undermining their ability ages (mayors, legislators, ward bosses) and government to gather information about options for alternative proj- technical agents themselves-this time from the SF.53 To ects, and to deliberate and choose among them. What's the extent that these actors determine the choice of proj- more, the prized autonomy and special privileges of SFs ects, rather than community members, it is not sur- also reduce the opportunities and likelihood of their prising that sustainability and ownership often fail to working complementarily with existing institutions of materialize. But continuing to describe these programs government, whose lack of involvement partly explains indiscriminately as reflecting community preferences the frequent failure of maintenance and operation. Also makes it difficult to understand exactly what goes problematic, the same donor accounts that laud the wrong-and, more importantly if things work well, what more flexible and speedy disbursement of SFs point dis- goes right. If there is a distinction to be made between approvingly to these traits as causing SF-served commu- SFs and other government programs, then, it may lie nities to have less choice-with less information about less in bringing beneficiary choice into the picture than options, and less time to deliberate.48 Finally, SF auton- in substituting another array of supply-driven forces. omy does not necessarily translate, as claimed, into more Supply-drivenness, however, is exactly the critique that is "apolitical," technocratic, or "private-sector-like" man- made of traditional government provision. This shifts the agement styles. The less noted flip side of this autonomy focus of attention from the matter of demand-driven vs. is the SFs' vulnerability to political manipulation. supply-driven, and the SF model vs. traditional govern- ment, to which kinds of supply-driven programs are bet- Turning Assumptions into Questions trta tes ter than others. The evidence on SFs and other decentralizing experi- Information Asymmetries: The New Spoilers ences is somewhat muddied by the fact that many ac- counts treat as assumptions matters that should be Some of the afflictions of SFs and other programs serv- treated as questions. For example, decentralization and ing poor and rural populations originate in "asymme- outsourcing to local providers are assumed to create a tries" of information and power. These figure impor- special dynamic leading to better service-a dynamic tantly in the recent literature of the new institutional that is said to be missing in traditional public programs. economics. Information asymmetries can cause trouble According to this logic, private firms and NGOs help in various kinds of contractual relationships-between reduce government's monopoly power as provider, an buyers and sellers, service providers and users, and cen- important source of its inefficiency. The ensuing com- tral governments and local governments. If one side petition among potential providers yields results that are knows much more than the other and keeps that infor- WHY ARE SOCIAL FUNDS SO POPULAR? II9 mation to itself, the underinformed and less powerful suggests that even when the decentralization dynamic party to the transaction does not fare well. In fact, the of the SFs is working as it should, it cannot produce the new institutional economists have warned that when desired results. these kinds of asymmetries prevail, the usual bets on the Unfortunately, SFs typically work in environments gains to be made from decentralization and its associated where asymmetries of information and power are par- measures are off. The economist Joseph Stiglitz played a ticularly common-rural areas and poor communities. seminal role in developing this literature of information Population densities are lower, illiteracy is higher, and asymmetries and transactions costs in the 1980s.54 Later travel and other means of communication are more dif- in the 1 990s, when Stiglitz was Senior Vice President and ficult. Government is less present, whether as provider Chief Economist for Development of the World Bank, or regulator, partly because this kind of presence is more he warned of the stranglehold that these asymmetries difficult and costly under such conditions. For this and could exert on the benefits normally expected from de- other reasons, then, the SF model might actually work centralization, as well as from privatization.55 least well in such environments, not better. An indica- Those who worry about asymmetries of information tor of this problem can be found in the remedies sug- and power typically assign to government the role of cor- gested by the donors to fix SF shortcomings. recting or counterbalancing them-as regulator, media- A representative sampling of such remedies includes: tor, and broker for the weaker party to the contract, and more monitoring and supervision, more transparent and as provider of information. Indeed, this is exactly what objective selection criteria for projects, more training, SF units are meant to do. Social Fund project designs, more public information campaigns about project choices or at least the evaluations of existing SF programs, typi- available to communities, more tolerance by project cally call for strong public information campaigns about managers for "participation," more poor-targeted selec- the new choices available to communities, and the pro- tion criteria, more "demand orientation" and commu- cedures for taking advantage of them. But the evidence nity participation in helping communities to choose shows information and community choice to have been their projects, more attention to organizing users around surprisingly low in many of the assisted communities, operations and maintenance or to committing line agen- even in the Brazilian programs of my field research cies to that responsibility, and finally, more coordination where project design included information campaigns with line agencies and their sectoral programs.57 Many and the donor paid serious monitoring attention to of these remedies would require a significant increase of them. Combined with the typical political, bureaucratic, SF agency presence in the countryside in terms of time, and market forces present in many SF environments, personnel, resources, and effort. This kind of change, ironically, the model may actually reinforce the very moreover, would compromise one of the SFs' most asymmetry of information that it was meant to reduce. acclaimed strengths-their "leanness" and low adminis- The findings from my Brazil research projects shed trative costs. More constructively, if becoming more light on this strange outcome. Three separate sets of ac- supply-driven is the solution rather than the problem, tors turned out to have a distinct interest in limiting in- these conclusions suggest that donors might look with formation rather than broadcasting it: contractors and greater enthusiasm and diligence to reform prospects project-design firms, elected leaders and other politi- within existing public institutions. cians, and SF management and staff.56 It is important Getting Out of the Fix to note that their information-limiting behavior was not simply motivated by rent-seeking and other forms of The difference between my judgments about SFs and mean-spiritedness. Just as often, these actors were doing those of the donors turns less on the empirical find- what they thought was right-maximizing profits or ef- ings-given that much of the aforementioned case evi- ficiency, enacting a particular vision of the public good, dence comes from the donor evaluations themselves- insisting on standards representing the profession's con- than on our differing interpretations of the problems sensus about best practice, or making the best of a work revealed by this evidence. To their credit, the donors environment in which demand for projects exceeded take the flaws they report seriously. But they also assume supply. This means that even when actors thought they that these flaws can be readily fixed without compromis- were serving the public good, or when firms were behav- ing the strengths of the model. I see these findings, in ing at their best, the results were not that different from contrast, as meriting greater pause. If informed commu- those said to be produced by rent-seeking behavior! This nity choice, sustainability, and reduction of inequities explanation for SF shortcomings is in some ways more of service provision are incompatible with the SFs' most troubling than that of rent-seeking and corruption: It marked achievements, then they are not much better in IZO LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION the long run than existing approaches to reforming tra- civil-service regulations. And it is financed mainly by ditional agencies, or in dealing with similar problems in the Chilean government, in comparison to the continu- the recent decentralization experience. ingly heavy donor funding of most other SFs.60 These Finding incompatibility among basic goals or instru- traits place the Chilean SF closer to existing government ments within a particular program is not unusual. than the average. Outlier cases like this raise questions Scholars of the behavior of organizations have long about what actually makes SFs work when they do. shown how the goals of any particular organization More importantly, these cases should be mined for an often conflict, one with the other. Indeed, a mix of con- understanding of how to reduce some of the important flicting goals often serves various organizational pur- limitations of SFs, while at the same time showing how poses and constituencies well. Treatises on organiza- the experience can be made more germane to reform ef- tional leadership show how the best agency managers forts in the traditional public sector. are those who manage these contradictions artfully, and The findings reported here, in sum, raise important in a way that presents a unified face to the outside questions about SFs and other reforms, whether decen- world. It is not the incompatibilities themselves that are tralizing, deconcentrating, or neither. If SFs and tradi- cause for concern, then, but the lack of recognition of tional public institutions all tend to be supply-driven, them and, hence, their implications. that is, then we should be asking which of these inter- Trimming the perception of SFs down to size would ventions leads to better results, in which particular cir- make it possible to look at the SF experience and that cumstances and places, and for which particular tasks. of traditional government agencies with a more open Correspondingly, what are the circumstances under and curious mind. At their best, for example, SFs may which private firms providing public goods and services represent a "deconcentrated" version of supply-driven are, a la Stiglitz, actually likely to be more responsive service-rather than demand-driven or decentralized than public institutions to user needs? These questions service-that leads to the construction of small infra- will not yield answers that are necessarily pro-SF or anti- structure projects more quickly and at lower cost. It SF, or pro- or antitraditional government. Importantly, must be kept in mind, however, that these programs however, they do not start out by assuming that pro- seem to do no better than traditional agencies in gener- grams with demand-driven designs actually are demand- ating maintenance and operational support. Social driven, or that private and decentralized provision al- Funds may also show the way for traditional public ways elicits more user choice and customer-tailored agencies to simplify procurement regulations. At the results. Turning the assumptions into questions instead same time, no such demonstration effect has yet been could generate considerable empirical material of value reported in SF evaluation studies. about how communities decide, how markets work The reported achievements of SFs in rapid disburse- under outsourcing and partial privatization, and how ment rates and lower unit costs, in turn, may represent politics influences outcomes-whether for the better or such significant progress in service delivery that the ap- for the worse. proach is well worth pursuing.58 If these gains are so pal- Gaining Favor pable, it is important to pay attention to how they can be applied to reducing costs and delays in existing agen- If SFs do not measure up to the broadly held under- cies and programs. At the same time, however, this standing of how they work and what they accomplish, mode of operation can also jeopardize participation, lo- then why are they so popular? Why do the very docu- cally tailored solutions, and sustainability. Similarly, SFs ments relating the shortcomings previously cited con- may work at their best only for certain kinds of commu- clude on such a positive note? If SFs are afflicted with nities-better off, better organized, or less remote. But some of the same grave problems that afflict existing what works best for the rest, who are less well off and so government agencies, then why has this model elicited much the object of policy concern? so much enthusiasm in comparison to efforts to reform Finally, it should be noted that one of the best SFs is traditional agencies working on similar problems? If actually most unlike the average SF-the Chilean SF. In many staff members of the donor agencies themselves this case, the SF is not an autonomous unit and works share these skeptical views about SFs, then why has the instead through a line agency of the central government. more enthusiastic interpretation won out, and so (A study comparing the Chilean with the Venezuelan SF strongly influenced policy? cited the unusual integration of the Chilean SF with the In part, the popularity of SFs is itself supply-driven, line ministries as an explanation for why it was more but in this case the "supply-drivers" are donor agencies successful.)59 It does not circumvent procurement and rather than traditional government agencies in borrower WHY ARE SOCIAL FUNDS SO POPULAR? I2I countries. The reason for SF popularity, that is, relates nurturing has led to the formation of an articulate and not only to the acclaimed features of the model, but also visible support group for SFs within and across borrower to the workings of the donors as large bureaucracies. As countries themselves, which must surely boost morale such, they are subject to the same dynamics that drive and enhance learning. Support for such sustained cross- all large organizations, including their complex relation- country networking is usually not available to public ship with and sensitivity to their outside environments. servants trying to carry out reform within line agencies, For perfectly understandable reasons, many donor pro- although they face at least as daunting and isolating an fessionals find it more satisfying to work with SFs and environment as those working at SFs. on SF projects than with traditional agencies. The man- Social Funds are also popular because of their crucial agement and staff of SF agencies in borrower countries role in helping donors sell austerity reforms to borrower have far fewer bureaucratic masters to serve than min- countries. Because donors present SFs to governments istries and other existing agencies, and are therefore as safety-net measures that help reduce poverty and more easily accessible to donor project officers and more unemployment (despite the evidence that SFs do not able, if willing, to carry out their suggestions for im- really fulfill these goals), they help counteract the wide- proved service. All this provides a greater sense of con- spread public criticism of these governments that aus- trol and accomplishment to donor project officers, mak- terity measures fall disproportionately on the poor. The ing it easier to design such projects and monitor them, belief that SFs will help ease the adoption of austerity and to see things happen before one's eyes. To dedicated packages, in turn, attracts support from macroecono- professionals working within the bureaucratic con- mists in international financial institutions whose task straints of a large donor organization, this makes for sat- is to convince countries to adopt the macroreform pack- isfying work. ages. This support represents an influential and power- Adding to this liberating effect, the regulations of ful voice for SFs-the voice of professionals who would both the World Bank and the IDB endow SFs with otherwise have no particular interest in such program- something akin to fast-track status compared to other matic interventions. categories of projects. The World Bank suspends condi- Social Funds also draw powerful support because tionality requirements on SF projects, which normally they help leaders out of the political dilemma created by slow down and complicate the trajectory of project ap- such austerity programs and other unpopular reform proval in the Bank. IDB rules require that funding for measures. The political assist from SFs does not neces- SFs and other projects falling in this more subsidized sarily work directly through their actual impact on the category must be approved before some other categories poor, but through their use by elected leaders to selec- of projects offered on less subsidized terms can go for- tively court groups of voters, whether among the poor- ward. This requirement turns a more numerous and in- est or not. In that the macroreform programs inevitably fluential group of IDB professionals and directors into cause governments to lose votes from important con- SF advocates-in itself contributing to faster project ap- stituencies, this leads to the aggressive courting by lead- provals. To the same end, the SF agency's association ers of new constituents to make up for the losses, often with an international donor is an important source of through the dispensing of patronage. Social Funds, like power and prestige in its own bureaucratic world. This all "distributive" programs that provide numerous indi- helps counterbalance the jealousies and resistance of vidual grants for small projects that are spatially dis- other bureaucratic actors on whom SF projects are de- persed, are an excellent vehicle for such patronage-as pendent for maintenance and operation, and who resent demonstrated in richly chronicled accounts of SFs and the special status of SFs and their freedom from civil- electoral politics in Peru and Mexico.63 That elected service and procurement regulations.61 leaders can distribute these funds in a discretionary Borrower-government elected leaders also support manner to some districts and not others, and that they SFs, though this is partly in reaction to donor sugges- are exempt from existing formulas for revenue-sharing tion. Because donors often suggest SFs as part of a larger with local governments, also contributes to their suit- lending package, borrower governments tend to favor ability for patronage purposes. their creation as a surefire way of obtaining donor fi- Social Funds' prized autonomy and freedom from nancing. In addition, donors have invested considerable bureaucratic encumbrances, in sum, also makes them effort in creating and supporting a network of SF pro- vulnerable to political meddling. While donors are fessionals from various countries, funding them to travel praising SF autonomy, presidents and other elected of- to international meetings where they share their experi- ficials are viewing it as almost the opposite kind of bless- ences and learn about best practices in other SFs.62 This ing. To them, SFs offer easier access than regulation- I22 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION bound ministries for meeting their electoral needs. This Barnett, Richard A. 1994. "Publicity Strategy Evaluation, North- politics-as-usual side of SFs tends to be obscured by the east Rural Development Project." World Bank, Recife Office. accounts of SFs as a "modern" model of management, August 1 1. and by their inclusion in packages of "modernizing" re- Bianchi, Tito. 1997. "Supply-led Technology Adoption in and by thirinluioipckgsfmoerizngMaranhao, Brazil: A Case of Development Trade in the forms to be carried out by "serious" leaders. Small-scale Furniture Sector." MIT/State of Maranhao Proj- The final reason for SF popularity among donors re- ect. Master's Thesis, Department of Urban Studies and Plan- lates to their effectiveness as a powerful "development ning, Massachusetts Institute of Technology. May. Cam- narrative," to use the words of Emery Roe and others, bridge, Massachusetts. Roewrites that in environments with great ambiguity Bianchi, Tito. 1998. "Built to Last: A Couple of Well-Known Roe writes Secrets About Cooperatives and What Keeps Them Alive." as to cause and effect, such narratives offer convincing Report to the MIT/BNB (Banco do Nordeste) Project. July and simple explanations for the causes of certain prob- 14. Cambridge, Massachusetts. lems and provide appealingly straightforward blueprints Bonfim, Washington L. de Souza. 1999. "Qual Mudan,a? Os for action (1991, 1999). Because of their power as nar- Empresarios e a Americanizaqao do Ceara." Unpublished ratives, he argues, these accounts are rather invulnerable Ph.D Dissertation in Political Science, Instituto Universitario ratives, he.argues, these accounts are rather invulnerable de Pesquisas do Rio de Janeiro (IUPERJ). Rio de Janeiro. to empirical evidence that challenges their accuracy- Boschi, Renato. 1998. "Democratic Governance and Partic- such as the evidence of SF shortcomings, and of in- ipation: Tales of Two Cities." Universidade Federal de significant impacts on poverty and unemployment. Minas Gerais/Instituto Universitario de Pesquisas do Rio de Roe develops three examples from Africa of such de- Janeiro. velopment narratives and contrasts them with extensive Cornelius, Wayne A., Ann L. Craig, and Jonathan Fox. 1994. empricl fndigs o te cntrry64 In each of these Transforming State-Society Relations in Mexico: The National empirical fndilngs to the contrary. nec o hs Solidarity Strategy. San Diego: Center for U.S.-Mexican cases, he shows how remarkably resistant these narra- Studies, University of California. tives have been to case evidence that contradicted them. Cornia, Andrea Giovanni. 1999. "Social Funds in Stabilization He draws on his empirical material to make program- and Adjustment Programmes." United Nations University/ matic suggestions that differ from the "blueprint for ac- World Institute for Development Economics Research, Hel- tion" attached to each of three narratives At the sinki. Forthcoming. In Andrea Giovanni Cornia, ed., In- tion"attachetowv heachidfthesthree ac rrade ices.a consu equality and Poverty Under the Washington Consensus. New same time, however, he chides academics and consul- York: Oxford University Press. tants like himself for "naively" thinking that carefully Cornia, Andrea Giovanni, and Sanjay Reddy. 2000. "The Im- gathered empirical evidence could have the same kind pact of Adjustment-related Social Funds on Distribution and of power the narratives have in influencing the way in- Poverty." Manuscript. stitutions think about and act upon such problems. The Costa, E. 1978. "An Assessment of the Flows and Benefits Gen- talk about SFs would seem to meet Roe's qualifications erated by Public Investment in the Employment Guarantee Scheme of Maharashtra." Working Paper No. 12. Geneva: of a successful development narrative, with a corre- International Labour Organisation/World Employment sponding blueprint for action. This also helps explain Programme. why donors remain so enthusiastic about SFs, and so ac- Damiani, Octavio. 1996. "Report for the Government of cepting of the causal assumptions underlying the model, Brazil-World Bank Commission: Rural Poverty." Cam- despite the questionable evidence that they themselves bridge, MA: Department of Urban Studies and Planning, have unearthed. Massachusetts Institute of Technology. February. Deolalikar, Anil B., and Raghav Gaiha. 1996. "What Determines Female Participation in Rural Public Works? The Case of Bibliography India's Employment Guarantee Scheme." University of Wash- Abrucio, Fernando Luiz. 1998. Os barges dafederacfo: Os gover- ington and the University of Delhi. April. nadores e a redemocratizapio brasileira. Departamento de D'Silva, E. H. 1983. "Effectiveness of Rural Public Works in Ciencia Politica da USP [Universidade de Sao Paulo]. Sao Labour-surplus Economies: Case of the Maharashtra Employ- Paulo: Editora Hucitec. ment Guarantee Scheme." Ithaca, NY: Cornell International Ames, Barry. 1995. "Electoral Strategy under Open-List Propor- Agricultural Monograph No. 97, Cornell University. tional Representation." American Journal of Political Science Echeverri-Gent, John. 1988. "Guaranteed Employment in an 39(2, May): 406-33. Indian State: The Maharashtra Experience." Asian Survey Ames, Barry. 2000. Institutions and Politics in BraziL Ann Arbor, 28(12, December): 1294-1310. MI: University of Michigan Press. (Relevant material drawn Echeverri-Gent, John. 1993. The State and the Poor: Public Pol- from Part 1, Chapter 4 of "Institutions and Politics in Brazil," icy and Political Development in India and the United States. draft manuscript, 1996.) Berkeley, CA: University of California Press. Angell, Alan, and Carol Graham. 1995. "Can Social Sector Re- Eicher, Carl K. 1999. "Institutions and the African Farmer." form Make Adjustment Sustainable and Equitable? Lessons CIMMYT [International Maize and Wheat Improvement from Chile and Venezuela." Journal of Latin American Stud- Center] Third Distinguished Economist Lecture. Mexico, ies 27(1, February): 189-219. DF: CIMMYT WHY ARE SOCIAL FUNDS SO POPULAR? 123 Fox, Jonathan, and Josefina Aranda. 1996. Decentralization and Hutchful, Eboe. 1994. " 'Smoke and Mirrors': The World Bank's Rural Development in Mexico: Community Participation in Social Dimensions of Adjustment (SDA) Programme." Re- Oaxaca's Municipal Funds Program. Monograph Series, 42. view ofAfrican Political Economy 62: 569-84. San Diego, CA: Center for U.S.-Mexican Studies, Univer- Inter-American Development Bank (IDB). 1997a. Social Invest- sity of California. ment Funds in Latin America: Past Performance and Future Gaiha, Raghav. 1998. "Do Anti-Poverty Programmes Reach the Role. By Margaret Goodman et al. Evaluation Office, Social Rural Poor in India?" New Delhi: Faculty of Management Programs and Sustainable Development Department. March. Studies, University of Delhi. May. Washington, D.C. Gay, Robert. 1990. "Popular Incorporation and Prospects for Inter-American Development Bank (IDB). 1997b. Chile: Fondos Democracy: Some Implications of the Brazilian Case." Theory de Solidaridad de Inversion Social (FOSIS). Informe de la Con- and Society 19: 447-63. sultora Srta. Dagmar Raczynski. Evaluation Office, EVO. Gay, Robert. Forthcoming. "The Broker and the Thief: A June 1996. In Social Investment Funds in Latin America: Past Parable (Reflections on Popular Politics in Brazil)." Luso- Performance and Future Role, Chapter 2. A Joint Project Be- Brazilian Review. tween the Evaluation Office and the Social Programs and Sus- Gay, Robert. Forthcoming. "Rethinking Clientelism: Demands, tainable Development Department. June. Washington, D.C. Discourses and Practices in Contemporary Brazil." European Inter-American Development Bank (IDB). 1998. "The Use of Review of Latin American and Carribean Studies. Social Investment Funds as an Instrument For Combating Gershberg, Alec Ian. 1994. "Distributing Resources in the Edu- Poverty." A Bank Strategy Paper. July. Washington, D.C. cation Sector: Solidarity's Escuela Digna Program." In Wayne Joshi, Anu. 1999. "Mobilizing the Poor? Activism and the Em- A. Cornelius et al., eds., Transforming State-Society Relations ployment Guarantee Scheme, Maharashtra." Draft. Institute in Mexico: The National Solidarity Strategy. San Diego, CA: of Development Studies, Sussex. Center for U.S.-Mexican Studies, University of California, Lowi, Theodore J. 1964. "American Business, Public Policy, pp. 233-54. Case-Studies, and Political Theory." World Politics 14(July): Gibson, Edward L. 1997. "The Populist Road to Market Re- 677-715. form: Policy and Electoral Coalitions in Mexico and Ar- Lustig, Nora, ed. 1995. "Coping with Austerity: Poverty and In- gentina." World Politics 49(April): 339-70. equality in Latin America." A study jointly sponsored by the Graham, Carol. 1994. "Mexico's Solidarity Program in Compara- Brookings Institute and the Inter-American Dialogue. Wash- tive Context: Demand-based Poverty Alleviation Programs in ington, D.C.: The Brookings Institution. Latin America, Africa, and Eastern Europe." In Wayne A. Cor- Lustig, Nora. 1997. "The Safety Nets which Are Not Safety nelius et al., eds., Transforming State-Society Relations in Mex- Nets: Social Investment Funds in Latin America." Draft. Oc- ico: The National Solidarity Strategy. San Diego, CA: Center for tober 31. Washington, D.C. U.S.-Mexican Studies, University of California, pp.309-28. Molinar Horcasitas, Juan, and Jeffrey A. Weldon. 1994. "Elec- Graham, Carol. 1998. "The Capitalization and Popular Partici- toral Determinants and Consequences of National Solidar- pation Programs in Bolivia." In Carol Graham, ed., Private ity." In Wayne A. Cornelius et al., eds., Transforming State- Markets for Public Goods: Raising the Stakes in Economic Society Relations in Mexico: The National Solidarity Strategy. Reform. Washington, D.C.: Brookings Institution Press, San Diego, CA: Center for U.S.-Mexican Studies, Univer- pp. 139-87. sity of California, pp. 123-42. Graham, Carol, and Cheikh Kane. 1998. "Opportunistic Gov- Morrison, Judith A. 1997. "What Works in Rural Afro-Brazilian ernment or Sustaining Reform? Electoral Trends and Public- Communities of Maranhao? Impressions of Successful Agency Expenditure Patterns in Peru, 1990-1995." Latin American Interventions." Master's Thesis, Department of Urban Studies Research Review 33(1): 67-104. and Planning, Massachusetts Institute of Technology. MIT/ Heller, Patrick. 1999. The Labor of Development: Workers and State of Maranhao Project. May 12. Cambridge, Massachusetts. the Transformation of Capitalism in Kerala, India. Ithaca, NY: Natalacchio, Marcela. 1998. "Organization and Empowerment Cornell University Press. of Rural Women in the Babassu Areas of Maranhao, Brazil." Herring, Ronald J. 1983. Lind to the Tiller: The PoliticalEcon- MIT/State of Maranhao Project. Master's Thesis, Graduate omy ofAgrarian Reform in South Asia. New Haven, CT: Yale School of Arts and Sciences, Boston University. Boston. University Press. Nelson, Joan. 1997. "Reforming Social Sector Governance: A Herring, Ronald J. 1998. "Agrarian Reform, Path Dependency Political Perspective." Paper prepared for a conference on and Persistent Poverty: Lessons from the United States and Governance, Poverty Eradication, and Social Policy, Harvard India." For the Conference "What Can be Done about University, 12-14 November. November 7. Overseas Devel- Poverty?," Institute of Development Studies, University of opment Council, Washington D.C. Sussex, June 29-July 1, 1998. Nordhaus, William D. 1975. "The Political Business Cycle." Re- Herring, Ronald J., and Rex M. Edwards. 1983. "Guaranteeing view of Economic Studies 42(2): 1969-90. Employment to the Rural Poor: Social Functions and Class Pinhanez, Monica F. 1997. "Shattered Power, Reconstructed Interests in the Employment Guarantee Scheme in Western Coalitions: An Analysis of Rural Labor Unions in Maranhao, India." World Development 11(7): 575-92. Brazil." MIT/State of Maranhao Project. Master's Thesis, De- Hesse, Fernando J. 1996. Against All Odds: Successful Cases of partment of Urban Studies and Planning, Massachusetts In- Collective Use of Tractors in Rural Ceara, Brazil." MIT/State stitute of Technology. May. Cambridge, Massachusetts. of Ceara Project II. Master's Thesis, Department of Urban Quir6s, Rosa. 1996. "Rural Water Supplies that Work, Endure Studies and Planning, Massachusetts Institute of Technology. and Reach the Poor: Lessons from Ceara, Brazil." MIT/State June. Cambridge, Massachusetts. of Ceara Project II. Master's Thesis, Department of Urban 124 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Studies and Planning, Massachusetts Institute of Technol- Tagle, Laura. 1996. "Inadvertently Reaching the Poor: The Dif- ogy. May. Cambridge, Massachusetts. fusion of Small Scale Irrigation in Northeast Brazil." MIT/ Roberts, Kenneth M. 1996. "Neoliberalism and the Transfor- State of Ceara Project II. Master's Thesis, Department of mation of Populism in Latin America: The Peruvian Case." Urban Studies and Planning, Massachusetts Institute of World Politics 48(1): 82-116. Technology. June. Cambridge, Massachusetts. Roberts, Kenneth M., and Moises Arce. 1998. "Neoliberalism Tendler, Judith. 1982. Rural Projects Through Urban Eyes: An and Lower-Class Voting Behavior in Peru." Comparative Po- Interpretation of the World Bank's New-style Rural Develop- litical Studies 31 (2): 393-407 ment Projects. Staff Working Paper No. 532. Washington, Roe, Emery M. 1991. "Development Narratives, Or Making the D.C.: World Bank. Best of Blueprint Development." WorldDevelopment 19(4): Tendler, Judith. 1993a. New Lessons from Old Projects: The 287-300. Workings of Rural Development in Northeast Brazil A World Roe, Emery M. 1999. Except Africa: Remaking Development, Re- Bank Operations Evaluation Study. Washington, D.C.: World thinking Power. New Brunswick, NJ: Transaction Publishers. Bank. Rogoff, Kenneth. 1990. "Equilibrium Political Budget Cycles." Tendler, Judith. 1993b. "Tales of Dissemination in Small-farm The American Economic Review 80(1, March): 21-36. Agriculture: Lessons for Institution Builders." World Devel- Rogoff, Kenneth. 1994. "Equilibrium Political Budget Cycles." opment2l(10, October): 1567-82. In Torsten Persson and Guido Tabellini, eds., Monetary and Tendler, Judith. 1997. Good Government in the Tropics. Balti- Fiscal Policy, Vol. 2, pp. 47-70. Cambridge, MA: MIT Press. more, MD: The Johns Hopkins University Press. [Paperback Schady, Norbert R. 1999. "Seeking Votes: The Political Econ- version, 1998; Portuguese version published by Editora omy of Expenditures by the Peruvian Social Fund (FON- Revan and ENAP (Escola Nacional de Administracao Pi- CODES), 1991-1995." Policy Research Working Paper blica) (Rio deJaneiro) 1998.] 2166. The World Bank. August. (Page references in the chap- Tendler, Judith. 1999. The Rise of Social Funds: What Are They ter are to a 1998 draft.) a Model Of? Monograph for the United Nations Develop- Scott, James. 1998. Seeing Like a State: How Certain Schemes to ment Program (UNDP). January. Improve the Human Condition Have Failed. New Haven, CT: Tendler, Judith, and Monica Amorim. 1994. "Small Firms and Yale University Press. Their Helpers: Lessons on Demand." World Development Seabright, Paul. 1997. "Is Co-operation Habit-Forming?" In 24(3, March): 407-26. Partha Dasgupta and Karl-Goran Maler, eds., The Environ- Tendler, Judith, and Sara Freedheim. 1996. "Trust in a Rent-seek- ment and Emerging Development Issues, Vol. 2, pp. 283-307. ing World: Health and Government Transformed in North- Oxford: Clarendon Press. east Brazil." WorldDevelopment22(12, December): 1771-91. Serrano Berthet, Rodrigo. 1996. Who Knows What's Bestfor the Thomas, John W. 1986. 'Food for Work: An Analysis of Current Poor: Demand-Driven Policies and Rural Poverty in Northeast Experience and Recommendations for Future Performance." Brazil. MIT/State of Ceara Project II. Master's Thesis, De- Development Discussion Paper No. 213. Cambridge, MA: partment of Urban Studies and Planning, Massachusetts In- Harvard Institute for International Development. Harvard stitute of Technology. May 22. Cambridge, Massachusetts. University. Serrano Berthet, Rodrigo. 1997. "Fruitful Policies: The Adoption United Nations Educational, Scientific, and Cultural Organiza- of Cash Crops by Subsistence Farmers in Maranhao, Brazil." tion (UNICEF). 1998. Social Funds in Developing Countries: First-year Doctoral Paper, Department of Urban Studies and Recent Experiences and Lessons. By Sanjay Reddy. UNICEF Planning, Massachusetts Institute of Technology. MIT/State Staff Working Papers. Evaluation, Policy and Planning Series of Maranhao Project. May. Cambridge, Massachusetts. Number EPP-EVL-98-002. June. New York. Skocpol, Theda, and Kenneth Finegold. 1982. "State Capacity Vicira da Cunha, Paulo, and Maria Valeria Junho Pefia. 1997. The and Economic Intervention in the Early New Deal." Politi- Limits and Merits of Participation. Policy Research Working cal Science Quarterly 97(2): 255-78. Paper No. 1838. The World Bank, Office of the Senior Vice Souza, Celina Maria de. 1997. Constitutional Engineering in President and Chief Economist. October. Washington, D.C. Brazil: The Politics of Federalism and Decentralization. New Vivian, Jessica. 1994. "NGOs and Sustainable Development in York: St. Martin's Press. Zimbabwe: No Magic Bullets." Development and Change Steffes, Ann. 1997. "An Organizational Analysis of Rural Drink- 25(1): 167-93. ing Water Provision in Ecuador and Maranhio, Brazil." De- Vivian, Jessica, and Gladys Maseko. 1994. "NGOs, Participa- partment of Urban Studies and Planning, Massachusetts tion and Rural Development: Testing the Assumptions with Institute of Technology. MIT/State of Maranhao Project. Evidence from Zimbabwe." United Nations Research Insti- February 10. tute for Development. Discussion Paper DP 49. United Na- Stewart, Frances, and Willem van der Geest. 1995. "Adjustment tions Research Institute for Social Development. January. and Social Funds: Political Panacea or Effective Poverty Re- von Braun, Joachim, Tesfaye Teken, and Patrick Webb. 1992. duction?" In Frances Stewart, ed., Adjustment and Poverty, "Labor-Intensive Public Works for Food Security in Africa: pp. 108-37. London: Routledge. Past Experience and Future Potential." International Labour Stiglitz, Joseph E. 1985. "Information and Economic Analysis: Review 131(1): 19-34. A Perspective." EconomicJournal 95: 24-41, Supplement. Webb, Kathryn Vandever, Kye Woo Lee, and Anna Maria San- Stiglitz, Joseph E. 1998. "More Instruments and Broader Goals: t'Anna. 1995. The Participation of Nongovernmental Organi- Moving toward the Post-Washington Consensus." World In- zations in Poverty Alleviation: A Case Study of the Honduras stitute for Development Economics Research (WIDER). Social Investment Fund Project. Discussion Paper No. 295. WIDER Annual Lectures 2. January. Washington, D.C.: World Bank. January. WHY ARE SOCIAL FUNDS SO POPULAR? 125 Weyland, Kurt. 1998. "Swallowing the Bitter Pill: Sources of of Management Governance and Development of the United Na- Popular Support for Neoliberal Reform in Latin America." tions Development Program. For support for my research with Comparative Political Studies 31(5, October). students and for leave time, I thank Bish Sanyal, Chair of the De- White, Robert, and Carl Eicher. 1999. "NGOs and African partment of Urban Studies and Planning of the Massachusetts In- Farmers: A Skeptical Perspective." Staff Paper No. 99-01. De- stitute of Technology. None of these persons or institutions is re- partment of Agricultural Economics. East Lansing, MI: Uni- sponsible for or necessarily agrees with my findings. versity of Michigan. January. 2. These numbers are approximate and a few years out of date Woodall, Brian. 1996. Japan Under Construction: Corruption, because of the difficulty of finding consistently defined cross- Politics, and Public Works. Berkeley, CA: University of Cali- donor data on SFs. In May 1997, the World Bank (WB) re- fornia Press. ported having committed US$1.367 billion to SFs in 34 coun- World Bank (WB). 1995. Decentralized Rural Development and tries (WB 1997a: 5, and Attachment 1, end-fiscal year 1996); the Enhanced Community Participation: A Case Study from North- Inter-American Development Bank (IDB) reported an additional east Brazil. Policy Research Working Paper No. 1498. By US$1.3 billion in 16 countries, 8 of which involved SFs to which Johan van Zyl et al. Agriculture and Natural Resources De- the WB also contributed, making a net total of 41 countries for partment, Sector Policy and Water Resources Division. these two largest donors (IDB 1997a: 10, Table 2.1). The IDB August. Washington, D.C. reported an additional US$558.2 million in its Latin American World Bank (WB). 1997a. "Portfolio Improvement Program Re- SFs as coming from "other donors"-mainly European donors, view of the Social Funds Portfolio." The Working Group for which does not include SFs funded outside Latin America by the Social Funds Portfolio Review, headed by Ishrat Husain these other donors. For the latter reason, and because the num- (PREM). Forthcoming as World Bank Technical Paper. May. bers do not capture the past few years' SF commitments in World Bank (WB). 1997b. "Innovations in the Delivery and Or- Southeast Asia and Eastern Europe in particular, the totals in the ganization of Social Assistance and Poverty-Targeted Pro- text are an underestimate. Also, in several countries, the WB and grams." In K. Subbarao et al., Safety Net Programs and Poverty IDB have more than one project, or have given second loans to Reduction: Lessons from Cross-Country Experience. Washing- existing SFs; in the 34 countries with WB-funded SFs, 60 proj- ton, D.C.: World Bank. ects have been funded-some representing follow-on loans and World Bank (WB). 1998a. Designing Rulesfor Demand-Driven others the creation of a second SF in that country. For the 16 Rural Investment Funds: The Latin American Experience. By countries with IDB-funded SFs, this includes 18 projects. Thomas Wiens and Maurizio Guadagni. World Bank Tech- 3. This chapter is not meant to be a thorough review of the nical Paper No. 407. May. Washington, D.C. SF experience or of the arguments for and against them. For this, World Bank (WB). 1998b. Social Funds and Reaching the Poor: the reader can turn to several comprehensive donor-funded re- Experiences and Future Directions. Anthony Bigio, ed., Wash- views of the evidence and other excellent studies of SFs by out- ington, D.C.: World Bank. side researchers, cited in the endnotes throughout or appearing World Bank (WB) . 1998c. Getting an Earful: A Review of Ben- in the Bibliography at the end of this chapter. eficiary Assessments of Social Funds. Social Protection Dis- 4. Nora Lustig (1997) contests this statement persuasively, cussion Paper No. 9816. By Daniel Owen and Julie Van which has been frequently repeated in donor documents. With Domelen. respect to the Latin American SFs, at least, she shows that donor- Wright, Gavin A. 1974. "The Political Economy of New Deal funded SF projects were actually underway well before the struc- Spending: An Econometric Analysis." The Review ofEconom- tural adjustment programs began to show any hint of adverse ef- ics and Statistics 56(1): 30-38. fects on the poor. Notes 5. For example, WB (1997b: 104) reports savings of 30-40 percent in school construction in Mexico's SF, PRONASOL; 1. I am most grateful for comments on an earlier version of and savings of up to 35 percent in Mexico's Mendoza Provincial this chapter from Meenu Tewari, Rodrigo Serrano, Hubert Program for Basic Social Infrastructure (MENPROSF). Schmitz, Anna Maria Sant'Anna, Sanjay Reddy, Sonia Ospina, (PRONASOL is one of the SFs initiated without donor assis- Joan Nelson, Anu Joshi, Maurizio Guadagni, Margaret Good- tance-though it has subsequently received donor funding-and man, Raghav Gaiha, Osvaldo Feinstein, Octavio Damiani, and to which the Mexican government has committed more funds John Briscoe. For a very detailed reading of the chapter and ex- than all of the Latin American SFs combined.) tensive comments, I am indebted to Mounir Tabet, Mick The WB itself also spends less on SFs for project preparation Moore, Richard Locke, Robert Kaufman, Anu Joshi, Steen Lau and supervision than on other projects run through existing min- Jorgensen, Carol Graham, Ruth Dixon, and Soniya Carvalho. I istries or agencies in education and health, economic infrastruc- also benefited substantially from feedback received at seminars ture, and for targeted or participatory poverty projects. The cost at the World Bank, the Inter-American Development Bank, the of WB input into the SF projects varied from 39 to 85 percent David Rockefeller Center for Latin American Studies at Harvard, of equivalent costs for comparator projects [WB (1997a: 42-43, the Institute for Development Studies at Sussex, and the School including Table 6)]. These lower costs, however, are not neces- of International and Public Affairs at Columbia University. I sarily related to the SF model in itself, but to the fact that the thank Sandra Hackman for excellent editing. WB does not make disbursements on SF loans contingent on Funding for the field research in Brazil was graciously provided "policy conditionality," which can slow down disbursements on by the state governments of Ceara and Maranhao, the Bank of the these other projects substantially [WB (1997a: 42, and note 55)]. Northeast, and the World Bank. For financial support for writ- 6. IDB (1997a: 71), and WB (1997a: vi). ing the larger document from which this chapter was drawn, I 7. See, in particular: (1) Portfolio ImprovementProgram Re- thank the MIT/UNDP Decentralization Project and the Division view of the Social Funds Portfolio (WB 1997a); see also WB 1998a iZ6 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION (by Wiens and Guadagni); WB 1998b (edited by Bigio et al.); chapter (see also his co-authored contribution, WB 1998a). The and WB 1998c (by Owen and Van Domelen); (2) Social Invest- programs I researched in Brazil, funded by the WB, are now ment Funds in Latin America: Past Performance and Future Role characterized in the WB SF lexicon as DRIFs. (IDB 1997a, 1997b; see also IDB (1998); (3) a chapter on SFs 13. This dynamic is chronicled for Peru's SF by Schady in Safety Net Programs and Poverty Reduction: Lessonsfrom Cross- (1999), Roberts (1996), Roberts and Arce (1998), and Graham Country Experience (WB 1997b); and (4) a review for UNICEF and Kane (1998); and for Mexico's SF, by Gibson (1997). See by Sanjay Reddy, Social Funds in Developing Countries: Recent also endnote 63. Experiences and Lessons (UNICEF 1998). All these studies are 14. IDB (1997a: 71). Lustig reports that even the most well- thoughtful and candid attempts to review the SF experience. To known, older, and highly praised Latin American SF, the Bo- the extent that half of the Latin American SFs are funded by both livian Social Emergency Fund (starting in 1986), employed the WB and the IDB (9 out of 18), there is a significant overlap roughly only 6-8 percent of workers in the two lowest income in the experience on which they both report. As of May 2000, deciles. The Honduran fund employed only 7 percent of the un- the Operations Evaluation Department of the World Bank is ini- employed (1990-1995), the Peruvian fund, 2.7 percent (1991- tiating a comprehensive study of SFs; in addition, the WB has 1995), and the El Salvadoran fund, 2.5 percent (starting in undertaken a set of six case studies of SFs to look into their rela- 1990). (For the Guatemalan fund, no data on employment gen- tionship to local government. eration were even gathered.) Data are from Lustig (1997: 4-5), For the interested reader, I have included in the Bibliography citing as sources the WB (1997a) for Bolivia; and IDB-funded to this chapter additional relevant works that were cited only in studies of El Salvador, Guatemala, and Honduras. its longer version (Tendler 1999). 15. IDB (1997a: 71). In a study of the employment-creating 8. See, in particular, Lustig (1995, 1997); Stewart and van works programs in various developing countries funded out of der Geest (1995); Cornia (1999); Cornia and Reddy (2000); and U.S. agricultural surpluses (Food for Work), Thomas (1986: 26) Hutchful (1994). reports an average 52 percent of total expenditures on labor, with 9. I report on this research in a longer version of this chapter a maximum of 77 percent; von Braun et al. (1992) stipulate at (Tendler 1999), where extensive evidence and citations for the least 60 percent for labor expenditures as desirable for African claims made here can be found; see also Serrano (1996). The programs. Studies of the Maharashtra Employment Guarantee field research looked into the SFs of four states in Northeast Scheme in India (Costa 1978; D'Silva 1983) show how labor in- Brazil-Ceara, Maranhao, Bahia, and Pernambuco. tensity varies with the kind of project-water projects using the 10. Not all SFs are explicitly demand-driven. A recent WB largest percentage (80 percent) and road projects the lowest (55 review reported that between 10 and 40 percent of the SFs use percent). More recently, the Maharashtra Scheme has required demand-driven mechanisms (WB 1997a: 24). The narratives that at least 60 percent of total costs be spent on unskilled labor about SFs and their strengths nevertheless often describe them (Deolalikar and Gaiha 1996). as "participatory," if not demand-driven. 16. IDB (1997a: 22, 71). The WB found similar results in 11. The evidence on quick disbursement is actually somewhat Honduras (Webb et al. 1995). The evaluators also note that es- mixed, as reported by the WB study of three social funds in Latin timates of SF job creation are often overestimated because of the America (1998: xvii), Stewart and van der Geest (1995), and in large amount of temporary employment that usually lasts only a the complaints of project agency managers about the way com- few months (p. 22). munity decisionmaking "slows down" the rate of disbursement. 17. In reporting these findings, Stewart and van der Geest The WB report attributes cases of slow disbursement to delays (1995) note that these unimpressive outcomes for benefits are by the central government in providing counterpart funding to partly a result of the fact that governments in SF countries com- the projects. Stewart and van der Geest attribute the problem mitted more resources to these non-donor-funded programs than to the demand-driven structure itself, which results in a time- they did to the SFs. But even if SF countries had committed consuming process of community- and municipal-level organiz- more resources, the authors claim that their calculations show ing and decisionmaking. They also point to the concern of proj- that the SFs would still have reached only a smaller share of the ect agencies about "clientelism" and political meddling in such unemployed in the lower deciles because of their greater diffi- decentralized project selection and location, which causes agency culty in targeting (p. 126). managers to impose criteria and requirements that slow things 18. IDB (1997a: 22-23). down. Their concern about reducing delay is at odds with the 19. IDB (1997a: 23). WB study (1997a) that suggests that more time and attention be 20. Peru is one of the exceptions, and perhaps for this reason paid to imposing project criteria that assure better participation one of the most studied cases with respect to this particular kind and inclusion of the poor. of outcome. Schady (1999), among others, show that the Peru- 12. Most importantly, the earlier funds were fashioned to be vian SF succeeded in concentrating a large percentage of its temporary and emergency interventions-usually bearing the expenditures in the poorest districts. (Because of difficulties name of Social Emergency Funds (SEFs). Subsequent SFs, or later of available data, as is typical, measures of the distribution of phases of earlier ones, were viewed as more permanent-both in project benefits within districts could not be made). terms of their institutional design and the community projects Given the bad name that political meddling has in these types themselves, and as reflected in the slight name change to Social of programs-and the donor claim that SFs can operate au- Investment Funds (SIFs); the 'WB made an even later distinction, tonomously-it is important to note that this success at target- stressing those SFs with particularly explicit demand-driven fea- ing is directly attributed to President Fujimori's electoral strat- tures and designating them as Demand-driven Rural Investment egy to court this most marginalized rural population in his Funds (DRIFs). I thank Maurizio Guadagni for pointing out this second electoral bid (they had not voted for him in a previous latter distinction to me in reacting to an earlier version of this referendum on allowing him a second term); more generally, he WHY ARE SOCIAL FUNDS SO POPULAR? 127 needed to find a new constituency to make up for voter disaffec- 29. WB (1998a: xv-xvi, edited by Wiens and Guadagni). The tion in urban areas resulting from his macroreforms (Schady WB text uses the acronym "DRIFs" rather than SFs; DRIFs are 1999; Roberts 1996; Roberts and Arce 1998). a subspecies of SFs called "Demand Driven Rural Investment Similarly, it is interesting that the Peruvian SF has also been Funds" that, according to this classification, support mainly pro- characterized negatively for this political direction as one of the ductive infrastructure and natural resource management. The most "politicized" programs-where support for projects was cen- study reports on three DRIFs in Latin America-in Mexico, trally directed according to political criteria (Graham and Kane Colombia, and Brazil-the latter being the same program looked 1998; Schady 1999). It is ironic, in this case, that one of the SFs' at in my Brazil research. (Actually, the Brazil SFs/DRIFs did not most important objectives-reaching poor communities-was exactly fit this particular description.) achieved through means that are considered antithetical to the 30. IDB (1997a: 71). program model: SFs are said to be better at service delivery, that 31. Lustig (1997: 2-4). Stewart and van der Geest (1995) ar- is, partly because they are "autonomous" and "apolitical." rive at similar conclusions in a study including African as well as 21. As reported by WB (1997a: 18, WB 1998a: xv), IDB Latin American countries. (1997a), Lustig (1995, 1997), IDB (1997a), and Stewart and van 32. IDB (1997a: 23). der Geest (1995). For the 1990-92 period with respect to Mex- 33. IDB (1997a). The citation (p. 16) comes from a Decem- ico's PRONASOL, Cornelius et al. report that middle-income ber 1996 version of this report, as cited in WB (1997a: 47, states received more funds per capita than poor states (as mea- note 58). sured in terms of indices of poverty and underdevelopment) 34. The evaluators also pointed to the inability to truly com- (1994: 22-3). Graham (1994) reports that, more generally, none pare the demand-driven SFs to other programs due to the lack of the poverty alleviation programs in Latin America, Africa, or or poor quality of the data, the classic apples-and-oranges prob- Europe has been particularly successful in targeting the poorest lem of such a comparison, and the limitations of their data and members of the population. The IDB study points out that even methodology. The sample size was small (ranging from 8 to 69); after its own calculations, it is very difficult to determine target- they did not compare SFs to non-Bank-funded programs [as ing from the data, which does not distinguish between rich and Stewart and van der Geest (1995) did]; and they were not able poor within municipalities or higher-level administrative units to separate out, on the SF side, the sectoral piece of the SF pro- from which the data are drawn. gram that corresponded to the comparator project in a functional 22. WB (1997b), as cited in WB (1997a: 18). ministry-health, education, water, roads, and so on. (They also 23. Lustig (1997: 5), citing WB (1997b). More recent com- did not rank the kinds of impacts of unemployment and poverty mentaries have suggested that in the Bolivian cases, the units of reported above.) WB (1997a) study are so large and diverse-whole departments-that in- 35. IDB (1997a: 68). The study notes that this is because of tradepartmental comparisons would be more valid. the reliance on ex-post beneficiary questionnaires for these evalu- 24. Some of the studies show that whereas the SFs did not ations, and the lack of ex-ante data. The report does mention, reach the poorest communities, they often reached communities however, that the impact evaluations are a valuable source of in- that were poor. The IDB study found that the poorest decile mu- formation on whether projects are operating and whether selec- nicipalities received less than the others, but that the nonpoorest tion and construction were satisfactory. poor received more than the best off. As noted earlier, a study of 36. IDB (1997a: 6, 46, 73). The evaluators attributed this the Peruvian SF FONCODES (Schady 1999) found that poorer finding to the "inflexibility" of the donors, and their "rules and communities actually get more SF funding per capita. limitations," which inhibited the ability of local officials to ex- 25. Lustig (1995, 1997) for the comparator programs and SFs periment with innovative solutions. One interesting example of in Latin America, and Stewart and van der Geest (1995) for com- this donor "inflexibility" related to the use of private contractors parator programs in Latin America, Africa, and South Asia. In for works projects. In trying to serve the poverty-reducing goals Lustig's review of the Latin American SFs (1995: 31), she noted of the SFs, donors typically emphasized works projects that that they "compare unfavorably" with such direct-transfer pro- trained and employed local people. This stipulation faced the re- grams (she is considering only the direct-transfer aspects of SFs sistance or noncompliance of private contractors, who usually in the comparison). Lustig, a researcher at the Brookings Insti- preferred bringing in their own workers from outside, particu- tution at the time of her study, drew on various SF evaluation larly for skilled work, and complained that this would compro- studies by the donors. mise their efficiency. In focus-group meetings convened by the 26. Cornia (1999) and Cornia and Reddy (2000). IDB, interestingly, mayors and community representatives ex- 27. The IDB review of SFs found that, for all but one of the pressed more concern about project quality than local employ- countries (Peru), it was not possible to determine the extent to ment, and therefore preferred that contractors use their own which those employed by SFs were poor. [In Peru, an unrelated skilled labor. With respect to "inflexibility," then, the IDB eval- survey from the ongoing WB Living Standards Measurement uators were making the same critique of the donors that the lat- Project had included a question about employment in the SF; 36 ter had been making of line ministries. percent of the jobs went to the extremely poor, and 57 percent 37. The WB review of African and Latin American projects to the poor (IDB 1997a: 32).] reported concerns about sustainability, particularly with respect 28. IDB (1 997a: 15). The study notes that baseline data are to the economic infrastructure and microfinance components of not available for employment and income in the regions served such projects, noting that such concerns had "been raised in by SFs, making the estimate of changes in poverty and income other reviews as well" (WB 1997a: vii, 15-16, including foot- not possible. Data have been collected in several cases, however, note 9). Another WB study (1998a: xvii-xviii, 46) found that on the employment and income generated by the projects them- none of the three Latin American projects (DRIFs/Demand- selves, their benefits, and surveys of project beneficiaries. Driven Rural Investment Funds) it reviewed "performed partic- I28 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION ularly well in achieving" sustainability, and that "information (1995: 128) reported that poorer communities present fewer pro- from local or partial surveys suggests that a high proportion of posals for funding than richer communities. Herring (1983) and subprojects may not be sustainable." A WB appraisal report for Gaiha (1998) found the same kinds of results in similarly targeted a Senegal SF/AGETIP noted that the "sustainability of many and deconcentrated programs in India. A study of the education AGETIP investments is uncertain," due to a lack of ownership projects financed by the Mexican SF found that the program's and participation in the project identification and preparation requirement that a community have an effectively functioning phase and in the post-project operations and maintenance phase Solidarity School Committee before seeking funding explained [WB Senegal PAR Public Works and Employment Project, 1996 why fewer per capita funds went to poor indigenous communi- draft, page 2 notes, as cited in WB (1997a: 15:note 9)]. ties as compared to others (Gershberg 1994: 249-51). The IDB came to similar conclusions (1997a: 35-41), and There may be an inherent tendency for exacerbation of this an earlier 1994 IDB study cautioned that "[s]ustainability re- problem in that the better-off communities that are successful in mains a potentially serious problem . . ." [Glaessner, Lee, San- getting one project will come back for subsequent ones, and pre- t'Anna, and de St. Antoine, "Poverty Alleviation and Social In- pare them better, while communities that are turned down or vestment Funds: The Latin American Experience," p. 22, as cited have a difficult time will become discouraged and desist, a point inWB (1997a: 15)]. made by Schady (1999). I thank Mick Moore for pointing out 38. WB (1997a: 31). the importance of looking at subsequent choices of projects be- 39. Eighty percent of the project descriptions did not men- cause of the learning that results from "repeated games." tion sustainability, nor concern themselves with its three key 47. IDB (1997a: 15). components: (1) evidence of demand (range of options offered, 48. In another twist on the tradeoff theme, the Stewart and van information made available, evidence of commitment through der Geest (1995) comparator study of SFs with earlier supply- contribution in cash or kind); (2) appropriateness of technical driven programs claims that SFs actually disburse less rapidly than standards; and (3) soundness of arrangements for operations and the comparators, and pointed to the consultation process and the maintenance (funding, and training) (WB 1997a: 30). demand-driven design: if taken seriously and at its best, it resulted 40. WB (1997a: 30-31). in a time-consuming process of organizing and decisionmaking by 41. WB (1997a: 15-16, and note 9). The WB evaluators re- communities or municipal councils. inforce their concerns about sustainability with citations from 49. See Serrano (1996), and Tendler (1999, Sections III their sister SF financing institution, the IDB, and from other re- and IV). viewers within the WB itself. They also question whether SF de- 50. IDB (1 997a: 39). In many communities, the report said, signers and managers even thought about project designs and NGOs are not that active. In addition, NGOs tended to special- technical standards that would be more likely to elicit user main- ize more in training and community development programs than tenance and financing for recurrent costs. They point out, it in managing the construction projects that constitute an impor- should be noted, that their findings relate more to "likely," as tant activity of many SFs. In the SFs where NGOs played a opposed to actual, sustainability, because only a limited number greater role, then, it was because the program did not focus on of the individual country evaluations it drew on involved SF building infrastructure (like Chile's FOSIS). Other exceptions projects with long-term objectives (p. 4). were cases in which the government was "institutionally ex- 42. WB (1997a). tremely weak" to the point that NGOs had more capacity to gen- 43. The aforementioned study of SFs now being studied will erate projects than government (Haiti) and, in general, because hopefully yield more evaluation information on this issue. In ad- the SF was formally required to use them. The usual tension that dition, six country studies now being carried out by the WB by exists between NGOs and government also seemed to get in the Andrew Parker (Africa) and Rodrigo Serrano (Latin America) to way. The NGOs disliked being the mere executors of a "pater- explore the impact of SFs on local governments may also provide nalistic" government program and wanted to participate more in illuminating, findings on the issue. early phases of the project cycle. The SF managers and staff, pre- 44. IDB (1997a: 74). In Latin America, out of 16 countries sumably, were not anxious to do this. and 17 SFs (Guatemala has two), Chile's FOSIS has the lowest 51. The IDB review reported that the "recurrent-cost prob- level of external financing-Il percent. The next lowest are lem" was most acute in the case of NGOs (IDB 1997a: 41). A Guatemala's FONAPAZ (12 percent), and Colombia's RED WB Project Completion Report [cited in WB (1997b: 107) and SOLIDARIDAD (20 percent). (The IDB evaluators, as already also in the previous IDB reference] noted a study of the Bolivian noted, ranked these three as the most successful in terms of in- SF showing NGOs to be disproportionately represented among novative practices.) For the rest, external financing ranges from the projects that were least likely to be sustained. The projects 58 to 94 percent, with only three countries being lower than 80 were in health and education, and the study was conducted one percent (albeit higher than 60 percent)-Peru, Uruguay, and to two years after completion. (Lower performers on the "sus- Venezuela (IDB 1997a: 10, T. 2.1). The Mexican SF, PRONA- tainability" measure also included projects requested by regional SOL, is also one of the SFs most "owned" by its government: It government, as opposed to central-government, institutions.) was also initiated by the Mexican government without donor 52. Reddy's review for UNICEF (1998: 58) noted that funding and is one of the largest in terms of absolute resources, "[flavouritism in the disbursal of contracts to NGOs" was a seri- share of the budget, and coverage (Cornelius et al. 1994: 14). It ous issue in various countries, as was the "proliferation of NGOs does not appear in this particular table of the IDB because it is of dubious grassroots credentials" as a result of the new availabil- currently not receiving donor funding, though it has in the past. ity of SF funding. 45. IDB (1997a: 64, 74). 53. Some SF managers expressed a certain distaste for, and 46. WB (1997b); IDB (1997a: 15, 43). Based on studies of therefore sometimes discouraged, genuine processes of commu- the Bolivian and Honduran SFs, Stewart and van der Geest nity decisionmaking. These processes, they said, "slowed down" WHY ARE SOCIAL FUNDS SO POPULAR? I29 the rates of disbursement so prized by these managers and their teragency cooperation. This same problem plagued the earlier in- donors (WB 1997a; IDB 1997a). [Other managers, interestingly, tegrated urban and rural development projects, so much in vogue actually liked the eligibility criteria which, even though slowing in the 1970s and 1980s, with their autonomous project units. down disbursement, gave them some kind of protection against The problems of this autonomy contributed to the keen distaste political interference (Stewart and van der Geest 1995)]. today among the donors for such integrated projects, and hence 54. See, for example, Stiglitz (1985). their rapid fall from grace in the late 1980s. An even earlier donor 55. See, for example, his speech to the World Institute for penchant for creating special autonomous state enterprises, in Development Economics Research (WIDER) (1998). the 1950s and 1960s, also fell out of fashion, for partly similar 56. Rodrigo Serrano (1996) reported this three-part set of ac- reasons. If donors were slapping their own hands about the pit- tors in driving community decisions, which emerged from the falls of independent units created solely for their programs in the larger Brazil research program; see also Tendler (1999), Sections 1980s, what is different about SFs or the 1990s that would re- III, IV, and V. The IDB evaluation also points to the importance verse this judgment? of contractors in driving community decisions in many cases. 62. One such recent meeting was held in Mexico in Decem- The practice was apparently common enough that the IDB eval- ber of 1999, and another is planned for Washington, D.C. in uators baptized it as "persuasion by contractor," and cautioned June of 2000. that "the real beneficiaries" in these cases might well be "the con- 63. These arguments have been made at great length and with tractors" rather than the final users (IDB 1997a: 41, 43). detailed case study evidence by Gibson (1997) for Argentina and 57. For example, WB (1997a: vii, ix, 15). Mexico, and, for Peru, by Schady (1999), Roberts (1996), and 58. Some question the evidence on this claim on the grounds Roberts and Arce (1998). In developing the evidence from their that systematic and comparable cost data are not available or not cases, these authors specifically link the compensatory patron- collected (sometimes even across different SF offices in the same age-dispensing policy to the SFs in Mexico and Peru. I elaborate country), and important costs-such as donor monitoring-are on this point and the evidence in Tendler (1999, Section VI). not included. Some SFs, for example, do not include their own 64. The examples are (1) the tragedy of the commons, which overheads in reporting unit costs; for Peru, see Schady (1998: 5). predicts that collective-action attempts to prevent overuse of 59. Angell and Graham (1995). grazing lands and other common-property resources will always 60. IDB (1997a: 34, and 1997b: 38-76, particularly pp. 46, be doomed to failure because of the free-rider problem; (2) the 48, 73, 74). Also different, the Chilean government viewed belief that the lack of secure and privately held land titles holds FOSIS as apennanent program from the start (it was created dur- back farmers from investing in increased productivity, which ing a time of high economic growth of 7 percent a year); this con- formed the basis for much policy advice and intervention di- trasts with the originally temporary status of the majority of rected to convert common holdings to individually held parcels; Latin American SFs and the origins of most SFs in "temporary" and (3) the concept that the integration of economic activities periods of low growth, high unemployment, and structural- (such as livestock management in his study) into larger systems adjustment or other crises. requires that programmatic interventions also be integrated by 61. It should be noted that this recognition of the problems bringing together various agencies in one program and requiring of special autonomous institutions is definitely not new, particu- them to coordinate-which led, in turn, to a generation of larly the way such special status exacerbates the fabled lack of in- "integrated" rural and urban development programs. I Building ompetitive Cities for t e 2 S1st Centur Ii.- chapter 15 What Distinguishes Success among Second-Tier Cities? Ann Markusen Urban and Regional Planning Program, Humphrey Institute University of Minnesota Council on Foreign Relations O_ n many countries "second-tier," or district, depicting the dominant relationships among midsized, cities have been growing firms and other agents in urban economies. We conclude faster than first-tier cities over the that national governments should continue to monitor past few decades. In Brazil, Japan, the and shape urban growth patterns, selectively intervening Republic of Korea, and the United in dynamic patterns that are worsening inter- and in- States, the employment growth rates traregional disparities. We warn that wholesale devolu- of at least half a dozen medium-sized cities have ex- tion, when unmatched by adequate financial resources, ceeded those of the largest metropolitan areas in each training, or institution-building at the subnational level, country (see Table 15.1). Why have these second-tier may paradoxically set back economic development. cities grown so fast? What can be learned from their pat- Three Urban Economic Prototypes terns of development and economic structure? This chapter summarizes the results of five years of Since the early 1980s, economic geographers and re- comparative research on the fastest-growing second-tier gional economists have elaborated on Alfred Marshall's metropolitan areas in Brazil, Japan, the Republic of insights to focus on the industrial district as a core com- Korea, and the United States by a team of economists ponent of economic regions (Markusen 1996). Relying and geographers based in universities in each country heavily on the experience of the "third Italy" and Silicon (Markusen, DiGiovanna, and Lee 1999). Specifically, Valley in "flexible specialization," some analysts have we research the structural and voluntaristic features of claimed that these "new industrial districts" constitute a three to five of the fastest-growing second-tier metro- novel regional form that is growing disproportionately politan areas in each country. Our research uncovers sev- fast (Scott 1988). However, we found that the character- eral prominent contributors to the success of such cities: istics associated with such districts explain only a mod- est number of rapidly growing second-tier cities. Many * Specialization in innovative and/or income-elastic eco- more fully reflect two other new industrial models: the nomic sectors hub-and-spoke structure and the satellite platform. * Favorable national government policies (investment, In investigating these three prototypical structures, infrastructure, and industrial and regional policies) as well as the keys to success of second-tier cities, we * Leadership in economic development by the local identified metropolitan areas with growth rates signifi- public or private sector cantly higher than the national average. We then chose * Flexibility and openness to new and emerging indus- a subset as case studies, including at least one in each of trial ensembles. the four countries that conformed to the new industrial districts (NID) formula, and three to five others whose In the course of our study, we develop two new struc- industrial structure and organization appeared to be dif- tural prototypes, contrasted to the Marshallian industrial ferent. Although we have not attempted to fit all fast- 132 WHAT DISTINGUISHES SUCCESS AMONG SECOND-TIER CITIES? I33 Table 15.1 Second-Tier City Employment, Brazil, Japan, Republic of Korea, and the United States, Selected Years 1970-90 Employment (thousands) % Change Employment (thousands) % Change Brazil 1970 1991 1970-91 Republic of Korea 1981 1991 1981-91 Second Tier Second Tier Manaus 10 57 477 Ansan na 134 na Londrina/Maringa 13 46 246 I Changwan 44 141 220 Curitiba* 38 128 237 I Taejon 122 284 133 Joinville 22 73 234 i Kwangju 127 275 117 Fortaleza* 25 80 223 Kumi 52 110 112 Caxias do Sul 23 75 221 Inchon 272 560 106 Belo Horizonte* 62 173 182 Pohang 55 107 95 Franca 9 27 182 Ulsan 111 207 86 Campinas 71 193 172 Masan 86 136 58 Piracicaba 19 50 166 First Tier Salvador* 28 71 156 Taegu 360 558 55 Blumenau 29 74 156 Seoul 2,372 3,629 53 Sorocaba 35 88 151 Pusan 734 1,108 51 Santos/Cubatao 35 88 151 Korea 6,602 11,356 72 Belem* 14 33 128 Americana 13 28 122 Sao Jose dos Campos 47 100 114 M First Tier Sao Paulo* 907 1,410 55 Rio de Janeiro* 224 313 39 Employment (thousands) % Change Employment (thousands) % Change Japan 1970 1990 1970-90 United States 1970 1990 1970-90 Second Tier Second Tier Sayama 13 23 73 Orlando, FL 164 569 246 Anjyo 24 41 70 Anaheim-Santa Ana, Hatano 12 21 69 CA 532 1,552 192 Komaki 20 33 65 Austin, TX 169 471 178 Kasugai 22 31 44 Reno, NV 57 145 155 Oita 18 26 40 San Jose, CA 445 1,015 128 Kawagoe 22 30 35 Albuquerque, NM 136 305 125 Yao 35 47 33 Tucson, AZ 142 316 123 Sagamihara 41 53 28 Raleigh-Durham, NC 230 513 123 Atsugi 20 25 26 San Diego, CA 635 1,397 120 Hiroshima 56 69 23 Seattle, WA 626 1,339 114 Kohriyama 22 27 22 Melbourne/ Suzuka 22 27 20 Titusville, FL 95 202 112 Kariya 40 47 18 Colorado Springs, CO 112 228 104 Nagaoka 21 24 17 i Huntsville, AL 93 163 76 Ueda 18 21 14 Madison, WI 151 262 73 First Tier Elkhart, IN 71 116 64 Nagoya* 981 957 -2 First Tier Osaka* 1,126 863 -23 Los Angeles, CA 3,333 5,200 56 Tokyo* 1,394 785 -44 Boston, MA 1,286 1,672 30 Japan 11,675 11,173 -4 Chicago, IL 2,986 3,673 23 New York, NY 4,671 4,765 2 United States 70,719 110,321 Sources and notes: Data are compiled and explained in separate tables in Ann Markusen, "National Contexts and the Emergence of Second Tier Cities," in Ann Markusen, Yong-Sook Lee, and Sean DiGiovanna, eds., Second-Tier Cities: Rapid Growth beyond the Metropolis, Minneapolis: University of Minnesota Press, 1999. * Brazil: connote metropolitan area data. Japan: Figures are for industrial employment only. Data for Tokyo, Nagoya, and Osaka are for the prefectures, not metropolitan areas. I34 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION growing cities into this typology, it is our working hy- Figure 15.1 pothesis that most such cities can be characterized as Alternative Regional Industrial Structures possessing one or an amalgam of these structures. Building on enterprise studies and corporate inter- views as methods for studying regions (Markusen 1994), A. Marshallian Industrial District we relied on interviews with business firms, trade asso- ciations, trade unions, and regional economy watchers to map the relationships among firms and other eco- nomic actors inside and outside the region. We specifi- cally focused on features in the NID formula: firm size, Suppliers Customers upstream and downstream industrial links, vertical inte- gration, networks among district firms, local gover- nance, innovative capabilities, and the organization of production. We also scrutinized the role of large firms B. Hub-and-Spoke District and that of the state at both the national and regional/ local level, as rule maker, producer and consumer of goods and services, and underwriter of innovation. Fi- nally, we investigated the longer-term dynamic of major industries to determine their resiliency. The Flexibly Specialized Industrial District The literature on new industrial districts focuses on re- gions where the business structure is composed of small, locally owned firms that make investment and produc- C. Satellite Platform District tion decisions locally (top of Figure 15.1). Scale econo- mies are relatively modest, forestalling the rise of large firms. Buyers and sellers transact substantial trade within the district, often based on long-term contracts or commitments, and often for eventual export from the region. Such a district also encompasses a relatively spe- cialized set of services tailored to its unique products and industries. They include local financial institutions of- fering "patient capital"-that is, they are willing to take longer-term risks because they have access to inside in- formation and trust local entrepreneurs. Q large locally headquartered firm What makes this type of industrial district so vibrant 0 small, local firm and productive, in Marshall's account, is the nature and quality of the local labor market, which is highly flexi- L branch office, plant ble. Individuals move from firm to firm, committed to the district rather than to any one company. Labor out- migration is minimal, while in-migration occurs as growth permits. Owners and workers live in the same community, where all benefit from the fact that "the se- keting, technical, and financial help, as well as forums crets of industry are in the air." This relatively stable to hammer out collective strategy. Local and regional community enables a local cultural identity and shared governments may be central in regulating and promot- industrial expertise to evolve. ing core industries (Markusen 1996). Scholars of new industrial districts hypothesize that robust growth rates are due, too, to cooperation among The Hub-and-Spoke Industrial District competitor firms to share risk, stabilize markets, and In the hub-and-spoke region, a number of key firms or share innovation. Disproportionate shares of workers are facilities act as hubs for the regional economy, with sup- engaged in design and innovative activities. Activist pliers and related activities spread out around them like trade associations provide management, training, mar- spokes of a wheel (B in Figure 15.1). The dynamism in WHAT DISTINGUISHES SUCCESS AMONG SECOND-TIER CITIES? 135 hub-and-spoke economies is associated with the posi- for firms bristling under high urban wages, rents, and tion of the large firms in their national and international taxation. Tenants of satellite platforms may perform markets. Key investment decisions are made locally but functions ranging from routine assembly to relatively so- their consequences are spread globally. For example, a phisticated research, but they must be spatially indepen- single large firm such as Boeing in Seattle or Toyota in dent of upstream or downstream operations in the same Toyota City buys from local and external suppliers and firm, and of agglomerations of suppliers, customers, and sells chiefly to external customers, who may be large (as competitors. in the case of Boeing) or numerous (as in the case of The most conspicuous feature of this model is the ab- Toyota). Other examples include central New Jersey in sence of any networks within the region and the pre- the United States, Ulsan and Pohang in South Korea, dominance of links to the parent corporation and branch and San Jose dos Campos in Brazil. plants elsewhere (C in Figure 15.1). Large, externally Unrelated or loosely linked hubs may coexist in a re- situated firms make key investment decisions, and or- gion. The Seattle economy, for instance, encompasses ders and commitments to local suppliers are noticeably paper manufacturer Weyerhauser as the dominant re- absent. Since platforms generally host heterogeneous source sector company, Boeing as the dominant indus- firms, they do not share risk, stabilize the market, or en- trial employer, Microsoft as the leading services firm, the gage in innovative partnerships. The labor market is in- Hutchinson Cancer Center as the progenitor of biotech- ternal to the vertically integrated firm rather than to the nology firms, and the Port of Seattle as the transporta- district, producing high rates of migration at the man- tion hub. Core firms or institutions are embedded non- agerial, professional, and technical levels. locally, with substantial links to suppliers, competitors, Almost all countries have satellite platforms. An and customers outside the district. Internal scale and outstanding U.S. example is North Carolina's much- scope economies are relatively high, and turnover of admired Research Triangle Park, a collection of unrelated firms and personnel relatively low. research centers of major multinational corporations Hub-and-spoke districts may exhibit intradistrict co- (Luger and Goldstein 1990). In South Korea, Kumi con- operation, but generally on the terms of the hub firm. stitutes a low-end textile and electronics platform, while Markedly lacking is cooperation among competitor Ansan operates as an intentional ensemble of disparate firms to share risk, stabilize the market, and share inno- industrial polluters (Park and Markusen 1994, 1999). vation. Larger firms are more apt to forge strategic al- Even Taeduck can be regarded as a satellite platform liances with partners outside the region. Workers' loyal- because R&D centers in the science park have few ties are to core firms first, then to the district, and only local links (Jeong and Park 1999). Some of the better- after that to small firms. If jobs open up in hub firms, performing Japanese technopoli, such as Oita and Ku- worker will often abandon smaller employers to get on mamoto, fall into this category (Markusen and Sasaki the hub firms' payroll, but turnover of firms and per- 1999). A remarkable Brazilian case is the state-sponsored sonnel is relatively low. expansion of Manaus as an import/export zone (Diniz In hub-and-spoke districts, the few trade associations and Borges Santos 1999). are relatively weak, often because top hub managers ab- Satellite platforms can and do evolve into rapidly sent themselves from their activities. Hub firms concern growing second-tier cities, but they have little "patient themselves with state and local governmental activities capital" to draw on because outside corporate headquar- that impinge upon their land use, taxes, and regulations, ters are the main sources of finance, technical expertise, and try to ensure that area politicians represent the in- and services. Satellite platforms also lack industry- terests of their firm and industry at the national and in- specific trade associations that would provide shared in- ternational levels. They may also be actively involved in frastructure and assist with management, training, and issues that affect their workforce and ability to do busi- marketing. Strong efforts by national or local govern- ness, especially in improving area educational institu- ments and services from Chambers of Commerce and tions and infrastructure. other associations of local fixed capital can compensate only partially for these drawbacks. The Satellite Platform A third distinctive industrial form is the satellite plat- Industrial Specialization form-a congregation of branch facilities of firms based One of our most interesting findings regarding second- elsewhere. National or provincial governments often as- tier cities, no matter what their economic prototype, is semble these platforms to stimulate economic develop- that they must specialize more than ever before, given ment in outlying areas and lower the cost of business rapidly integrating world trade and investment. Some 136 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION continue to rely on hinterlands that specialize as agri- role in that city's growth (Diniz and Razavi 1999), while cultural processing centers or mining districts, but most Korean investment in research institutes and the Korea have lost hinterland markets to competitors, as distance Institute of Technology in Taeduk have spurred growth costs or hinterland competitiveness shrinks. in their host cities (Jeong and Park 1999). In general, the more successful cities-especially However, the presence of excellent research universi- those we classify as Marshallian or hub-and-spoke-are ties does not guarantee superior economic performance. built on one or more specializations that are well posi- Many of America's best research universities have had tioned in national and international trade because they no appreciable impact on regional development, while are innovative or income-elastic, or simply because they cities like Colorado Springs have thrived without re- exhibit comparative advantage (Howes and Markusen search universities. Still others, such as San Diego and 1993). Autos, for instance, are an income-elastic sector Seattle, have used revenues to create such universities because as people's income rises in host and neighbor- after the fact. ing nations, they spend a larger share of their earnings Explicit regional policy, often as a component of na- on cars. Port cities with containerization facilities and tional industrial policy, has played a strong role in shap- good rail and highway links to the interior are similarly ing the placement and growth of second-tier cities favored by growth in the volume of trade (Noponen, (Markusen 1995). Examples are Korea, where the cen- Markusen, and Driessen 1997). Examples of successful tral government built and still runs entire industrial specialization include autos and auto parts in the Re- complexes within which firms are directed to locate public of Korea's Ulsan and Changwon, electronics and (Park and Markusen 1994), and Japan, where incentives telecommunications in Brazil's Campinas (Diniz and for industrial location have long been used to stimulate Razavi 1999), electronics in Japan's Oita and Kuma- regional growth, most recently via the technopolis pro- moto (Markusen and Sasaki 1999), and aircraft and gram (Funaba, Sasaki, Lee, and Markusen 1999). software in Seattle (Gray, Golob, and Markusen 1996). Public investments in infrastructure have also been sig- Of course, not all cities specialize, especially satellite nificant in explaining the relative growth of cities within platforms. But when they do, specializations tend to nations. The U.S. investment in railroads-and more re- emerge on their own, as did pharmaceuticals and elec- cently the interstate highway system-made a major con- tronics in North Carolina's Research Triangle Park tribution to postwar population growth in second-tier (Luger and Goldstein 1990). A number of government cities (Isserman, Rephann, and Sorenson 1989). Brazil- attempts to confine new industrial complexes to certain ian infrastructure investments have done likewise, al- sectors have failed or required modification. For exam- though since states have contributed much of the invest- ple, the Republic of Korea's Kumi, designed as an elec- ment, decentralization has tended to occur mostly within tronics export complex, ended up with a significant richer regions such as Sao Paulo (Diniz 1994). number of textile plants (Park and Markusen 1999). Of course, not all second-tier cities owe their success to high levels of public investment or activity. The Re- Public-Sector Investment and Directives search Triangle area, while it did land a major federal fa- Most of the cities we studied benefited extraordinarily cility early on, today attracts very low levels of defense well from public-sector investment, procurement, and/or or other procurement spending. The park owes its suc- regional programs, including national defense. United cess to other features, particularly its pioneering activi- States examples include Silicon Valley, Seattle, and Col- ties as a satellite platform (Luger and Goldstein 1990). orado Springs (Gray, Golob, Park, and Markusen 1999; Nor do public-sector investments and directives in- Gray, Golob, and Markusen 1996; Markusen and Gray evitably turn out well. Despite heavily subsidizing Man- 1999). Brazil's San Jose dos Campos, which hosts the aus as an export platform in the Amazon, and dracon- nation's publicly owned defense-related aircraft sector ian attempts to suppress manufacture of consumer (Diniz and Razavi 1999), and the Korean city of Chang- electronics elsewhere, the Brazilian government has cre- won, built in the 1970s to serve as the nation's defense ated a deformed industrial district that is now largely an industrial complex (Markusen and Park 1993), are other entrepot for electronics imports and a major drain on examples. the national treasury (Diniz and Borges 1999). Land-grant universities in the United States, funded first by the national government and later by state gov- Public and Private Leadership ernments, have contributed to the decentralization of Porter (1990), among others, has shown that state and urban growth and the remarkable postwar success of local politicians, business leaders, and economic devel- several dozen cities. Similarly, investment by the state of opers have often designed unique and effective strate- Sao Paulo in its university in Campinas played a pivotal gies by playing to their regions' strengths. And indeed, WHAT DISTINGUISHES SUCCESS AMONG SECOND-TIER CITIES? 137 subnational leadership helps explain the relative success we do not yet know much about this process, certain re- of some cities in our study. A U.S. governor along with gions seem to possess an ability to learn and to shift gears local developers played a key role in creating Research to new types of activities when threatened with obsoles- Triangle Park (Luger and Goldstein 1990). A Japanese cence or decline caused by trade or shifting national pol- governor in the prefecture of Kumamoto played a sim- icy. Seattle had the good fortune that Boeing, despite its ilar role (Markusen and Sasaki 1999). In Colorado active suppression of local suppliers and spin offs from Springs, a private economic development group com- the parent company, was relatively indifferent, if not posed mainly of local business interests ("fixed capital") welcoming, to the rise of Microsoft and the region's new, recruited the Air Force Academy and the North Amer- small biotech industry (Gray, Golob, and Markusen ican Space Command, and later electronics branch 1996). plants from Silicon Valley (Markusen, Hall, Campbell, The most outstanding instance of a deliberately- and Deitrick 1991). In the case of these satellite plat- even aggressively-flexible strategy is Colorado Springs. forms, leaders have focused on recruiting firms from Having once hedged against its dependence on military outside the area. However, other localities cannot always installations by attracting electronics branch plants, it emulate such strategies, as the number of beneficiaries responded to the exit of a number of such plants by re- depends upon the available private and public facilities. cruiting telemarketing firms, then sports organizations Leaders in some cities have attempted to retain and (including the U.S. Olympic Committee), and most re- strengthen existing, often mature firms and encouraged cently Christian organizations such as Focus on the entrepreneurship and startups by existing residents, Family, with its publishing and TV empire. In this "re- managers, and employees. In the northeastern United cruit and parlay" activity, local business elites trans- States, this form of development leadership has emerged formed fixed assets (physical structures, labor force, cul- as prospects for successfully competing with southern ture) accrued in one era into attractors for a next round and overseas localities have declined. of economic activity (Markusen and Gray 1999). To confound matters, we identified robust instances We found instances of this in other countries as well. of metropolitan growth without much public- or private- In Changwon, the complex was able to parlay its ma- sector leadership at all. Seattle has benefited from very chining expertise into attracting an auto parts sector little concerted economic development activity over the (Markusen and Park 1993). In Campinas, a similarly decades. And, with the possible exception of Boeing's at- open environment has encouraged diversification of sec- tention to the region's elementary and secondary educa- tors in that metropolitan area (Diniz and Razavi 1999). tional system, neither Boeing nor Microsoft has exerted Masan, South Korea, originally designed for Japanese much local leadership (Gray, Golob, and Markusen branch plants but subject to recent plant closings on a 1996). relatively large scale, has attracted smaller Korean firms Private-sector entrepreneurship-the incidence of in diverse sectors (Lee 1999). startup firms-has proved important in some American The antithesis of this kind of culture is that of De- cities, but the conditions that promote (or suppress) troit, where a dominant, often oligopolistic industry such entrepreneurship are not well understood (Gray, uses markets in land, labor, capital, and politics to sup- Golob, and Markusen 1996). Such entrepreneurship press other sectors (Chinitz 1960; Checkland 1975; proved important in Seattle, for example, through the Markusen 1985). work of Bill Boeing and Bill Gates. However, in other U.S. cases-including Research Triangle and Colorado Implications for Urban Policy in the Springs-and most of the other countries we studied, Twenty-First Century private-sector entrepreneurship was not particularly im- Several conclusions from our study are germane to the portant in explaining the growth of second-tier cities. prospects for subnational economic development. First The relative absence of small entrepreneurial firms in of all, while there are remarkably successful second-tier lead sectors in second-tier cities led us to speculate that cities, many benefited from one-of-a-kind circum- startups are more likely to cluster in the largest metro- stances, making their success difficult for other regions politan areas. to replicate. What's more, there are no magic bullets. Al- though some factors contributed to growth of multiple Openness to New and Emerging g .g Industrial Ensembles second-tier cities, no one factor was present in all cases. However, investments by national governments in Second-tier cities that have done well, especially more the form of infrastructure, facilities, procurement, and mature ones facing declining sectors, do have business revenue sharing have clearly been important determi- and public cultures that embrace new sectors. Although nants of metropolitan growth. Action by urban and re- I38 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION gional leaders has also been important, although of cupations), guidance in economic development, and as- somewhat more limited effect. Overall, we found ample sistance in building their capacity to meet this challenge evidence that good regional policy still matters. Smart and informed choices by national and urban leaders re- Bibliography garding the number of urban centers to nurture, the op- Checkland, S. 1975. The Upas Tree. Glasgow: Glasgow Univer- timal location of infrastructure and other public invest- sity Press. ments, and the deployment of development incentives Chinitz, Benjamin. 1960. "Contrasts in Agglomeration: New man havents major,impact on the distribution of eco York and Pittsburgh." American Economic Association, Papers can have a major impact on the distribution of eco- and Proceedings 40: 279-89. nomic activity as well as on national growth. Diniz, Clelio Campolina. 1994. "Polygonized Development in Successful U.S. second-tier cities like Seattle, Col- Brazil: Neither Decentralization nor Continued Polariza- orado Springs, Silicon Valley, and Research Triangle tion." InternationalJournal of Urban and Regional Research have served as objects of admiration and emulation for 18: 293-314. economic developers around the world. But admiration Diniz, Clelio Campolina, and Fabiana Borges Santos. 1999. "Manaus: Vulnerability in a Satellite Platform." In Ann for the U.S. system must be coupled with a recognition Markusen, Sean DiGiovanna, and Yong Sook Lee, eds., Sec- that significant national public investments, accompa- ond Tier Cities: Rapid Growth Outside the Metropole in Brazil, nied by two centuries of practice in democratic local Korea, Japan and the U.S. Minneapolis: University of Min- governance, are largely responsible for the sustained nesota Press: 125-46. growth of multiple metropolitan areas. Diniz, Clelio Campolina, and Mohamad Razavi. 1999. "San Jose Overall, our results cast doubt on the ability of decen- dos Campos and Campinas: State-Anchored Dynamos." In Ann Markusen, Sean DiGiovanna, and Yong Sook Lee, eds., tralization and devolution to adequately spur urban de- Second Tier Cities: Rapid Growth Outside the Metropole in velopment, balance regional growth rates, and ensure na- Brazil, Korea, Japan and the U.S. Minneapolis: University of tional economic efficiency in allocating resources. For Minnesota Press: 97-124. example, both Korea and Japan have arguably used top- Funaba, Masatomi, Masayuki Sasaki, Yong-Sook Lee, and Ann down means to disperse employment more successfully Markusen. 1999. "Japanese Technopolis Policy: A View from than the more federally structured Brazil. Devolution Four Cities." In Ann Markusen, Sean DiGiovanna, and Yong Sook Lee, eds., Second Tier Cities: Rapid Growth Outside the does not automatically foster balanced regional develop- Metropole in Brazil, Korea, Japan and the US. Minneapolis: ment. In Brazil, where federal and state governments University of Minnesota Press: 239-66. have long shared responsibilities and taxing powers, de- Gray, Mia, Elyse Golob, and Ann Markusen. 1996. "Big Firms, velopment remains concentrated in a polygon around the Long Arms, Wide Shoulders: The 'Hub-and-Spoke' Industrial Sao Paulo-Rio-Belo Horizonte axis (Diniz 1994). This District in the Seatde Region." Regional Studies 30 (7): 651-66. has occurred mainly because the "haves," such as the state Gray, Mia, Elyse Golob, Ann Markusen, and Sam Ock Park. 1999. "New Industrial Cities? The Four Faces of Silicon Val- of Sao Paulo, are able to use state tax revenues to build ley." In Ann Markusen, Sean DiGiovanna, and Yong Sook infrastructure and educational institutions. Economic ac- Lee, eds., Second Tier Cities: Rapid Growth Outside the Metro- tivity has decentralized from the city of Sao Paulo, but pole in Brazil, Korea, Japan and the U.S. Minneapolis: Uni- mainly to medium-sized cities within the state. versity of Minnesota Press: 267-90. Our research documents the past success of and con- Howes, Candace, and Ann Markusen. 1993. "Trade, Industry temporary need for explicit regional policy at the na- and Economic Development." In Helzi Noponen, Julie Gra- temporary need for explicit regional policy at the na- ham, and Ann Markusen, eds., Trading Industries, Trading tional level. The recent trend toward devolution, accom- Regions. New York: Guilford. panied by a retreat from regional policy, may result in Isserman, Andrew, Terance Rephann, and David Sorenson. wasteful allocations of resources, especially if dynamics 1989. "Highways and Rural Economic Development: Results are set in place that eschew some communities in favor from Quasi-Experimental Approaches." Discussion Paper, of concentrated and high-income growth in others. For University of West Virginia, Discussion Paper 8907. Jeong, Jun Ho, and Sam Ock Park. 19993. "Taeduk Research devolurion to work, It must be complemented with Park: Formation of Spin-offs and Local Linkages." In Ann shared revenues, training, and subnational institution- Markusen, Sean DiGiovanna, and Yong Sook Lee, eds., Sec- building. Expecting urban areas to be responsible for ond Tier Cities: Rapid Growth Outside the Metropole in Brazil, their own growth and livability is naive now that-more Korea, Japan and the U.S. Minneapolis: University of Min- than ever before-they are subject to the eclectic choices nesota Press: 199-222. of outside producers, consumers, governments, and in- Lee, Yong-Sook. 1999. "The Masan Free Export Zone: Conflict ternational institutions. Since urban economies must and Attrition." In Ann Markusen, Sean DiGiovanna, and Yong Sook Lee, eds., Second Tier Cities: Rapid Growth Out- specialize to ensure growth, their leaders will require side the Metropole in Brazil, Korea, Japan and the U.S. Min- good economic data (on regional trade, industry, and oc- neapolis: University of Minnesota Press: 183-98. WHAT DISTINGUISHES SUCCESS AMONG SECOND-TIER CITIES? 139 Luger, Michael, and Harvey Goldstein. 1990. Technology in the Markusen, Ann, Peter Hall, Scott Campbell, and Sabina Garden. Chapel Hill, NC: University of North Carolina Press. Deitrick. 1991. The Rise of the Gunbelt. New York: Oxford Markusen, Ann. 1985. Profit Cycles, Oligopoly and Regional De- University Press. velopment. Cambridge, MA: MIT Press. Markusen, Ann, Sean DiGiovanna, and Yong Sook Lee, eds. Markusen, Ann. 1994. "Studying Regions by Studying Firms." 1999. Second Tier Cities: Rapid Growth Outside the Metropole The Professional Geographer 4(4): 477-90. in Brazil, Korea, Japan and the U.S. Minneapolis: University Markusen, Ann. 1995. "The Interaction of Regional and Indus- of Minnesota Press. trial Policies: Evidence from Four Countries." Proceedings, Noponen, Helzi, Ann Markusen, and Karl Driessen. 1997. Annual World Bank Conference on Development Economics "Trade and American Cities: Who Has the Comparative Ad- 1994:279-98. vantage?" Economic Development Quarterly 1 1(1): 12. Markusen, Ann. 1996. "Sticky Places in Slippery Space: A Ty- Park, Sam Ock, and Ann Markusen. 1994. "Generalizing New pology of Industrial Districts." Economic Geography 27(3): Industrial Districts: A Theoretical Agenda and an Applica- 293-313. tion from a Non-Western Economy." Environment and Plan- Markusen, Ann. 1999. "Fuzzy Concepts, Scanty Evidence, ning, A27: 81-104. Wimpy Policy: The Case for Rigor and Policy Relevance in Park, Sam Ock, and Ann Markusen. 1999. "Kumi and Ansan: Regional Studies." Paper presented at the American Associa- Dissimilar Korean Satellite Platforms." In Ann Markusen, tion of Geographers. Boston. March. Sean DiGiovanna, and Yong Sook Lee, eds., Second Tier Markusen, Ann, and Mia Gray. 1999. "Colorado Springs: A Mil- Cities: Rapid Growth Outside the Metropole in Brazil, Korea, itary-anchored City in Transition." In Ann Markusen, Sean Japan and the U.S. Minneapolis: University of Minnesota DiGiovanna, and Yong Sook Lee, eds., Second Tier Cities: Press: 147-62. Rapid Growth Outside the Metropole in Brazil, Korea, Japan and Porter, Michael. 1990. The Competitive Advantage of Nations. the US. Minneapolis: University of Minnesota Press: 311-32. New York: Free Press. Markusen, Ann, and Sam Ock Park. 1993. "The State as Indus- Porter, Michael. 1996. "Competitive Advantage, Agglomeration trial Locator and District Builder: The Case of Changwon, Economies and Regional Policy: A Response to Ann South Korea." Economic Geography 69(2): 157-8 1. Markusen," InternationalRegional Science Review 19(1 & 2): Markusen, Ann, and Masayuki Sasaki. 1999. "Oita and Ku- 85-94. mamoto: Technopolises without Brains." In Ann Markusen, Saxenian, Annalee. 1994. RegionalAdvantage: Culture and Com- Sean DiGiovanna, and Yong Sook Lee, eds., Second Tier Cities: petition in Silicon Valley and Route 128. Cambridge, MA: Rapid Growth Outside the Metropole in Brazil, Korea, Japan and Harvard University Press. the U.S. Minneapolis: University of Minnesota Press: 223-38. Scott, Allen. 1988. "Flexible Production Systems and Regional Markusen, Ann, Peter Hall, and Amy Glasmeier. 1986. High Development: The Rise of New Industrial Space in North Tech America: The What, How, Where and Why of the Sunrise America and Western Europe." International Journal of Ur- Industries. Boston: Allen and Unwin. ban and Regional Research 12(2): 171-86. chapter 16 Globalization and Urbanization: The Case of Thailand Utis Kaothien National Economic and Social Development Board Royal Thai Government Douglas Webster National Economic and Social Development Board Royal Thai Government and Asia Pacific Research Center, Stanford University H hailand exemplifies a new form of ur- and the national economy (see Map 16.1). As a result banization that is occurring through- of government policy, as well as cooperation between in- out East and Southeast Asia. Although ternational lenders and multinationals investing in Bangkok-with some 8 million peo- Thailand, the center of manufacturing has shifted to the plC-still reigns as the primate city, three Eastern Seaboard provinces of Chonburi, Chacho- most new urbanization is occurring in engsao, and Rayong. In fact, this periurban area is emerg- areas extending beyond traditional municipal boundaries. ing as the manufacturing center of Southeast Asia, not This periurbanization is driven by manufacturing, which just Thailand. One can argue that the Eastern Seaboard, is increasingly located in modern industrial estates, with with 2.5 million people, is the second-largest city. As a residential development and commercial facilities such as result, Thailand's urban system is becoming less primate megamalls following, often in a dependent and non- nationally. planned fashion. Although periurbanization has been an important This new form of urbanization differs from the edge factor in gross domestic product (GDP) and employ- cities found in western countries, particularly the ment growth and declining poverty levels, it has gener- United States, in that periurban areas generally have ated social costs in terms of low environmental quality minimal business services and thus lack office (particu- and inadequate services.' For example, while the num- larly regional headquarters) and research and develop- ber of people living in slums in Bangkok fell during the ment parks. Meanwhile, growth rates are slowing in core 1990s, slums are proliferating throughout the industrial Bangkok, which is based on tertiary-or knowledge- heartland to the north and east of core Bangkok, along activities. Much of the deconcentration is occurring roads, rivers, and around industrial estates. within 200 kilometers of core Bangkok, driven by mo- Periurbanization also means that essentially rural-ori- torization and changes in communications and trans- ented, low-capacity governments are attempting to deal portation infrastructure, production processes, housing with the cutting edge of globalization, while more so- preferences (people's desire for more space at an afford- phisticated local governments manage slower-growing able price), and government location incentives (which core areas. Urgent measures must be taken to shape and are being withdrawn consistent with World Trade Or- better manage periurban development while Bangkok ganization regulations). itself builds on its attempts to become a knowledge However, growth around the intermediate-sized cities economy supporting a largely middle-class population. of the Northeast-the poorest region of Thailand- Thailand will especially need to ensure the competitive- now outpaces even that in the extended Bangkok region. ness of its extended urban economy as lower-cost pro- This growth is fueled by rural-urban migration and the ducers such as China and Indonesia emerge. Although increasing integration of the southern portions of the the private sector will be the prime investor in restruc- northeast corridor into the core Eastern Seaboard Area turing the economy of the Bangkok region and boost- I40 GLOBALIZATION AND URBANIZATION: THE CASE OF THAILAND I41 Map 16.1 Dynamic Urban Regions: Thailand Twin Cities e LAMPHUN t )LAMPANG < X 9V9~~~~~~~~~UDON THANI | \ ~~~~~~~KHON KAENA Northeaist 07 / / ~~~~~~~~~Corridor ~ < ~~~~NAKHON /O j ~~~~~RATCHASIMA _ a ~ ~~BANGO MPotS "" Eastern Seaboard " ?[ ~~(ESB) PHUKETW ;70_ ,X,HAT YAI/SONGKHLA Source: Urban Development Co-ordination Division, National Economic and Social Development Board, 1998 142 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION ing its position among Southeast Asian cities, the role experiencing strong development pressures such as of the public sector in strategically shaping the extended the Eastern Seaboard, and in the Hat Yai extended urban region will continue to be essential. urban region in the South (see Webster 1999). Al- though the new law does not provide for multiprovince Building Regional Coordination SLAOs, it could be amended to create SLAOs covering To strengthen the ability of local governments to deliver dynamic multiprovince regions such as the Eastern infrastructure and services, the Thai government has Seaboard. made decentralization a priority. However, the problem Addressing Quality of Life in Extended in Thailand is a plethora of local governments. The ex- Ur tended Bangkok region contains close to 1,000 local ban Regions governments, and even intermediate-sized urban regions Slum communities are only a rough surrogate for have 10 or more. Thus, decentralization could enable poverty in many emerging economies, as some house- local areas to tailor development strategies to global op- holds prefer to invest in small businesses and their chil- portunities, or it could result in lost opportunities if ex- dren's education rather than higher-quality housing. tended urban regions lack the capacity and resourtes to In fact, in Bangkok, one-third of slum renters-usually support new economic activities. considered the poorest residents-have household in- What is needed are new forms of regional coordina- comes higher than the mean (Pornchokchai 1992). tion, especially in larger extended urban regions such as What's more, the percentage of people living in the Eastern Seaboard. A variety of possibilities to address poverty, the number of people living in slums, and the this challenge exist, from regional government to re- number of slum communities in the BMA declined dur- gional development corporations to voluntary associa- ing the 1980s and 1990s (although the recession has tions of local governments. In the case of Thailand, pure reversed these gains somewhat).2 These changes reflect regional government and regional development corpo- reduced migration to the area as well as upward socio- rations are unlikely to be politically feasible or locally economic mobility supported by the buoyant economic acceptable. However, in November 1999, the King signed conditions from 1984 to 1996. the National Decentralization Committee law, which Nevertheless, urban slum communities do reveal makes possible the creation of special local administra- Thailand's failure to deliver acceptable living standards tion organizations (SLAOs) modeled on the Bangkok to substantial numbers of citizens-approximately 20 Metropolitan Administration (BMA). SLAOS would percent of the population of the extended Bangkok re- take responsibility for regional-scale governance, with gion. And slums are dispersing to suburban and exurban local councils operating below the SLAO level to pro- areas, a process that is not receiving enough attention. vide meaningful local participation, as is currently the These slums are prevalent along expressways, rivers, case in the BMA. Thus, services that are best planned or roads, and near industrial estates. For example, in the delivered at a regional scale, such as treatment of waste- Rayong urban area, as much as 50 percent of the popu- water and solid and toxic waste, transportation, and de- lation may be "unregistered"-mostly living in slums- velopment, could be coordinated by one regional-scale creating significant demand for services while providing government. little local government revenue.3 The proliferation of SLAOs would oversee urban regions that are eco- local governments has not helped the situation.4 nomically and politically more developed, often where a Unfortunately, fewer state agencies, nongovernmental special economic zone exercises considerable influence organizations (NGOs), and community-based groups are on developmental dynamics. SLAOs coterminous with available to address this challenge in newly urbanizing an existing province would exercise approximately the areas. NGOs, government agencies, and international de- same wide range of powers and have the same revenue- velopment agencies must recognize that the geography of generating opportunities as the BMA will have under the poverty is shifting and change their priorities accordingly new decentralization law. SLAOs smaller than a province to help communities deal with this challenge. In particu- would hold the powers of a senior municipality. lar, government grants based on registered rather than ac- The first SLAO will probably be implemented tual population hamper the ability of fast-growing com- in Phuket, and the second in the Nakhon Ratchasima munities to address the needs of the poor. Furthermore, (see Map 16.1), which contains Thailand's second- the high capital costs for social services and infrastructure largest city (excluding the Eastern Seaboard), with ap- associated with fast-growing urban populations should proximately 1 million people in the extended urban be recognized in dynamic areas such as the Northeast region. If successful, the concept can be applied in areas Corridor and the Eastern Seaboard. GLOBALIZATION AND URBANIZATION: THE CASE OF THAILAND 143 To address the problem of urban poverty, the Na- and disposal, and air pollution. Urban residents are also tional Economic and Social Development Board and demanding more public space, particularly green, in the the National Housing Authority established the Urban city, including access to the river. Community Development Office. This agency, which The national government was relatively slow in iden- began operating in 1992 with startup capital of US$53 tifying the key role that quality of the urban environ- million, offers residents of poor urban areas microcredit ment plays in attracting corporate headquarters, for livelihood improvement-such as vocational or tourism, and producer services. However, high-profile small business training, or loans to establish or upgrade reports such as the 1996 Urban Environment Strategy a small business-and for housing and community im- highlighted the importance of amenities such as public provement, including security of tenure. To be eligible, space and the need to address brown issues. Although a low-income community must first establish a commu- considerable progress has been made, further alleviating nity savings group, which lends money to households or traffic congestion-which accounts for a loss of at least groups within the slum community. The agency focused 2.1 percent of gross regional product in the Bangkok its first activities in Klong Tooey, Bangkok's largest slum, area alone-is an obvious step. Although Bangkok still adjacent to the port, but its activities are spreading trails some competitor cities in environmental quality, throughout the extended Bangkok region and beyond public commitment to improving the urban environ- to smaller cities, increasingly in the rapidly urbanizing ment-as well as developing human resources-is now but poor Northeast. evident. For example, in the most recent Asia Week poll What is lacking in Thai urban areas are public- of Asia's best cities, Bangkok climbed from a 26th-place private partnerships that develop and operate large- ranking in 1998 to 13th in 1999 despite (or perhaps be- scale projects for the public good (see Kaothien and cause of) economic constraints imposed by the eco- Webster 1996). Hopefully, public-sector reform, such nomic recession. as legal change enabling private enterprises to hold more than 50 percent equity in public-private partnerships, Developing Human Resources in Extended and allowing public servants to openly negotiate with Urban Regions the private sector, will make such partnerships more at- The Royal Thai Government has made considerable tractive to the private sector. Even more importantly, progress in improving enrollment rates at both primary governments need to establish a track record of deliver- and secondary school levels through stay-in-school pro- ing on plans and honoring commitments, thus making grams that include student loans. (Many children from the private sector more willing to engage in joint ven- poor families do not attend or drop out of secondary tures. For its part, the private sector will need to better school because of lost income, plus the costs of modest recognize emerging opportunities and community needs school fees, books, school clothes, and transportation.) and act prudently; its past performance in overbuilding The challenge is to develop post-secondary technical property in inappropriate places indicates that it could education that will dovetail with existing and future em- benefit from paying closer attention to government de- ployment needs in Thailand's diverse urban regions. Be- velopment plans and socioeconomic data. cause of a shortage of qualified instructors-over 30 As the global economy demands new forms of urban percent of post-secondary faculty positions are vacant infrastructure, such as convention districts with high- outside Bangkok-this will require more emphasis on level amenities, public-private cooperation will become cutting-edge telecommunications infrastructure to allow not an option but a necessity. Until recently, decision- distance learning. makers in Thailand have consistently underestimated Distance learning can also help improve education in the role that such amenities play in attracting high-value rural areas, indirectly benefiting urban regions through activity to urban regions. Numerous surveys of Japan- rural-urban linkages and migration. Today there are ese, American, and European corporate leaders have enormous differences in educational attainment be- shown the value they place on amenities in deciding tween rural and urban areas: Some 33 percent of urban where to locate production and knowledge-based activi- residents have attended secondary school, while only 11 ties (Webster 1992). Those that control capital are de- percent in rural areas have done so; 27 percent of urban manding even more amenities for high-level corporate, residents have some post-secondary education versus 6 service, and production activities, and Thailand's urban percent in rural areas. citizenry is increasingly demanding that government Considerable promise also lies in public-private part- confront traffic congestion and "brown issues"-waste- nerships that involve corporate leaders in operating water treatment, garbage pickup, solid waste processing state-of-the-art training facilities. General Motors and I44 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Honda, as part of their decisions to locate or expand pro- Figure 16.1 duction for Southeast Asian markets in Thailand, are de- Oversupply to Remain for a Long Time veloping automotive technical colleges in the periurban area. Similarly, UNOCAL is operating a training facility Units (thousands) Excess Supply 1%) in partnership with the Department of Mineral Re- 400 - - 1,400 sources in Songkhla. Governments must maintain regu- 1,200 - - 1,200 latory control over such human development facilities, but they do represent an effective means for technology 1,000 - - 1,000 transfer, and almost invariably provide a springboard to 800 - - 800 good employment after graduation. An education sys- tem that produces entrepreneurs and managers as well as 600 - - 600 technical personnel will set the stage for more Thai own- ership of firms in industries such as automobiles, petro- 400 - - 400 chemicals, tourism, and agribusiness. A current weakness 200 - - 2 in Thailand's development is that small manufacturing 200- -200 enterprises are not developing rapidly enough to supply o I I - o lead firms in periurban industrial clusters such as auto- 1992 1993 1994 1995 1996 1997 1998E mobiles and electronics. Forecasted real = Actual houses +- Excess Harnessing Global Capital Flows housing demand constructed supply (%) When Thailand suffered a severe recession in the early Note: Forecasted real housing demand contrasted to actual housing 1980s, the national government decided to implement construction in the extended Bangkok region from 1992 to 1998. Excess supply peaked in 1996. an export-led development policy propelled by foreign Source: Government Housing Bank, National Housing Authority, Real Estate direct investment (FDI). This policy was more success- Sub-committee. ful than even its strongest supporters expected. Virtu- ally all the resulting FDI flowed into Thai urban areas. Inadequate regulation of capital flows, particularly port- biles made living or producing in more remote loca- folio investment, clearly underlay the urban bubble that tions possible. burst in 1997. The Thai Government also recognized a decade ago FDI was dominantly in manufacturing early on but it that telecommunications would become vital in inter- shifted toward real estate by the latter 1990s. The impacts national competition, and partially deregulated and pri- of this shift-as well as a 19-fold increase in commercial vatized the telecommunications industry. Private opera- lending from foreign banks between 1991 and 1997, re- tors installed wideband fiber optics networks in many sulting from introduction of the Bangkok International urban areas and in corridors connecting major urban Banking Facility in 1991-on the property market of the areas, often using railway rights-of-way.5 Ironically, extended Bangkok region were enormous. However, the middle-income developing countries such as Thailand, high levels of investment did enhance Bangkok's role as particularly their urban areas, often leapfrog technolog- a financial center and diversified its rapidly growing ser- ically. In this case, backlogs in orders for fixed telephone vice economy. The result was a property bubble: Approx- lines have encouraged rapid acceptance of new tech- imately 340,000 housing units remain vacant in the ex- nologies such as mobile phones. Today virtually all set- tended Bangkok region (see Figure 16.1). Although tled areas possess state-of-the-art wireless systems, often overbuilding can make middle-class housing and of- more advanced than those in Europe.6 fice and industrial space more affordable, it encourages urban sprawl, misallocates capital better used to fund cusing on Software instead of Hardware employment-creating enterprises, and results in "ghost" Recent Thai urban development has focused largely on buildings throughout the urban landscape. "hardware"-or physical systems. In cases where full or A secondary but significant use of available financ- nearly full cost recovery is possible, such as urban ex- ing was for purchase of consumer durable goods, pressway projects serving core urban areas are operating mainly automobiles. Thus, the opening up of Thailand effectively, but new expressways into lesser-developed to capital flows and motorization were powerful forces areas are having difficulties achieving full cost recovery. in shaping Thailand's urban region, particularly the ex- However, in other cases, such as wastewater treatment tended Bangkok region. Inexpensive loans for automo- and solid waste disposal, expensive facilities are nor op- GLOBALIZATION AND URBANIZATION: THE CASE OF THAILAND I45 erating well because local operating funds or technical capital-especially by producing and widely disseminat- skills are not available, and because appropriate policy ing information on flows and uses of capital. For exam- frameworks, including incentive structures, are not in ple, the government can educate the banking industry place. For example, Bangkok's wastewater treatment sys- on the need to consider the region's supply and demand tem is operating at 20 percent of capacity because there in making loans for urban development projects. Even is no regulation, or incentive, for households or firms to at the microlevel of loan applications, future income hook up (see Webster 2000). This problem is exacer- flows based on business plans and market analysis bated by international companies that promote inappro- should take precedence over collateral-based loans. priate technology such as costly solid waste incinerators, All these efforts could result in more effective use of multilateral and bilateral development agencies that capital in structuring and shaping urban systems. To a favor loans for such hardware, and central and local gov- large extent, urban regions rather than nations will be ernment politicians and bureaucrats who stand to gain competing. They cannot afford to grossly misallocate through large-scale capital projects. scarce capital in either the private or the public sector. Managing demand in vehicle use and consumption of electricity and water must become a priority. Al- The Bangkok Plan though demand management will obviously not negate The case of the extended Bangkok region demonstrates the need for incremental infrastructure, it will dampen how positive outcomes and additional challenges may the need for costly new hardware. The fact is that the arise from such policy approaches. In 1995 the Ministry price of many conventional technologies, such as waste- of the Interior gave the Bangkok Metropolitan Admin- water systems, which cost as much as US$1,000 per istration permission to prepare its own urban plan. This household, or urban incinerators, which usually cost at was an important precedent in Thailand, as it allowed a least US$50 million, are disproportionately high rela- local authority to establish priorities and a vision of the tive to incomes in Thai urban areas. Thus, capital costs future. Until this point all urban plans in Thailand had alone make cost recovery difficult, and in times of eco- been produced centrally. nomic recession subsidies are not available to keep these The rapid development of the Bangkok financial sec- systems operating. tor reflected a wider trend toward a knowledge econ- Decentralization should reduce this problem, as local omy. The National Economic and Social Development authorities may be less likely to invest in facilities that Board and the Bangkok Metropolitan Administration they cannot maintain. Policymakers, politicians, and therefore promoted Bangkok as a "knowledge city"-a managers responsible for Thai urban regions must de- characterization that increasingly came to reflect reality. termine whether there are more cost-effective means of Governor Krisda Arunvongse of the BMA commis- delivering urban infrastructure, such as through demand sioned the Massachusetts Institute of Technology, in management, recycling, more informed choice of tech- conjunction with BMA's city planning department, to nology, market-based incentives, or user fees. Globaliza- develop a physical plan based on a vision of Bangkok as tion forces will demand better urban management; a knowledge-based economy playing a key role in given real resource constraints, this will force a move Southeast Asia-and even the world. The ensuing ad- away from a simplistic focus on urban hardware to a ministration of Governor Bichit Rattakul augmented more challenging focus on urban software. this plan with local district-level initiatives and strong Thailand's overall goal should be to aim for higher- measures to make the city more livable and green. value-added urban activities, employment based on For example, hundreds of thousands of trees have re- skilled human resources, and amenities such as access to cently been planted in the city and sidewalks have been a full range of social and recreational services. All these upgraded with interlocking bricks. Measures are being activities need an urban framework: Comprehensive taken to limit growth in flood-prone areas on the pe- planning may be dead for good reason, but strategic riphery of the city. In 1997 Bangkok initiated a city- planning/management and coordinated programming wide antilittering campaign based on extensive public at local and regional levels are essential. awareness, including posters, street booths, and airport The reality is that today government has limited advertisements, as well as strict enforcement, with vio- power to control capital flows; the notion that govern- lations incurring 2,000-Baht fines. The program has ments in guided capitalist systems in East and Southeast been a success: Bangkok streets are litter free. Other ef- Asia can pick winners and allocate capital accordingly forts to make BMA more attractive to global knowledge has largely been abandoned. The government's role will industries include improved sidewalks, more landscap- instead be to inform, advise, and regulate key users of ing, and new "pocket" parks. In fact, the Bangkok ad- 146 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION ministration justifies the initiatives as both improving residents would benefit from such quality-of-life initia- the quality of life for local residents and positioning tives as well as the resulting income, as employment and Bangkok as a leading knowledge city.7 educational levels would rise along with the influx of The Bangkok Transit System (BTS) is a direct re- high-value businesses. sponse to the rapid growth of the knowledge economy, Rama III, the new business district, is emerging to the particularly the financial sector. Bangkok has no one dis- east of the present Silom area, known as the Wall Street tinct downtown but rather several key areas specializing of Thailand. This new district is almost completely the in different kinds of business such as media, finance, product of corporate interests such as Esso Petroleum fashion, tourism, international governance, and diplo- and local property developers, who responded to a new macy. The 24-kilometer elevated BTS, which opened in road into the area built by BMA. As more multinational December 1999, connects these areas, with the excep- and national corporations are attracted to the area, the tion of the new Rama III business area. The system was BMA has belatedly tried to shape this new business dis- driven and justified by the fact that office-based employ- trict and improve access to it by proposing a light-rail ment and tourism generated traffic beyond road capac- connection. ity. To increase densities demanded by the need for face- However, the business community is wary of govern- to-face contact in these industries, the government was ment planning and management of urban infrastruc- forced to build a mass-transit system linking the city's ture-the chief complaint being that the government key nodes. often does not deliver on its commitments, leaving pri- The BTS system, opened in November 1999, with a vate investments stranded. Private entities are also reluc- capacity of 700,000 riders per day, is attracting world- tant to become involved in municipal ventures they wide attention because it aims to recover its costs fully- would not control, as regulations prohibit more than 49 a rarity in urban rail systems. This is considered possi- percent private ownership. Nevertheless, business com- ble because time-savings afforded by the system will be munities increasingly drive local priorities. For example, among the highest in the world. However, full cost re- businesses are promoting high-quality industrial estates covery might not be feasible in the near-term because with cutting-edge infrastructure including social facili- vehicle traffic is growing more slowly than before the re- ties and programs, as well as vocational training focus- cession, ;several new toll freeways have recently opened, ing on the region's comparative advantages. and a portion of the loans is denominated in U.S. dol- lars, while revenue will be in Baht. Nevertheless, the Shaping the Future project will undoubtedly be highly economical. As the dominant city-region in a country that is one of The missing link is commuter rail. As the urbanized the most open in East Asia, Bangkok has seen its econ- area grows far faster than population, the failure to de- omy and society change rapidly over the past 15 years. velop commuter-rail systems becomes more serious. Rail The resulting demands, such as improved job opportu- systems would economize on energy and human time and nities for households, more amenities and higher envi- provide urban structure by stimulating higher densities ronmental quality, regional-scale infrastructure delivery around stations. The State Railway of Thailand is now in periurban areas, and faster transportation in the core, taking action to revive a 60-kilometer elevated commuter have required numerous changes in governance, service rail project to serve the northern and eastern suburbs. delivery, and the mix of public-sector programming. Overall, the Bangkok metropolitan area, like the na- The BMA, suburban and periurban governments, as tional Thai government, is beginning to understand the well as the national government, have often been in a re- importance of amenities in attracting and retaining high- active position, running to catch up with these changes. value economic activity, but much remains to be done. However, much progress has been made, reflected in In many cases, small amounts of money would make a the rapid climb of Bangkok in city-quality ratings in significant difference. For example, several hundred kilo- Asia. For example, the SLAO form of organization could meters of elevated expressways and rapid-transit systems improve governance considerably in periurban areas, a crisscrossing the Bangkok area are greatly improving the rapid transit system and several new toll expressways are quality of life. However, landscaping of the vast concrete now operating, the city's streets have been cleaned and structures has been very limited, and the chaos created trees planted, low-income households have access to mi- underneath these behemoths, resulting from population crocredit to improve their income capacities, and tech- relocation and construction, is not being adequately ad- nical education is being improved to better match the dressed despite the fact that doing so would represent emerging knowledge-based economy in the Bangkok only a trivial percentage of the cost of the projects. Local area. The challenge now facing those responsible for the GLOBALIZATION AND URBANIZATION: THE CASE OF THAILAND 147 Bangkok region's future-and with decentralization and Pornchokchai, Sophon. 1992. Bangkok Slums: Review and Rec- the growth of civil society, there are many stakeholders- ommendations. Bangkok: Agency for Real Estate Affairs. is to move policy and programming to a more proactive Rohlen, Thomas. 1995. A Mediterranean ModelforAsian Region- alism: Cosmopolitan Cities and Nation States in Asia. Asia Pa- stance. It IS to be hoped that, within a decade, local gov- cific Research Center, Stanford: Stanford University, Mono- ernments in the extended Bangkok region will be shap- graph (The Urban Dynamics of East Asia Series). ing socioeconomic and physical development, not just United Nations Conference on Trade and Development (UNC- responding to the strong and often beneficial but rapidly TAD). 1996. World Investment. New York and Geneva: changing forces of globalization. United Nations. Webster, Douglas. 1992. "The Role of Amenity in Canadian Re- Bibliography gional Development." Plan Canada. July. Webster, Douglas. 1998. "Infrastructure Development Policies Bangkok Metropolitan Administration (BMA). Various years. and Economic Development: A Perspective from Thailand." Statistical Profile of BMA. Bangkok: BMA, Department of Paper presented to Workshop on ASEAN Infrastructure Policy and Planning, published annually. Planning and Management, Bangkok, September 30. Bank of Thailand. Various issues. Monthly Bulletin. Bangkok: Webster, Douglas. 1999. "Strengthening of Area Based Devel- Bank of Thailand, published monthly. opment in Thailand." In Proceedings of Policy Workshop on Dowall, David. 1992. "A Second Look at the Bangkok Land and Area Development and Rural-Urban Transition, Bangkok: Housing Market." Urban Studies 29(1): 25-38. National Economic and Social Development Board. Garreau, Joel. 1988. Edge City: Life on the New Frontier. New Webster, Douglas. 2000. Financing City Building: The Bangkok York: Doubleday. Case. Asia Pacific Research Center, Stanford: Stanford Univer- Graham, Stephen, and Simon Marvin. 1996. Telecommunica- sity, Monograph (The Urban Dynamics of East Asia Series). tions and the City: Electronic Spaces, Urban Places. London: World Bank. Various issues. Thailand Economic Monitor. Wash- Routledge. ington, D.C.: The World Bank, published quarterly from Kaothien, Utis, and Douglas Webster. 1996. "Public-Private October 1998. Cooperation in Urban Development. "Proceedings of the 10th AAPH Biennial Convention on the Impact of Globalization on Not Urban Development in the ASEAN Region. Bangkok: National Otes Housing Authority. 1. Numerous high-profile environmental problems have been Kaothien, Utis, and Douglas Webster. 2000. "The Bangkok Re- documented in the Extended Bangkok Area. For example, a gion." In G. Hack and R. Simmonds, eds., Global City Re- school in Map Ta Phut had to be moved because of severe air gions. London: Routledge. pollution from a nearby petrochemical complex. Kaothien, Utis, Douglas Webster, and John Lukens. 1997. "In- 2. Because BMA's population grew substantially during the frastructure Investment in the Bangkok Region." In Proceed- 1980s and 1990s, the percentage of people living in poverty and ings of Conference on The Infrastructure Instrument: How to slums declined much faster than the absolute numbers. See Sta- Use It to Manage the Function and Form of Global City Regions. tistical Profile of BMA (published annually). Global City Regions Project, Madrid. Madrid: Universidad 3. Personal communication, mayor of Rayong, September Complutense. 1998. Kaothien, Utis, Douglas Webster, and Voravit Vorathanyakit. 4. Introduction of SLAOs, discussed previously, would solve 1996. "Country Report of Thailand." In Royston Brockman the government proliferation problem. and Allen Williams, eds., Urban Infrastructure Finance. Ma- 5. Because deregulation often resulted in regional monopo- nila: Asian Development Bank. lies, the price of international telecommunications has remained National Economic and Social Development Board (NESDB)/ high relative to competitor cities such as Singapore, but the Baht Leman Group. 1995. Chao Phraya Multipolis Structure Plan devaluation has brought the cost of Thai-sourced international (Metropolitan Region Structure Study). Bangkok: NESDB. telecommunications close to international prices. For details, see National Economic and Social Development Board (NESDB)/ Kaothien, Webster, and Lukens 1997, and Kaothien and Web- Danish Cooperation for Environment and Development. ster 2000. 1996. Urban Environmental Management in Thailand: A 6. For a description and analysis of the acceptance of new Strategic Planning Process. Bangkok: NESDB. technologies in Thailand, see Kaothien and Webster 2000, and National Economic and Social Development Board (NESDB)/ Kaothien, Webster, and Lukens 1997. Norconsult. 1997. A Spatial Development Strategy for Thai- 7. Personal communication, Deputy Governor Banasopit land, Bangkok. Bangkok: NESDB/Norconsult. Mekvichai, November 20, 1998. chapter 17 The Impact of Decentralization on Cities' Fiscal Health: Lessons from the United States Andrew Reschovsky Robert M. La Follette Institute of Public Affairs Department of Agricultural and Applied Economics University of Wisconsin-Madison N hroughout the world, in both devel- the nonpoor to escape fiscal responsibilities for the poor oped and developing countries, re- by moving to suburban communities where the poor are sponsibilities for the provision of often effectively excluded through the use of zoning or- public services are being shifted to dinances and the existence of housing market discrimi- subnational governments.' Examples nation. The fiscal health of central cities can then be fur- of decentralization can be found in a ther weakened because the out-migration from the city number of developing countries, for example, in South of both businesses and moderate- and high-income fam- Africa and Kenya in Africa, and in China and India in ilies creates fiscal externalities by further weakening the Asia. Even in the United States, which has long been fiscal capacity of the central city and raising the average highly decentralized, recently passed legislation has cost of providing public services. shifted primary responsibility for cash assistance for the The governance of most American metropolitan areas poor (welfare) from the federal government to the states. is highly fragmented. For example, the Chicago metropol- Decisions to decentralize are usually motivated by a itan area contains 262 separate general-purpose govern- desire to improve the effectiveness of governments at all ments. When one counts all school districts and special levels. Transferring the responsibility for both the provi- districts, there are nearly 1,200 different governmental sion and financing of public services from central gov- bodies within the metropolitan area. The existence of a ernments to subnational governments means that deci- large number of governments is not necessarily bad. A sions about the mix and the level of service will more large choice among governments enhances consumer closely reflect the preferences and tastes of the residents well-being by matching public good preferences and will- of each local community. As local governments are pre- ingness to pay. Furthermore, competition among govern- sumably more in touch with the desires of their con- ments may force local governments to operate more effi- stituents, they should be in a better position than ciently.2 Fiscal externalities, however, may lead to central higher-level governments to develop innovative and city fiscal problems when local revenue sources provide a cost-effective methods for providing public services. substantial portion of the financing of public services pro- The strength of a decentralized fiscal system must be vided by local governments, and when local governments tempered, however, by the realization that when urban are ultimately responsible for providing services for large areas are divided into a number of fiscally independent concentrations of the poor and otherwise needy. local governments, each local government has an incen- In this chapter, I argue that many U.S. cities are in a tive to exclude those individuals who require extra ex- structurally weak fiscal position, and their long-run fis- penditures in excess of their marginal contributions to cal health may be further weakened by the recent devo- locally raised revenues. If, as in the United States, the lution of responsibilities for welfare to the states. The poor tend to be concentrated in the central city and in U.S. situation provides a cautionary tale for countries a few older suburbs, there exists a strong incentive for throughout the world that are in the process of decen- 148 THE IMPACT OF DECENTRALIZATION ON CITIES' FISCAL HEALTH: LESSONS FROM THE UNITED STATES 149 tralization. By understanding the structural fiscal prob- erally referred to as fiscal disparities. The existence of fis- lems faced by many central cities in the United States, cal disparities has both equity and efficiency implica- it is hoped that other countries will be able to avoid tions. Fiscal disparities result in horizontal inequities some of the hardships and distortions that result from among metropolitan area residents. These inequities the fiscal system that has evolved within metropolitan occur when residents of two metropolitan area commu- areas in the United States. nities face identical tax rates but receive different levels of public services. Alternatively, inequities exist when Measuring the Fiscal Condition residents of communities that provide similar levels of of Local Governments public services face different tax rates. In State of the Cities in 1998, the U.S. Department of The claim is sometimes made that these inequities are Housing and Urban Development (1998) declared that short-run phenomena that will be ameliorated by market "driven by a robust national economy, cities are fiscally forces as individuals living in communities in weak fiscal and economically the strongest they've been in a decade." condition are motivated to move to places in stronger fis- Yet despite this rosy assessment, many U.S. cities con- cal health.4 The existence of zoning and other land-use tinue to struggle. Even though their residents face rela- control mechanisms in the United States, however, places tively heavy tax burdens, many provide inadequate levels severe constraints on the residential mobility of low- of public services-police are not available, trash is col- income households and helps explain why metropolitan lected infrequently, and students perform poorly.3 area fiscal disparities are not self-correcting. The use of These problems stem largely from the fact that decen- zoning regulations allows suburban communities to ef- tralization is burdening many cities with growing service fectively set a minimum price (and rent) for housing responsibilities even as federal assistance is declining. within their boundaries, thereby providing an effective While the tax base (or more broadly, the revenue-raising way to exclude low-income households. There is also con- capacity) of most cities is growing, the rate of growth is siderable evidence that racial discrimination in the hous- generally less rapid than in surrounding communities ing and rental markets is widespread (Yinger 1997). with whom cities must compete for both residents and These discriminatory practices make it more difficult for business. minority residents to move out of central cities in order It is helpful to summarize these developments by using to find housing in communities in better fiscal health a single measure to represent and compare the structural than the central city. fiscal condition of local governments. Drawing on the To the extent that individuals and businesses make work of Bradbury et al. (1984), and Ladd and Yinger locational decisions within metropolitan areas based on (1991), a need-capacity gap can be calculated for each fiscal considerations, a pattern of inefficient location de- local government. It is defined as the gap between ex- cisions is likely to occur.5 By encouraging suburbaniza- penditure need and the revenue-raising capacity of each tion, fiscal considerations may result in a pattern of busi- local government. Expenditure need indicates the mini- ness and residential locations that increases metropolitan mum amount of money a government must spend per area congestion and environmental degradation. To the resident in order to provide a standard or average level extent that high-income residents and businesses are of public services for which it is responsible. Revenue- most sensitive to fiscal conditions, their out-migration raising capacity indicates the amount of revenue per from central cities will have the effect of further exacer- resident a local government has available if its residents bating the fiscal health of the city. These fiscal externali- face a standard or average tax burden. Revenue-raising ties are likely to occur because private decisions to leave capacities can be enhanced by cities' receipt of grants the city will not only reduce the revenue-raising capacity from higher-level governments. It is important to em- of the city, but by changing the mix of residents remain- phasize that this measure of the fiscal condition of local ing in the city, may result in increases in the average per government focuses on factors that are generally outside capita cost of providing public services. For example, be- the control of local government officials. In this way, cause of "peer-group" effects in education, the departure need-capacity gaps provide an objective measure of the of middle-class children from central city schools is likely structural fiscal problems faced by local governments. to raise the costs of educating those children who remain Fiscal Disparities within Metropolitan Areas (Henderson, Mieszkowski, and Sauvageau 1978). Flscal~ ~~~~~~~~~Resn foranle Weaki Fiscaita Health Within metropolitan areas, differences in the fiscal con- Reasons for Weak Fiscal Health ditions of local governments, whether measured by In the following paragraphs I consider a number of rea- need-capacity gaps or by alternative measures, are gen- sons why many American central cities are in weak fis- 150 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION cal health relative to their suburbs. They can be grouped is that the central cities' share of metropolitan-area em- into the following areas: (1) relatively low revenue- ployment fell in 82 percent of the metropolitan areas raising capacities in many cities; (2) growing service re- studied. sponsibilities; (3) higher uncontrollable costs in cities Not only the level, but the structure of employment relative to their suburbs; and (4) policies of higher-level in cities is changing. Cities are losing manufacturing governments. jobs while gaining some white-collar employment in business services, finance, insurance, and real estate. The holders of high-paying city jobs in these sectors often Urban economists have argued that rising incomes and prefer to live in the suburbs and commute to work. declining transportation and communication costs in- Cities' ability to capture a share of the higher wages paid duce both individuals and businesses to move away from by new high-productivity industries therefore depends the city center and toward outlying areas where land is on whether they can tax the income of nonresidents. generally less expensive. Of the 24 U.S. cities with pop- Yet doing so is difficult because city governments that ulations over one-half million, 8 had smaller popula- want to levy an income or sales tax must seek authoriza- tions in 1994 than in 1980. The population of these 24 tion from state government. And suburban-dominated central cities grew an average of 6 percent during those legislatures often refuse to allow cities to expand their years, while the population of their suburbs grew 25 tax base because such a move would mean higher taxes percent annually-over four times faster. for suburban residents. Only 8 of the nation's 24 largest Of the 10 most populous cities in the Northeast and cities impose an income or wage tax, and those 8 cities Midwest, only one, Columbus, Ohio, experienced a tax income earned by nonresidents at a very low rate or population increase between 1970 and 1990. Population not at all. Even if a city does succeed in taxing earnings, in the other 9 cities fell an average of 17 percent. The businesses may move to the suburbs to avoid paying a population loss was mostly caused by out-migration of wage premium to attract workers. And if a city imposes high- and middle-income families and individuals; in all a sales tax, the higher the tax the smaller the chance that 9 cities the ratio of per capita income in the city to per suburban residents will choose to shop in the city. capita income in the suburbs fell between 1960 and While central cities may no longer dominate their re- 1989. gions' economy as they did in the past, they continue to The out-migration of middle-income city residents serve as their regions' cultural and entertainment cen- tends to reduce the capacity of the city government to ters. Although, cities' museums, concert halls, and sports raise revenue, especially when people with lower in- facilities continue to be popular with residents, subur- comes replace departing residents, or no one replaces banites, and tourists, the fiscal benefit to cities of these them at all. To compensate, the city must increase tax facilities is often limited. To the extent that these facili- rates or reduce public spending-further convincing ties are owned by governments or nonprofit organiza- middle-class residents to leave. tions, they are exempt from property taxation. As a re- Regardless of what taxes city governments use, con- sult, cities that rely heavily on the property tax get centrations of poor people reduce their capacity to raise limited fiscal benefit from their cultural, educational, funds. Yet reducing spending on direct services to the and sports facilities. In general, tax-exempt property is poor in response may prove self-defeating. For example, concentrated in central cities. For example, in New York while eliminating youth recreation or summer job pro- City nearly one-third of property value is exempt from grams may save money in the short run, cities may have taxation, while 13 percent of property value is exempt to spend more on public safety and gang control in the from taxation in suburban Nassau County and 22 per- long run. cent in Westchester County. The traditional role of central cities as centers of em- ployment is also diminishing. Brennan and Hill (1999) Broad Service Responsibilities studied private-sector job growth in 92 large metropoli- In the United States, local governments are responsible tan areas between 1993 and 1996-a period of rapid for providing a wide array of public services. In fact, economic growth in the United States. They found that many of the core services most people associate with 23 percent of the central cities in their sample lost em- governments are provided, and in most cases financed, ployment during this period, while their suburbs gained by local governments. Although the assignment of func- employment. The number of jobs increased in 52 per- tions differs across states, local governments generally cent of the central cities, but at a slower rate than in play the role of service provider of last resort, required by their suburbs. The result of these patterns of job growth state governments or by the courts to provide shelter to THE IMPACT OF DECENTRALIZATION ON CITIES' FISCAL HEALTH: LESSONS FROM THE UNITED STATES 151 the homeless and medical aid to the indigent. Policy bundle of public services. Economists refer to the mini- changes at higher levels of government often end up hav- mum amount of money that a government must spend ing fiscal implications for local governments. In practice, in order to provide any given level of public services as expanded public service responsibilities often come in the costs of public services. the form of mandates from both the federal government Although it certainly is not difficult to find examples and state governments. For example, the widespread de- where city government spending is inflated due to inef- institutionalization of the mentally ill that has occurred fective management, inefficient and outdated union over the past couple of decades in effect forced cities to work rules, and wasteful administrative structures, there deal with the mentally ill who ended up on the street, also exists strong evidence that central cities face above- became public nuisances, committed crimes, or needed average costs due to factors over which they have no medical care. control. On the basis of a number of econometric stud- Service responsibilities tend to be greater in central ies, it is possible to conclude that costs tend to be higher cities than in most suburban communities. As demon- in central cities for the following reasons.7 First, the strated by recent research by Anita Summers and her costs of achieving any given level of public safety or of colleagues at the University of Pennsylvania, one reason educating children to meet any given level of educa- why public expenditures tend to be high in central cities tional performance are generally higher in locations with relative to their suburbs is that city governments finance concentrations of low-income households. Not only is a number of direct services to poor persons, especially the incidence of crime higher in poor neighborhoods, in the areas of public welfare and public health.6 In but community attributes associated with poverty, such 1998, the average poverty rate in American central cities as high density and poor housing conditions, increase was 18.5 percent, a rate that is over twice as high as the the amount of resources required to provide public average suburban poverty rate (U.S. Bureau of the Cen- safety in these neighborhoods. Studies also suggest that sus 1999). Not only are the poor concentrated in cen- smaller class sizes, specially trained teachers, and extra tral cities, but many of the nation's social problems, classes are necessary to compensate for the social and problems that hardly existed 20 years ago, such as crack economic disadvantages faced by most children from cocaine, homelessness, and the AIDS epidemic, also poor families.8 tend to be spatially concentrated in central cities. Second, in many cases, cities have higher costs than The fiscal consequence for central-city governments their suburbs because their infrastructure is older, and that must provide special services to citizens with vari- consequently the costs of maintenance, and often of fire ous social and economic problems is that other city res- prevention, are higher. idents and businesses must either pay higher taxes or Third, there is evidence that, at least for some public contend with lower levels of basic public services, such services such as education, costs in large central cities as public safety and sanitation. The fact that the con- are above average because of diseconomies of scale at- centration of the poor within central cities results in tributed to their large size. broader service responsibilities for central-city govern- Finally, costs, measured on a per resident basis, tend ments relative to their suburbs serves to weaken the rel- to be higher in central cities relative to suburbs because ative fiscal condition of cities. cities must provide services for significant numbers of nonresidents, whether they be suburbanites commuting to central-city jobs or taking advantage of the city's cul- To the extent that fiscal considerations influence loca- tural, entertainment, and commercial attractions. In tional decisions, it is reasonable to imagine that both particular, nonresidents contribute to the costs of pub- businesses and individuals compare the level and mix of lic safety, sanitation, and cultural and recreation services public services the city provides and the taxes and fees provided by city governments. they must pay to receive these services. Available evi- To the extent that city governments need to spend dence suggests that the relationship between benefits re- more money than their suburban neighbors in order to ceived and expenses incurred is generally less favorable provide services for the poor and for nonresidents, there in central cities than in their suburbs. This central-city are fewer resources available for improving public ser- fiscal disadvantage may occur either because city govern- vice delivery for businesses and for the middle class. City ments operate inefficiently compared to the average sub- governments face the difficult task of having to either urban governments or because factors beyond city con- cut services or raise taxes, either of which may increase trol require that city governments spend more money the chances of out-migration by these relatively mobile than suburban governments in order to deliver the same groups. I52 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Policies of Higher-Level Governments eral government has increased the concentration of the One policy response to the relatively low fiscal capacity; poor in urban neighborhoods and also fostered immo- broad public service responsibilities, and high costs lityamongproject residents. This occurs because proj- faced by central-city governments would be for the fed- ect residents pay a rent that is restricted to 30 percent of eral and state governments to provide cities with direct their income, while many low-income urban residents grants. In principle, city governments' relatively weak not living in public housing projects must pay over half fiscal conditions could be improved by a lump-sum their monthly income as rent.el grant allocated in proportion to a city's fiscal condition, A number of features of the federal tax code also may with local governments in weaker fiscal positions receiv- have an implicit pro-suburban bias. The most impor- ing large pern capita grants. tant of these are the tax treatment of owner-occupied In contrast to many, if not most, other countries, the housing. In the United States, homeowners are entitled federal government in the United States provides its to deduct the value of their mortgage interest and prop- local governments with no general-purpose aid. Most erty tax payments from their incomes subject to taxa- federal funds that are allocated to local governments are tion. Given the progressive structure of the U.S. income in the form of categorical grants designated for specific tax system, this feature of the tax code provides more purposes, such as mass transit, education for disabled generous subsidies to middle- and upper-income fami- children, and community development. Since the late lies than to those with modest incomes. Ongoing theo- 1970s, financial assistance from the federal government retical and empirical work by Joseph Gyourko and Todd to city governments has declined sharply as a percentage Sinai suggests that deductibility is providing larger sub- of city government spending. The cuts in direct federal sidies to suburban than to central-city residents and may assistance have been particularly steep for the nation's be having the effect of encouraging suburban homeown- 23 largest cities. The share of expenditures financed by ership among middle-income households. 12 the federal government in those cities fell from 14 per- Continued Decentralization in the United States cent in fiscal year 1977 to under 7 percent in fiscal 1996 (the latest year for which such data are available). Although the United States is already one of the world's Grants-in-aid from state governments to their cities most decentralized countries, policy changes imple- have in aggregate remained quite stable over time. In a mented in the past few years have resulted in further de- few cases, state financial assistance to city governments centralization. In 1996, the Congress enacted the Per- has taken the form of direct state financing of services sonal Responsibility Act. This legislation, commonly previously funded by city governments. This is the case referred to as "welfare reform," devolved responsibility in New York City, where the state assumed responsibil- for the major program providing cash assistance to low- ity for most of the operating costs of higher education income families from the federal government to the and the courts in response to the fiscal crisis of the mid- states. Prior to 1996, families that met income and other 1970s. The role played by state financial assistance in eligibility criteria were entitled to a monthly cash grant city budgets varies considerably among the states, de- funded jointly by the federal and state governments, pending in part on the range of services provided by city with the amount of the grant determined by the legisla- governments.9 With the exception of Boston, Detroit, tures in each state. With the passage of the welfare re- and Milwaukee, however, state general-purpose aid to form legislation, this entitlement was eliminated, and cities is quite limited, accounting for only 3.5 percent of states became responsible for determining the level and city direct expenditures in the nation's 23 largest cities."0 the scope of cash transfers. Prior to 1996, the financing Perhaps even more important for the long-run fiscal of these transfers was in the form of a matching grant health of cities than direct federal aid are the unintended from the federal government to the states. The new leg- pro-suburban spatial biases in a number of federal poli- islation replaced the matching grant with a block grant cies. The most salient examples are found in transpor- to the states. tation, housing, and tax policies used by the federal Due both to the current strength of the American government. These policies tend to reinforce both the economy and to stringent new eligibility requirements market incentives that cause middle- and upper-income that limit the lifetime receipt of cash assistance to five residents to migrate to the suburbs and the growing con- years and require most welfare-eligible adults to seek centration of the poor in the central cities (McGeary, work, the number of welfare cases has fallen dramati- 1990; Chernick and Reschovsky, 1997). For example, cally.13 At least in the short run, the decline in caseloads by building most large low-income public housing proj- combined with the switch to a system of block grants ects in central-city low-income neighborhoods, the fed- fixed in nominal value has resulted in substantial fiscal THE IMPACT OF DECENTRALIZATION ON CITIES' FISCAL HEALTH: LESSONS FROM THE UNITED STATES 153 windfalls in most states. The evidence suggests that at reform will also prove successful in reducing poverty is least in some states these excess funds have been used to unclear, however. Several recent studies that have been finance general tax relief or to fund programs unrelated monitoring the employment and earnings history of in- to the well-being of the poor. dividuals who are leaving welfare have found that earn- The important question, however, is whether this de- ings, even for those with full-time jobs, are very modest centralization of cash assistance for the poor to the states (Meyer and Cancian 1998; Parrott 1999; Cancian et al. will in the long run have an adverse effect on the fiscal 1999). These findings suggest that even in those cities health of central cities. Although it is obviously too early with large drops in caseloads, fiscal need may not drop to have a definitive answer to this question, on both em- proportionately to caseload declines. pirical and theoretical grounds there are some indica- In the long run, the greatest danger to central-cities' tions that welfare reform will further weaken the fiscal fiscal health may come from the conversion of federal aid condition of many central cities. for cash assistance to the poor from an open-ended matching grant to a block grant. Because of their rela- Thentra Cimactiof Devosution ontheFiscalHealthof tively short work histories and limited skills, individuals who have recently left welfare will be most vulnerable to Although welfare caseloads have been falling every- unemployment when the economy slips into recession. where, recent data indicate that the rate of decline is Under a system of matching grants, the amount of fed- more rapid in the suburban ring of metropolitan areas eral money states receive increases if an economic down- than in central cities. The consequence of this pattern is turn increases the need for welfare. The conversion of that the remaining caseloads are becoming more con- matching aid to block grants implies that state govern- centrated in central cities. Katz and Allen (1999) report ments must now bear the full fiscal burden of any in- that between 1994 and 1998 welfare caseloads in the creased spending on the poor.15 An economic recession counties containing the 30 largest American cities fell will put further strain on the fiscal health of states because by 35 percent, while the caseload in the states that are state tax revenues generally dedine during recessions. 16 home to these counties declined by 44 percent. In on- Even in the absence of a recession, economic theory going research Howard Chernick and I are finding evi- suggests that the long-run state government response to dence that when one divides urban counties into their a shift from matching grants to a block grant for welfare central-city and suburban portions, caseloads are falling will be a reduction in state government spending on the less rapidly in central cities. For example, in July 1995, poor. Under a matching grant, an extra dollar of welfare the City of Milwaukee, with less than 12 percent of the benefits paid to the poor will cost state taxpayers less state of Wisconsin's population, was home to 51.2 per- than a dollar, while an extra dollar spent on most other cent of the entire state's welfare caseload. By July 1998, state-provided public services will cost a full dollar. The Milwaukee housed 85.1 percent of Wisconsin's welfare shift to a block grant means that the tax price of welfare population.14 rises to a dollar and becomes identical to the tax price Although city governments are not responsible for the of spending on other state objectives.17 Although eco- direct payment of cash transfers to the poor, the grow- nomic theory indicates that this increase in tax price will ing concentration of families on welfare in central cities lead to a reduction in spending on the poor, empirical has several important fiscal implications. First, those in- estimates of the magnitude of the reduction vary dividuals remaining on welfare are probably those with widely.18 Although a precise estimate of the long-run the severest problems-the least education, the highest state government fiscal response to welfare reform is not probability of drug or alcohol dependency, and the possible, the weight of evidence supports the contention greatest incidence of mental illness or retardation. It is that state government spending on the poor will decline precisely these people who are likely to place the great- over time (Chernick and Reschovsky 1999). est demands on city services. To the extent that they are There are likely to be both direct and indirect conse- becoming increasingly concentrated in central cities, the quences for central-city fiscal health if state government relative fiscal position of central cities within their met- support for the poor is reduced. First, as the poor are ropolitan areas will diminish. concentrated in central cities, any policies that reduce The central goal of welfare reform in the U.S. is to the level of economic resources flowing to the poor will move people away from dependency on government cash negatively impact on the revenue-raising capacity of city assistance and into permanent jobs. The falling caseload governments. Furthermore, as local governments serve indicates an initial success in moving a large number of as public service providers of last resort, the reduction poor individuals into the labor force. Whether welfare of federal and state programs to aid the poor will in ef- 154 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION fect shift the responsibility for the poor to municipal of many central cities is growing more slowly than the governments. Even if welfare reform results in a perma- population of their states suggests that city governments nent reduction in the number of individuals dependent may be losing power in state legislatures (Weir 1996). on government transfer payments for support, central To the extent that most state legislators vote for changes cities will in all likelihood remain the place of residence in state and local grants only if the residents of the com- of a disproportionate share of those most in need of gov- munities they represent gain, it is not surprising that re- ernment assistance. To the extent that central cities serve cent changes in intergovernmental aid policies are not increasingly as the home of those most in need, the fis- favorable to central cities. If current population trends cal position of cities will worsen relative to that of their continue, it is unlikely that state governments will play suburbs. an important role in ameliorating the structural fiscal One possible response to devolution would be for problems faced by many central-city governments. state governments to provide additional fiscal assistance to central-city governments. As already indicated, rela- Lessons from the U.S. Experience tively few states provide much general-purpose assis- I have argued in this chapter that the decentralization tance to their local governments. Although there are no of expenditure responsibilities and taxing power to local nationwide data available on recent trends in the spatial governments within metropolitan areas in the United patterns of state assistance, an analysis of the evolution States often results in a situation where central cities are of state assistance in two states suggests that state finan- in weak fiscal health relative to their suburbs. Unless cial support for suburban governments may be increas- some remedial actions are taken, the structural fiscal ing at a faster rate than support for central cities. The problems of cities are likely to continue to worsen. The first example comes from the State of Wisconsin, where concentration of the poor in many central cities will between 1991 and 1997 per capita aid to Milwaukee, only serve to heighten existing inequalities in the provi- the state's largest city, increased by 64 percent, while per sion of basic public services. capita aid to Milwaukee's suburbs rose by 117 percent. It is very important, however, to emphasize that by This distribution of aid, highly favorable to the suburbs, highlighting problems with the American fiscal system, occurred despite the fact that between 1990 and 1997, I am not suggesting that fiscal decentralization should be per capita property values grew over 21/h times faster in avoided. Quite the contrary, the decentralization of both Milwaukee's suburbs than in the central city. In New revenue-raising authority and expenditure responsibili- York State, two recent policy changes appear to weaken ties has large potential benefits, especially in developing the fiscal position of New York City relative to its sub- countries that have a history of highly centralized gov- urbs. First, a school property tax relief initiative is re- ernment. Many developing countries are characterized placing a portion of local property taxes with increased by great local and regional differences. A decentralized state aid. The initiative has been structured so that the system of local government finance has the advantage greatest relief goes to communities with the highest rates that decisions concerning the provision of public services of homeownership, highest property tax rates, and high- can better reflect local economic, social, cultural, and po- est housing values. Because New York City has a rela- litical conditions. Decentralization also has the potential tively low rate of homeownership and low property tax benefit of increasing citizen participation in fiscal deci- rates on owner-occupied housing, in part because of its sionmaking and in enhancing political stability in coun- heavy reliance on nonproperty taxes, relatively little of tries where there exist strong pressures to break up along the new state aid will flow to New York City (City of ethnic or geographic lines (World Bank 1999). New York, Independent Budget Office 1997). Second, The fiscal problems of central cities that I have high- the New York State Legislature has recently passed leg- lighted here do not result from decentralization per se. islation that repeals a commuter tax New York City They occur because of the way decentralization inter- levied on suburban New York residents who worked in acts with a set of political, economic, and institutional the City. This tax had, at least in part, reimbursed the factors that characterize American metropolitan areas in City government for the costs of services it provided to the beginning of the twenty-first century. As an exam- suburban residents. ple, zoning ordinances frequently used by suburban The recent experiences in Wisconsin and New York local governments serve to concentrate the poor, and in are not sufficient evidence to prove that state govern- particular, racial minorities, in central cities. And as we ments in the United States are systematically shifting have seen, the concentration of the poor in central cities state government assistance away from central cities. The has important implications for the fiscal health of city fact, however, that around the country the population governments. THE IMPACT OF DECENTRALIZATION ON CITIES' FISCAL HEALTH: LESSONS FROM THE UNITED STATES 155 The general lesson from the American experience is to a community in stronger fiscal health. If a local gov- not that developing countries should not pursue decen- ernment's weak fiscal condition were caused by ineffi- tralization, but rather that great care must be taken to cient government actions, then such intrajurisdictional design an intergovernmental fiscal system that avoids mobility would be efficiency enhancing. If, however, some of the pitfalls that characterize metropolitan-area weak fiscal health is due to factors outside the control public finance in the United States. Implementing change of local government officials, then this mobility can fur- in a mature intergovernmental system, such as that in ther weaken city fiscal health and augment metropoli- the United States, is extremely difficult. As most changes tan-area fiscal disparities. One policy response to this involve winners and losers, the potential losers will work situation would be the establishment of a system of local very hard to impede any change. Developing countries government grants from a higher level of government, often have the opportunity to design intergovernmental with the grants being targeted to governments in the arrangements and grant systems where none existed be- weakest fiscal health. fore. This puts these countries in a unique position to In the United States, there are no general-purpose be able to learn from the experiences of other countries grants to local governments. While state government with well-developed systems of intergovernmental fi- grants to local school districts are found in all states, only nance and to act upon the lessons learned. a handful of states distribute general-purpose aid to their In the next few paragraphs, I suggest three specific local governments. Even when state aid does exist, the lessons that can be drawn with from the American ex- formulas used to distribute the aid usually account only perience of financing central-city governments. for differences across communities in property values per capita and ignore differences in costs and service respon- sibilities.2' A number of other countries, such as the One of the basic precepts of the theory of fiscal federal- United Kingdom, France, and Switzerland, distribute ism is that within an economy, the central government general-purpose grants to local governments. In France, should pursue any efforts to redistribute resources. The the largest government grant to local governments, the argument, first articulated by Musgrave (1959) and Oates Dotation Globale de Fonctionnement (DGF), is distrib- (1972), is that all attempts by subnational governments uted using a formula that includes a measure of "needs" to redistribute resources through spending programs or (Gilbert 1999a). In a study based on the calculation of progressive taxation are doomed to fail because high- Gini coefficients, Guengant (1996) has found that the income individuals can leave for other jurisdictions which DGF reduces disparities in the fiscal capacity of local pursue fewer pro-poor policies. In the United States, governments in France by about 40 percent.22 spending on redistributive programs is found at all lev- els of government.19 By shifting more responsibility for Metropolitan Cooperation cash assistance for the poor to the states, the recently en- To avoid some of the problems faced by U.S. cities and acted welfare reform legislation has obviously moved in to ensure that city governments will be able to continue the opposite direction from the principles articulated by to provide adequate levels of essential public services to Musgrave and Oates.20 One way to improve the fiscal their citizens, countries must work hard to counter the health of central cities and to reduce the incentive for fiscal externalities that arise when 'local governments local governments to attempt to exclude individuals within metropolitan areas compete for tax bases and when with low incomes, is for the central government to play the provision of services to the poor and the needy be- a dominant role in financing redistributive programs. It comes an important responsibility of local governments. should be emphasized that central government financ- One important strategy is to take steps to enhance ing of poverty alleviation and other social programs does metropolitan cooperation. On one extreme would be the not imply that these programs should be administered creation of metropolitan-area governments that would by the central government. In most cases, program ad- be given the responsibility for the provision of core local ministration, especially when the delivery of social ser- government services. This approach has been taken in vices is involved, is more efficiently done at either the South Africa, which has established metropolitan-area local or regional level. governments that are responsible for the local public good provision in the central city and the surrounding communities. These communities include both high- As long as there are differences in the fiscal conditions income (mainly white) suburbs and the former black of local governments within a metropolitan area, both townships. It is too early to know whether these new businesses and individuals have an incentive to relocate metropolitan governments will be able to reduce the 156 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION great inequities in public-service delivery (and in capital suburbs. One way to overcome this opposition would infrastructure) that currently exist within metropolitan be to grant suburban residents at least a partial income South Africa. or property tax credit against the payment of the city in- In most countries local governments in metropolitan come or payroll tax. The problem with implementing areas are already well established. This will make it po- such a scheme is that it would be opposed by suburban litically very difficult to establish metropolitan govern- governments that would, in effect, be transferring fiscal ments. Nevertheless, various forms of metropolitan-area resources to the central city. One possible way to over- cooperation are possible. On the one hand, these can in- come this suburban opposition would be for the state volve metropolitan-wide provision of specific services. or central government to provide incentives to suburban Ideal candidates for metropolitan provision are public governments to accept this type of tax-sharing scheme. services for which scale economies exist or public ser- The incentives could take the form of direct grants to vices that are characterized by metropolitan-wide exter- suburban governments or income tax credits to subur- nalities. On the other hand, cooperation can take the ban residents (Chernick and Reschovsky 1997). form of suburban government contributions to the fi- nancing of central-city public services, especially those Conclusion that provide direct benefits to suburban residents or that Compared to most countries in the world, the fiscal respond to problems that are arguably metropolitan in structure of the United States is highly decentralized. scope, such as homelessness. Local governments are responsible for providing a wide In the United States, with the exception of metropol- range of public services, including many services pro- itan governments in Indianapolis and Jacksonville, and vided by higher-level governments in other countries. a tax-base sharing scheme in the Minneapolis-St. Paul Although the property tax is the most important local metropolitan area, there are very few examples of met- source of revenue, local governments, especially central ropolitan-area governance. In a number of other coun- cities, also employ a wide range of other taxes and user tries, however, one finds examples of intergovernmental fees. As is well known, fiscal decentralization is assumed grant systems within metropolitan areas. In these sys- to enhance economic efficiency and encourage political tems of horizontal equalization, relatively rich commu- participation in government decisionmaking. I argue in nities or districts contribute funds that are distributed this chapter that given the political and cultural institu- directly to poorer municipalities or districts within the tions present in the United States, fiscal decentralization same metropolitan area. This type of metropolitan-area within American metropolitan areas has led to long-run sharing scheme can be found in Beijing, China and in fiscal problems for many central cities. some cantons in Switzerland.23 Central cities often bear the responsibility of provid- The lesson that can be drawn from the experience of ing public services for the metropolitan area's neediest the United States and from other countries is that met- people in an environment where both businesses and in- ropolitan cooperation, especially cooperative arrange- dividuals can easily move to suburban communities that ments that are beneficial to central cities, is unlikely to provide a more favorable fiscal climate. As a result, city occur unless local governments are provided with spe- residents often face a below-average level of public ser- cific incentives to cooperate. In recent years the French vices, yet an above-average rate of local taxation. I have government has been trying to induce local governments suggested that one likely impact of the recently enacted to cede both service responsibilities and a portion of decentralization of responsibility for cash assistance to their tax base to new metropolitan or regional govern- the needy will be further fiscal pressure on central-city mental bodies (Gilbert 1999b). As the incentive mecha- governments. nism, the government provides an increase in grants to The fiscal problems of American central cities are not local governments that agree to cooperate (mutualisa- an inevitable outcome of decentralization. Newly decen- tion volontaire). It appears that this incentive mecha- tralizing countries and countries in the midst of reform- nism has been quite successful in encouraging metropol- ing their fiscal arrangements can learn from the U.S. itan cooperation in France. experience. Countries should assign responsibility for An alternative form of metropolitan cooperation that providing services for needy populations to the central would benefit central cities would be the imposition by government, implement a system of equalizing grants local governments of income or payroll taxes that are from higher-level governments that accounts both for payable in the jurisdiction where income is earned. Such the expenditure needs and revenue-raising capacity of a tax levied by central-city governments would be op- local governments, and take appropriate steps to inter- posed by individuals who work in the city but live in the nalize fiscal and economic externalities within metro- THE IMPACT OF DECENTRALIZATION ON CITIES' FISCAL HEALTH: LESSONS FROM THE UNITED STATES 157 politan areas. Such countries should then be able to ex- Green, Richard K., and Andrew Reschovsky. 1992. "Fiscal Assis- ploit the considerable benefits of decentralization and tance to Municipal Governments." In Donald A. Nichols, ed., enjoy fiscally strong central cities Dollars & Sense, Volume III. Madison, WI: The Robert M. La Follette Institute of Public Affairs, University of Wisconsin. Bibliography Guengant, Alain. 1996. "Evaluation du Pouvoir P&equateur de la Dotation Globale de Fonctionnement des Communes." In Bahl, Roy. 1999. Fiscal Policy in China: Taxation and Inter- Guy Gilbert, ed., La Pe'requation Financiere Entre les Collec- governmental Fiscal Relations. South San Francisco: The 1990 tivite's Locales. Paris: Presses Universitaire de France. Institute. Gyourko, Joseph, and Richard Voith. 1997. "Does the U.S. Tax Bradbury, Katharine L., Helen F. Ladd, Mark Perrault, Andrew Treatment of Housing Promote Suburbanization and Cen- Reschovsky, and John Yinger. 1984. "State Aid to Offset Fis- tral City Decline?" Working Papers. Economic Research Di- cal Disparities Across Communities." National Tax Journal vision, Federal Reserve Bank of Philadelphia, Philadelphia: 37(June): 151-70. September. Brennan and Edward W. Hill. 1999. "Where Are the Jobs?: Henderson, Vernon, Peter Mieszkowski, and Yvon Sauvageau. Cities, Suburbs, and the Competition for Employment." Sur- 1978. "Peer Group Effects and Educational Production vey Series. Center on Urban and Metropolitan Policy, The Functions." Journal of Public Economics 10(August): 97-106. Brookings Institution, Washington, D.C.: November. Jeanrenaud, Claude, and Andreas Spillmann. 1997. La Pere'qua- Cancian, Maria, Robert Haveman, Daniel R. Meyer, and Bar- tion Financiere dans le Canton de Berne. Stuttgart: Verlag Paul bara Wolfe. 1999. "Before and After TANF: The Economic Haupt. Well-Being of Women Leaving Welfare." Institute for Re- Katz, Bruce and Katherine Allen. 1999. "The State of Welfare search on Poverty Report, University of Wisconsin-Madison, Caseloads in America's Cities: 1999." Survey Series. Center December 14. on Urban and Metropolitan Policy, The Brookings Institu- Chernick, Howard. 1998. "Fiscal Effects of Block Grants for the tion, Washington, D.C.: February. Needy: An Interpretation of the Evidence." International Tax Ladd, Helen F., and John Yinger. 1991. America ' Ailing Cities; and Public Finance 5(May): 205-33. Fiscal Health and the Design of Urban Policy. Baltimore, MD: Chernick, Howard, and Therese J. McGuire. 1999. "The State, Johns Hopkins University Press. Welfare Reform, and the Business Cycle." In Sheldon H. Ladd, Helen F., Andrew Reschovsky, and John Yinger. 1992. Danziger, ed., Economic Conditions and Welfare Reform. Kala- "City Fiscal Condition and State Equalizing Aid: The Case mazoo, MI: W. E. Upjohn Institute for Employment Re- of Minnesota." Proceedings of the 84th Conference on Taxation search, pp. 275-304. of the National Tax Association, 1991. Columbus, OH: Na- Chernick, Howard, and Andrew Reschovsky. 1997. "Urban Fis- tional Tax Association. cal Problems: Coordinating Actions Among Governments." McGeary, Michael. 1990. "Ghetto Poverty and Federal Policies In Burton Weisbrod and James Worthy, eds., The Urban Cri- and Programs." In Laurence Lynn, Jr. and Michael McGeary, sis: Linking Research to Action. Evanston, IL: Northwestern eds., Inner City Poverty in the United States. Washington, University Press. D.C.: National Academy Press. Chernick, Howard, and Andrew Reschovsky. 1999. "State Fiscal Meyer, Daniel R., and Maria Cancian. 1998. "Economic Well- Responses to Block Grants: Will the Social Safety Net Sur- Being of Women Following an Exit from AFDC." Journal of vive?" In Max Sawicky, ed., The End of Welfare? Consequences Magavge anhd tA 19a5m9 The Theoy of Public Finance. New of Federal Devolution for the Nation. Armonk, NY: M.E. York: McGraw-Hill. Sharpe, pp. 157-93. Oates, Wallace E. 1972. Fiscal Federalism. New York: Harcourt City of New York, Independent Budget Office. 1997. "School Brace Jovanovich. Tax Relief and Education Aid Proposals: Impacts on New Oates, Wallace E. 1999. "An Essay on Fiscal Federalism." jour- York City." IBO Fiscal Brief, New York: April. nal ofEconomic Literature 37(September): 1120-49. Council of the Great City Schools. 1999. "Closing the Achieve- Pack, Janet R. 1995. "Poverty and Urban Public Expenditures." ment Gaps in Urban Schools: A Survey of Academic Progress Working paper no. 197, Wharton Real Estate Center, Uni- and Promising Practices in the Great City Schools." Prelimi- versity of Pennsylvania, October. nary Report, October [available at http://www.cgcs. org/ Parrott, Sharon. 1998. Welfare Recipients Who Find Work: What Do ACHGAP.pdfl. We Know About Their Employment and Earnings? Washington, Duncombe, William, and John Yinger. 1997. "Why Is It So D.C.: Center on Budget and Policy Priorities, November 16. Hard to Help Central City Schools?" Journal ofPolicy Analy- Reschovsky, Andrew, and Jennifer Imazeki. 1998. "The Devel- sis and Management 16(1): 85-113. opment of School Finance Formulas to Guarantee the Provi- Gilbert, Guy. 1999a. "Local Taxation and Intergovernmental sion of Adequate Education to Low-Income Students." In Fiscal Relations in France." In Amedeo Fossati, Amedo Fos- Developments in School Finance, 1997. Washington, D.C.: sati, and Giorgio Panella, eds., Fiscal Federalism in the Euro- National Center on Education Statistics, U.S. Department of pean Union. Routledge Studies in the European Economy, Education, 1998: 121-48. No. 9, London: Routledge. Summers, Anita A. and Lara Jakubowski. 1996. "The Fiscal Bur- Gilbert, Guy. 1999b. "Quelles reformes pour le Financement den of Unreimbursed Poverty Expenditures in the City of des Collectivites Locales?" Cahiers Franfais 298(octobre- Philadelphia: 1985-1995." Working paper no. 238, Whar- decembre): 61-69. ton Real Estate Center, University of Pennsylvania, August. Glaeser, Edward L. 1998. "Are Cities Dying?" Journal of Eco- U.S. Bureau of the Census. 1997. StatisticalAbstract ofthe United nomic Perspectives 12(Spring): 139-60. States 1997. [available at http://www.census.gov/statab/www]. I58 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION U.S. Bureau of the Census. 1999. Poverty in the United States, tially more state financial assistance than city governments with 1998, Current Population Reports, Series P60-207, Wash- narrower responsibilities. ington, D.C.: U.S. Government Printing Office. 10. Excluding Boston, Detroit, and Milwaukee, general- U.S. Department of Health and Human Services, Administration purpose state aid equaled only 2.5 percent of direct expenditures for Children and Families. 1999. "Change in Welfare Case- in the nation's largest cities. These calculations are based on data loads Since Enactment of New Welfare Law." [available at from the U.S. Census Bureau's City Government Finances for http://www.acf.dhhs.gov/news/stats/aug-sept.htm], December. 1993-94 (available at http://www.census.gov/govs/city/). U.S. Department of Housing and Urban Development. 1998. 11. Although recent federal housing policy has emphasized State of the Cities in 1998. Washington, D.C.: Office of Pol- housing subsidies that low-income households are able to use in icy Development and Research. city as well as suburban locations, existing evidence suggests that Wasylenko, Michael. 1997. "Taxation and Economic Develop- this policy has done little to stimulate mobility by low-income ment: The State of the Economic Literature." New England households to the suburbs. Economic Review (March/April): 37-52. 12. For some early work on this subject, see Gyourko and Weir, Margaret. 1996. "Central Cities' Loss of Power in State Voith (1997). Politics." Cityscape: A Journal of Policy Development and Re- 13. Between August 1996 (the date of the signing of the wel- search 2(May): 23-40. fare reform legislation) and June 1999, the number of welfare World Bank. 1999. World Development Report 1999/2000. recipients in the United States has declined by 44 percent (U.S. Washington, D.C. Department of Health and Human Services 1999). Yinger, John. 1997. Closed Doors, Opportunities Lost: The Con- 14. We have also found that the rate of decline in welfare tinuing Costs of Housing Discrimination. New York: Russell caseloads between 1995 and 1998 was less in the eight largest Sage. central cities in California than in their respective suburbs. Notes 15. The current welfare reform legislation does allow limited state borrowing (with interest) from a "rainy day" fund if a state's 1. For an excellent review and analysis of decentralization ef- economy slips into recession. forts around the world, see Chapter 5 of the World Bank's World 16. See Chernick and McGuire (1999) for a discussion of the Development Report 1999/2000. empirical issues involved in estimating the impact of a recession 2. See Oates (1999) for a good discussion of the conditions on state welfare spending in light of the shift to a block grant for under which competition among local governments will be effi- welfare. ciency enhancing. 17. In fact, the story is more complex. Because increases in 3. In a recent report, the Council of the Great City Schools family income lead to reductions in the amount of Food Stamps (1999) reports on the low test score results achieved by many stu- to which a family is entitled, the tax price of a state increase in dents in the public schools of the nation's largest cities. cash assistance is actually greater than one (Chernick 1998). 4. Fiscal advantages and disadvantages are likely to be at least 18. See Chernick (1998) for a comprehensive review of the partially capitalized into housing prices. While this capitalization econometric literature. will reduce the cost of housing for individuals living in commu- 19. Edward Glaeser (1998) has suggested that cities' fiscal nities in weak fiscal condition, it does not eliminate intercom- health is improving because they have learned the Musgrave and munity inequities in both access to public services and tax prices Oates lesson that attempts at redistribution are destined to fail. faced by residents. Glaeser ignores the fact that although some cities, such as New 5. There is a large empirical literature in both the United York, have reduced programs targeted to the poor, many cities States and Europe addressing the role of fiscal factors in the in- have limited ability to do so because state governments or court trametropolitan locational decisions of households and busi- rulings mandate city responsibility for the homeless, the men- nesses. Although these studies present a wide range of findings, tally ill, and others in need of some type of public assistance. there appears to be broad support for the contention that fiscal 20. In a recent article, Oates (1999) suggests that by shifting factors play a significant role in locational decisions within met- the responsibility for welfare to the states, the U.S. Congress may ropolitan areas. For a good summary of the U.S. literature on the have wanted to adopt a system that would (in theory) encourage role of taxes in locational choices, see Wasylenko (1997). states to experiment with new ways to end dependency and alle- 6. After completing a detailed analysis of the budget of the viate poverty. City of Philadelphia, Summers and Jakubowski (1996) con- 21. In a study of the distribution of state aid to local govern- cluded that in 1995 the City devoted 7.6 percent of its own ments, Ladd, Reschovsky, and Yinger (1992) found that per source revenues to direct poverty-related services. In another capita aid was uncorrelated (r = 0.10) with an index of local gov- study, Pack (1995) reports that larger cities spent more money ernment fiscal health that included measures of local government per capita on direct poverty functions than smaller cities. revenue-raising capacity, costs, and service responsibilities. 7. Examples of these studies include Bradbury et al. (1984), 22. In ongoing research Gilbert and Guengant are evaluating Ladd and Yinger (1991), Ladd, Reschovsky, and Yinger (1991), the entire system of grants to local governments in France. They and Green and Reschovsky (1992). are attempting to assess the extent to which grants reduce in- 8. Two recent studies that estimated cost functions for public equalities in both the fiscal capacity and expenditure needs of education and found a strong relationship between concentrated local governments. poverty and educational costs are Duncombe and Yinger (1997) 23. See Bahl (1999) for a description of the grant system used and Reschovsky and Imazeki (1998). in Beijing metropolitan area and Jeanrenaud and Spillmann 9. City governments responsible for elementary and sec- (1997) for a discussion of the system used in the Berne metro- ondary education and for public health generally receive substan- politan area. chapter 18 How Singapore Regulates Urban Transportation and Land Use Sock-Yong Phang School of Business Singapore Management University _ ingapore is a small island city-state that such taxation would abolish poverty and economic with a population of 3.9 million and crises, as the latter, he said, resulted from speculation in a land area of 648 square kilometers. land values. In its approach to urban transportation and Without any natural resource except land regulation, Singapore has embraced a large dose of its strategic location, the country is George's prescription (Phang 1996). The state not only economically dependent on trade and owns four-fifths of all land but also controls the rights capital flows. To attract mobile capital, the government to road use. has tried to ensure that immobile factors-including land This approach began in 1965 when the country faced use, housing, infrastructure, public services, and labor- severe unemployment and an acute housing shortage complement growth. In particular, the city-state has pur- after gaining independence from Malaysia and Britain. sued a highly regulated approach to urban transportation State ownership and control of land were considered es- and land use-an approach that contrasts sharply with sential to attracting industrial investments and building its willingness to embrace foreign direct investment. public housing on a large scale. These policies have yielded stunning success. The In 1966 the government enacted the Land Acquisition 1998 World Development Report ranks Singapore fourth Act, which permitted the state and its agencies to acquire worldwide in gross national product (GNP) per capita, land for any public purpose. A 1973 amendment set pay- after Switzerland, Japan, and Norway. Annual growth ments independent of market conditions and the has averaged 7 percent per capita since the 1960s, and landowner's purchase price. In 1951, to better enforce the government has posted a budget surplus since 1968. government control over private land use and develop- Income tax rates have declined steadily, corporate in- ment, the British colonial government enacted the Plan- come taxes have dropped from 40 percent to 26 percent, ning Ordinance (now the Planning Act). The Planning and there is no payroll tax. Meanwhile, over 5,000 for- Act then spawned a Master Plan, which designates the eign companies have located in Singapore, and many type and intensity of use for privately owned areas. have made it their regional operating and manufactur- Higher densities and zoning alterations incur a surcharge, ing base. These experiences reveal that, despite the city- and all developers must obtain written permission before state's somewhat unique size and sovereign status, the ar- subdividing their holdings or making any alterations. gument for extensive government control and planning The government reviews the Master Plan every five years. remains persuasive in the area of urban development.1 Public-sector projects were subject to the 1971 Con- Controlling Land Use cept Plan, a strategic blueprint for land use and urban transportation for the entire city-state. The revised Con- Nineteenth-century economist Henry George made a cept Plan, released in 1991, includes comprehensive strong case for 100 percent taxation of rents from land, urban-design guidelines for a projected population of 4 which he regarded as common property. George claimed million people. I59 i6o LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION However, unlike in former socialist cities where gov- tion and ensures affordable housing prices and rents in a ernment ownership of land and the absence of markets land-scarce city. Some 22 public housing estates and new undercut efficiency, productivity, and environmental towns are comprehensively planned, with the number of quality, Singapore actively relies on markets to promote dwelling units in each ranging from 20,000 to 62,000. efficient land use. First, the one-fifth of land and associ- Yet more than 90 percent of households are homeown- ated real estate that belong to the private sector provides ers: The state leases 90 percent of the housing it builds valuable price information to urban planners. Second, on a 99-year basis to eligible household owners. as in Hong Kong where the state owns all the land, the The government offers price discounts on newly con- Singapore government uses a competitive bidding structed public housing and grants to households that process to allow private development of state land. The purchase housing on the resale market. Households, in bidding process for a particular plot of state land is ad- turn, finance their purchases through a compulsory sav- ministered by either the Urban Redevelopment Author- ings scheme known as the Central Provident Fund, ity or the Housing and Development Board as agent for which requires all employees and their employers to the Land Office. The reserve price for each plot is deter- contribute 20 percent of monthly wages. (The employ- mined by the chief valuer. The bidding process is usu- ers' share recently dropped from 20 to 10 percent be- ally a closed tender exercise for which detailed proce- cause of the 1997 recession.) People may withdraw these dures for the agent and bidders have been laid down as compulsory savings at age 55, use them to finance mort- financial regulations. The process is subjected to both gage payments, education, and medical expenses, or in- internal and external inspection and audit to ensure vest them in approved financial assets. competition and fairness. Invitations to tender or par- Rules encourage submarkets in government land and ticipate in open auctions (which are sometimes held in- property leases but discourage speculation. For exam- stead) are given the widest possible circulation to attract ple, there is an active resale market for government- the largest number of competitive tenders. Government- leased flats-but only after owners occupy the units for owned and government-linked companies also com- a minimum period of time. In fact, the volume of trans- pete with each other and with privately owned compa- actions on the housing resale market exceeds the volume nies to develop both private and public land. For ex- of new flats, and resale prices help the government set ample, the Jurong Town Corporation develops and prices of new flats. (Similarly, a successful bidder for the manages 30 industrial estates, including high-tech and development of a government site cannot resell to an- business parks. other single party before completing the project with- The government announces the total supply of units out permission from the government, but after complet- that its land sales program is expected to generate one year ing it faces no restrictions on resale.) ahead of time and conducts a midyear review before fi- The government uses zoning and careful screening of nalizing the program for the latter half of the year. Ten- hazardous industries to control emissions and imposes der documents for government land sales typically spec- buffer distances from housing according to industries' ify use, plot ratio, building height and layout, parking, classification as light, general, special, and hazardous. time frame for completion, and open spaces. Through Together these land-use and planning efforts have com- this process government leases have allowed private pletely eliminated slums and squatter settlements, and development of offices, hotels, shopping centers, ware- the population enjoys universal access to clean water and houses, recreational facilities, and residential projects. sanitation. Such leases, which typically run for 99 years or less, di- rect development toward priorities such as the new down- Regulating Urban Transportation town and the financial district, conservation and renewal Despite a gross domestic product (GDP) per capita of of specific buildings and areas, and new housing (Urban S$37,214 in 1999, only one Singaporean in ten owns a Redevelopment Authority 1995), as do tax exemptions car. To allocate rights to car ownership, Singapore again and deductions. Government receipts from such leases to- combines market mechanisms with taxation and active taled S$ 11 billion in 1995-96-nearly half of the gov- restrictions designed to contain congestion. The rationale ernment's operating revenue of S$28 billion that year. for such an approach is simple: Roads already constitute Financing Public-Built Housing 12 percent of the island's area, about the same percentage Financingme Publrshic-Bu Housinas housing, so room for continued expansion is clearly and Home Ownership limited. To the extent that road use rights represent ac- Singapore's extensive public housing program provides cess to a land-related resource, taxation of those rights is living quarters for more than 80 percent of the popula- completely in line with Henry George's prescription. HOW SINGAPORE REGULATES URBAN TRANSPORTATION AND LAND USE i6i A quota system restricts the number of motor vehi- monopoly on bus routes while a government-owned cles: Before buying a vehicle, a prospective owner must company operates the rapid-transit system and a linked obtain a certificate of entitlement through competitive 8-kilometer light-rail system. The government has an- bidding. This bidding set the price of a certificate for a nounced plans to allow the development of two major medium-sized car at S$43,802 in January 2000. The multimodal public transport operators. Because of this owner must also pay an import tax of 41 percent, a reg- comprehensive approach to transport, levels of carbon istration fee of 140 percent, and annual road taxes that monoxide, sulfur dioxide, nitrogen oxide, and lead in depend on the vehicle's engine capacity. As a result, a typ- the air all fall within the guidelines of the U.S. Environ- ical medium-sized car costs up to three times per capita mental Protection Agency. GDP-in most high-income countries that ratio is less than one. For example, a new Toyota Corolla with an im- port value of S$18,000 cost approximately S$1 10,000 Singapore's ability to pursue such effective planning (or US$65,000). stems from a network of competent, noncorrupt institu- The government also tackled road congestion di- tions that together provide rich public-sector capacity. rectly-and early, in 1975-by instituting an "area li- Public-sector entrepreneurs at numerous government- censing scheme" in which motorists entering heavily linked companies (such as the world-renowned national trafficked sectors had to purchase and display a coupon. airline) also compete with private firms both domesti- Electronic road pricing (ERP) replaced this scheme in cally and internationally in activities as diverse as tele- 1998. ERP charges vary by type of vehicle, time, and lo- communications, airlines, shipping, banking, shipyards, cation, and involve a combination of radio frequency, manufacturing, health care, and trading.2 Government- optical detection, imaging, and smart card technologies. linked companies are self-financing and most public- ERP relies on a pair of gantries: The first has antennas sector institutions have also been structured to be self- that check smart cards on approaching vehicles and financing. Table 18.1 lists the key public-sector agencies debit the card. The second gantry pinpoints the location involve in urban transport and land-use activities. of the vehicle, identifies the vehicle type, and verifies Why and how does the public sector in Singapore that the correct deduction has been made. A controller perform so well? The government has developed a rig- links information from the antennas and detectors to orous recruitment system: It sponsors the nation's "best check for possible violations (no smart card, insufficient and brightest" to attend top universities in the United balance on the card), and cameras transmit an image of Kingdom and United States, who then return to become the rear license plate to a central computer in the event public officials. Senior civil servants earn perhaps the of a violation. highest salaries among bureaucrats in the world, with The Land Transport Authority conducts quarterly re- their pay benchmarked against private-sector salaries.3 views to adjust charges to yield targeted "optimal" speed This system helps the government minimize corruption ranges of 45 to 65 kilometers per hour for expressways and recruit and retain talent, including midcareer offi- and 20 to 30 kilometers per hour for arterial roads. cials from the private sector.4 And recruitment and pro- Other charges for vehicles include parking fees and fuel motion are based on merit rather than seniority, kinship, taxes, which are high enough that the tank of any vehi- or ethnic ties. Supervisors assess employees' ultimate po- cle leaving the country must be three-quarters full. Total tential and promote high flyers quickly so they will levies on motor vehicles generally accounted for nearly reach the top ranks by age 45. Finally, extensive com- one-quarter of the government's operating revenue and puterization enhances public employees' productivity. nearly one-half of the government's operating expendi- The civil service has developed a culture where cor- ture in recent years. ruption is viewed as a "high risk, low reward" activity Because of such policies, Singaporeans use public (Quah 1996, p. 76). The powers of the Corrupt Practices transport to make 5 million of 7 million trips daily-of Investigation Bureau are defined under the Prevention which 3.1 million occur on buses, 0.9 million are on of Corruption Act 1960. The Corruption (Confiscation rapid transit, and 0.9 million are in taxis. The govern- of Benefits) Act of 1989 provides for the confiscation of ment completely funded the cost of the 83-kilometer benefits derived from corruption. Whether the Singapore rapid-transit system, and fares are low, ranging from 35 style of public administration is transferable elsewhere is cents to $1.60 per trip, compared with the quality of debatable because of the high economic and political service. An integrated ticketing system allows com- costs involved (Quah 1996). Competitive public-sector muters to use a single stored-value card to travel on both pay and large-scale computerization are economically buses and trains. Two major private companies enjoy a costly and may be politically costly as well. Introducing i62 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Table 18.1 Public Sector Institutions and Agencies Involved in Urban Transportation and Land Use Activities Institution/agency Primary responsibilities Ministry of Law: Primary responsibilities Land Office * Oversees land use and registration Registry of Land Titles and Deeds * Acquires private land under the Land Acquisition Act Survey Department * Manages and controls state land Land Surveyors Board * Alienates state land to both the public and private sectors for development Land Systems Support Unit * Collects land revenue Appeals Board (Land Acquisition) Strata Titles Board Urban Redevelopment Authority: * Acts as the national land-use and conservation planning authority statutory board under Ministry of * Formulates long-range land-use Concept Plan and translates it into specific planning National Development guidelines for local areas, which then form the Master Plan * Enforces planning regulations and determines and collects development charges * Acts as agent to the Land Office in "packaging" and selling state land to the private sector to facilitate focused development Jurong Town Corporation: * Acts as main developer and owner of industrial land in Singapore statutory board under Ministry of * Manages 30 industrial and business parks Trade and Industry Housing and Development Board: * Develops comprehensive estates that house 86 percent of the population and include statutory board under Ministry of infrastructure facilities, commercial, industrial, recreational, and public premises National Development * Provides subsidized mortgage loans that enable people to purchase its homeowner- ship flats - Administers the allocation system for subsidized public housing - Administers resale transactions - Provides housing grants for resale units - Upgrades and redevelops older housing estates - Acts as an agent of the Land Office in the sale of state land for private residential development and land reclamation Land Transport Authority: * Acts as agent of the government in administering, assessing, collecting, and enforcing statutory board under Ministry of motor vehicle taxes, fees, and charges Communications and Information * Plans, designs, constructs, manages, and maintains road and rail network, and Technology coordinates land transport services Public Transport Council: * Regulates entry and exit of bus service providers by issuing bus licenses statutory board under Ministry of * Approves increases in fares for public transport Communications and Information * Ensures that bus operators comply with PTC's service standards, which include route Technology planning and design specifications, service efficiency, hours of operations, affordability, and service information a meritocratic system and minimizing corruption also re- in former socialist countries and in Asia are also major quire the widespread support of the population. landowners, yet the absence of markets often makes Transferability of the Singapore . Experiencethese cities inefficient (Bertaud and Renaud 1995). Those cities can learn much from Singapore's planning Not every government may find it feasible to make process and its active role in creating markets. housing and competitiveness in attracting foreign in- Larger countries cannot raise taxes on vehicle owner- vestment high priorities. Indeed, one could argue that ship to Singaporean levels because doing so would im- the draconian land policies that help facilitate integrated pose a heavy penalty on intercity transport. However, planning in Singapore work because of its peculiar cir- the Singapore experience does show that larger cities can cumstances. For example, few large private landowners use road pricing to manage congestion and that rev- existed in the 1 960s and 1 970s, when separation from enues can cover and even exceed the cost of providing Malaysia and the withdrawal of British troops created a and maintaining roads.5 Despite this experience, only a sense of urgency. However, numerous city governments few other governments have implemented congestion HOW SINGAPORE REGULATES URBAN TRANSPORTATION AND LAND USE I63 pricing: Congestion pricing has been implemented on Oldridge, B. 1994. "Congestion Metering in Cambridge City, the intercity A- I highway in France and on toll rings in United Kingdom." In Borje Johansson and Lars-Goran Matt- the Norwegian cities of Bergen, Oslo, and Trondheim. son, eds., Road Pricing: Theory, Empirical Assessment and Pol- Winning political approval *fr congestion pricig is .icy. Boston: Kluwer Academic Publishers. winning political approval for congestion pricing is Phang, S. Y. 1996. "Economic Development and the Distribu- admittedly difficult: In Hong Kong, Stockholm, Lon- tion of Land Rents in Singapore: A Georgist Implementa- don, and the Randstad, intense controversy defeated or tion." The American Journal of Economics and Sociology. Oc- delayed congestion pricing proposals. Many motorists tober: 489-501. feel they stand to lose under such a policy, particularly if Quah, J. S. T. 1996. "Public Administration in Singapore: Man- the government does not return the toll revenue to them aging Success in a Multi-Racial City-State." In A. S. Huque, J. T. M. Lam, and J. C. Y. Lee, eds., Public Administration in some form. According to Oldridge (1994), "No city in the NICs: Challenges and Accompishments. Basingstake: anywhere in the world that is run on democratic princi- Macmillan Press. ples will be politically able to accept congestion pricing Toh, R., and S. Y. Phang. 1997. "Curbing Urban Traffic Con- unless it is gained through a slow evolutionary process." gestion in Singapore: A Comprehensive Review." Transporta- Such political problems, though difficult, are not in- tionJournal37(2): 24-33. soluble, particularly in developing countries where rates Urban Redevelopment Authority, Singapore. 1991. Living the Next Lap: Towards a Tropical City ofExcellence. of car ownership are still relatively low. Those countries Urban Redevelopment Authority, Singapore. 1995. Changing should note that Singapore introduced its area licensing the Face of Singapore through the URA Sale of Sites. scheme some 25 years ago. Growing cities would do well World Bank. 1998. World Development Report 1998. Washing- to begin to regulate and enforce road-use pricing now, ton, D.C. as well as develop their ability to manage efficient mass www.gov.sg for web sites of Singapore government ministries 6 ~~~~~~~~~~~~~~~~and statutory boards. transit. Yuen, B., ed. 1998. Planning Singapore: From Plan to Implemen- The tactics on which Singapore relies-state land tation. Singapore Institute of Planners, pp. 81-132. ownership, regulatory takings, and compulsory savings, complemented by an extensive public sector-could Notes easily have spawned widespread inefficiency and corrup- 1. Recent literature on various aspects of land use and urban tion, as they have in other countries. But voters have re- transport regulations in Singapore includes a collection of papers turned the People's Action Party to power continuously edited by Yuen (1998), which details different dimensions of the planning process, including sustainable development, the Mas- since *because i is honest, efficient, and ter Plan, renewal of public housing estates, urban transport plan- effective in delivering a high quality of life. This politi- ning, and urban conservation. Toh and Phang (1997) reviews cal stability contributes to the city's ability to attract for- the literature analyzing policies to curb traffic congestion in Sin- eign investment. gapore. Quah (1996) describes and analyzes the contribution of Intervention by cities whose government and leader- the public sector to the economic success of Singapore. The Web ship are weak and corrupt. may carry a higher cost than sites of various government ministries and public institutions (see ship are weak and corrupt may carry a higher cost than Table 18.1), found at www.gov.sg, contain much detailed infor- inaction. However, many cities, regions, and countries mation on current policies, prices, and other indicators of mar- would benefit by developing efficient and effective public- ket conditions. sector institutions and emulating the coincidence of inter- 2. A government-linked company is one in which the govern- ests between Singapore's government and its people. ment has equity holdings and is formed under the Companies Act. Bibliography 3. A new minister earns two-thirds of the average income of the top four individuals in six sectors: banking, accountancy, en- Bertaud, A., and B. Renaud. 1995. Cities Without Land Markets: gineering, law, local manufacturing, and multinational corpora- Location and Land Use in the Socialist City. World Bank Pol- tions. Meanwhile a good administrative officer at age 32 earns icy Research Working Paper 1477. the average income of the 15th person aged 32 in the same six Cmd 13. 1994. Competitive Salaries for Competent and Honest sectors (Cmd. 13, 1994). Other salaries are extrapolated from Government: Benchmarks for Ministers and Senior Public Offi- these two points. In 1992 the private benchmark figure for a new cers. White Paper, 21 October, Singapore. minister was S$800,000 (US$500,000) and that of a 32-year-old George, H. 1879. Progress and Poverty. New York: Robert public administrator $199,000 (US$120,000). Schalkenbach Foundation, 1979 edition. 4. The ability to attract private sector expertise into the pub- Hau, T. D. 1992. Congestion Charging Mechanismsfor Roads: An lic sector has been particularly important for the judiciary and Evaluation of Current Practice. WPS 1071. Washington, for central banking. D.C.: The World Bank. 5. For more on various combinations for cost recovery, see Heggie, I. G., and P. Vickers. 1998. Commercial Management Heggie and Vickers (1998). and Financing of Roads. World Bank Technical Paper No. 6. See Hau (1992), pages 62-63, for guidelines on conges- 409. Washington, D.C.: The World Bank. tion charges appropriate for developing countries. chapter 19 To Plan or Not to Plan: Southeast Asian Cities Tackle Transport, Communications, and Land Use Peter J. Rimmer Department of Human Geography, Research School of Pacific and Asian Studies The Australian National University Howard Dick Department of Management University of Melbourne he megacities of Southeast Asia in- opment Project (Crooks Michell Peacock Stewart 1971). trude into world consciousness at The infant republic had just separated from the larger opposite extremes (Rimmer 1986; state of Malaysia-an event soon followed by the staged Dick and Rimmer 1998 1999, forth- withdrawal of British forces. To avoid soaring unemploy- coming). Singapore has earned a jus- ment and political tensions, the government sought to tifiable reputation as a model world attract foreign investment, especially in labor-intensive city. Bangkok has become notorious as arguably the manufacturing. One aspect of a reliable workforce was a globe's most congested city. Meanwhile Jakarta and transport system that could get people to work on time, Manila resemble Bangkok rather than Singapore, while especially from new high-rise public housing. Hong Kong and Kuala Lumpur are closer to the Singa- Singapore's transport policy focused on both medium- pore model. What accounts for such a marked differ- and long-term horizons while also benefiting from con- ence among these major cities? tinuous learning. The first measures involved rational- This chapter will show how different cities utilize ization of the existing slow-speed bus transport system. land use, transport, and communications-as well as During the 1970s the government merged four main planning underlying all three-to produce very differ- bus companies into one public corporation, while bus ent outcomes. We will also contrast the experiences of routes integrated with housing estates and new towns megacities with those of medium-sized and smaller and industrial estates to allow for affordable and reliable cities. Although cities of varying sizes often need to commuting. focus on different priorities, all would gain from greater Although the government invested substantial funds access to local revenue sources and more responsiveness in new highway links, by the mid-1970s it recognized to local constituencies. that accommodating rapid motorization would only lead to worsening congestion. The subsequent initiative and Kuala Lumpur for tackling downtown congestion restricted access to the central business district through the pioneering In Singapore urban planning is a task of national gov- Area Licensing Scheme and raised license and parking ernment and is a national priority. The land constraint charges until the cost of car ownership became proba- is so tight and reclamation so expensive that the cost of bly the highest in the world. In 1998 electronic moni- planning errors is high. Moreover, the city itself is Singa- tors began to record and charge vehicles entering the pore's prime resource: To compete on the world stage it central business district-an innovation in which Sin- must coordinate port, airport, and telecommunications gapore again leads the world. with industry, commerce, and housing. The government also decided that the transport sys- Integrated transport and land-use planning began in tem's dependence on buses and private cars was no longer Singapore in 1967 with the Urban Renewal and Devel- a viable long-rerm strategy. Construction of a mass I64 SOUTHEAST ASIAN CITIES TACKLE TRANSPORT, COMMUNICATIONS, AND LAND USE I65 rapid-transit system began in 1983 at a cost of US$2.3 density city-region with a high rate of vehicle ownership billion, and the first trains ran in 1989 (Rimmer 1986; that resembles a small Los Angeles. Authorities are now JUTS 1997). Ten years later this network has become the trying to retrofit the 24-kilometer elevated light-rail backbone of a well-articulated transport system. Plan- system opened in 1996 to serve this less-than-ideal ning has also ensured that most new housing and em- situation for public transport use. Authorities are also ployment locates along the rapid-transit corridors, reduc- making another attempt to retrofit the recently electri- ing demand for cross-city movement. Both the port and fied commuter railway into the Kelang Valley corridor, airport are situated within these axes. No other South- which combines infrastructure, manufacturing, and east Asian city has achieved this level of sophistication. both middle-class and working-class housing, along Singapore has also led the way in telecommunica- with a seaport and domestic airport. tions development, pushing beyond basic functions into Like Singapore, the Malaysian government aims to data networks that facilitate global trade and finance. develop a high-tech information corridor in the hinter- The government has recently eased its restrictive policy land of Kuala Lumpur on the model of Silicon Valley on Internet access and media censorship, reinforcing its (Malaysia 1996). Taking advantage of large tracts of land commitment to promoting Singapore as a high-tech in- formerly used for plantations, the Multimedia Super formation processing and research center. Corridor (MSC) covers an area stretching 50 kilometers The case most comparable to Singapore is Hong by 15 kilometers. The corridor encompasses the twin Kong. Both function as transport hubs, international Petronas Towers of the Kuala Lumpur City Center, the business centers, and tourist destinations. And both have new administrative capital of Putra Jaya, the new man- experienced the switch from basic to high-value manu- ufacturing center known as Cyber Jaya, and the relo- facturing, services, and logistics and the need to develop cated Kuala Lumpur International Airport (Sirat 1998). cross-border links as part of their larger regional roles. Almost 300 information technology companies-in- Like Singapore, Hong Kong planned infrastructure de- cluding 30 world-class enterprises-have been attracted velopment along expressway and fixed-route public- by the liberal cyber laws and received MSC status. The transport corridors. High-capacity metro, rail, light-rail, degree to which these firms will attract the latest infor- and cross-harbor tunnel projects extend the reach of mation technology and stimulate the Malaysian econ- older bus, ferry, paratransit, and tram networks. The omy will depend on the recovery in foreign investment long-term aim is to reduce commuters' reliance on road- after the Asian crisis. based transport by integrating employment centers with rail stations, adding pedestrian walkways to selected U C roads, and providing segregated overhead walkways. Bangkok, Jakarta, and Manila, in contrast, have applied Despite their similarities in linking transport and in- modern transport technologies piecemeal-uncoordi- dustrial development, Hong Kong and Singapore do nated by any consistent long-term strategy. Policy has show marked differences in long-term planning. There focused on accommodating rapid growth in private ve- is more private ownership of land in Hong Kong. The hicle traffic. Jakarta has built the most integrated toll- latter also relies more heavily on private investment to road network, with port, airport, and radial expressways fund infrastructure (such as in developing a container connecting via a central ring road. Manila has built an port). Hong Kong has preferred franchised private bus elevated ring road that connects with northern and and ferry companies, accompanied by public regulation southern toll roads. Bangkok has managed to build only of any transport monopolies. And Hong Kong has made a few sections of freeway network, which have, if any- more use of paratransit, though its operation in the met- thing, exacerbated inner-city congestion. These cities ropolitan area has been constrained to scheduled ser- have also implemented one-way streets, truck bans in vices on fixed routes at fixed fares. After some early ex- daylight hours, dedicated bus lanes, minimum car oc- periments, Hong Kong has not applied stringent traffic cupancy, odd and even access days, and freight distribu- restraints in downtown areas. tion terminals, but have paid almost no attention to the The only other capital city in Southeast Asia that has needs of pedestrians and bicyclists. The urban poor have implemented long-term planning is Kuala Lumpur. Be- thus seen their access impeded and dangers from vehi- cause land is cheap, transport development has until re- cle traffic worsen. cently been almost entirely road-based and facilitated Public transport in these cities has until very recently settlement at lower density than in any other Southeast relied on buses supplemented by paratransit such as Asian city. The city's freeway/tollway system has accom- minibuses and jeepneys. Ambitious plans to modernize panied low-rise new towns, producing a sprawling low- bus fleets have invariably foundered on lack of sustained i66 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION funding and bad management. Competition has pro- developing into a minimal network. In December 1999 duced some air-conditioned, limited-stop buses charging the first 10 kilometers of the new elevated mass rapid- premium fares, but most passengers put up with crowded, transit system opened along the ring road between Que- poorly maintained, and heavily polluting vehicles. zon City and Makati; the full loop line is scheduled to Notwithstanding the lack of comfort, flat-rate fares connect back into the original line in 2002. In Bangkok of the bus systems do allow cheap urban commuting- the 23-kilometer elevated Skytrain also commenced op- albeit at slow speed. And though criticized as a cause of eration in December 1999, but connections to main traf- traffic congestion, minibuses (in Bangkok and Jakarta) fic nodes await integration with two other projects de- and jeepneys (in Manila) carry a large number of pas- ferred because of the Asian economic crisis. Jakarta has sengers over short-to-medium distances with stops on rehabilitated its electric railway system, including an demand and minimal waiting times. The perception urban loop line, as a potentially high-volume commuter that such vehicles obstruct private vehicles and cause network, but further investment will be needed to har- congestion-though influential in the making of trans- vest significant passenger flows. Funding constraints, po- port policy-ignores the fact that paratransit vehicles litical squabbling, and outright corruption have further have much higher productivity per unit of road space ensured that little has been built. The region's economic and dollar of road investment. Their drawback is slow- crisis will further defer completion indefinitely and tempt ness over long distances. policymakers to restrict infrastructure spending to incre- Despite substantial investment in urban roads and mental development of expressways. expressways, speed per kilometer in these three cities has However, Manila, Bangkok, and Jakarta should seize remained constant or, as in Bangkok, even fallen. Ex- the opportunity to formulate 20-year land-use plans. pressways typically feed into and out of bottlenecks, The clear lesson from Singapore and Hong Kong is that so travel time is unpredictable and consists largely of efforts to funnel new industry and housing along pri- idling, which is costly and leads to scheduling problems, mary transport corridors optimize returns from long- despite the widespread use of mobile phones. These and term investments in rail infrastructure. Incremental road other Southeast Asian cities can further streamline traf- and expressway investment and the lack of a long-term fic, but even costly expansion of expressway systems can transport plan, in contrast, allow private and often spec- only delay other solutions to the problem. ulative land-use decisions to shape the urban structure. Telecommunications access is much better than trans- Transport planning becomes reactive and unstable as port: All three cities have installed widespread basic in- each new increment reinforces the existing pattern. frastructure. Satellites provide television and telephone The underlying problem is that Southeast Asia has access to virtually the entire urban population. The less seen both industrialization and rapid urbanization over well-off gain access to the network through public tele- the past three decades. High rates of investment in and phone booths-even using them to make international around main capital cities have attracted steady in-migra- calls. (Because of the high number of overseas Filipino tion, producing spectacular growth in city size, whether workers, the Philippines has one of the highest rates of measured by population, area, or regional domestic prod- incoming international calls in the world.) uct (Ginsburg, Koppel, and McGee 1991). The urban economies of Bangkok and Jakarta are now larger than the economy of Singapore, and greater Kuala Lumpur Because a fixed-route public-transport network is either and Manila are following close behind (Table 19.1). nonexistent or rudimentary in Southeast Asian cities Urban sprawl is exacerbated by governments' ten- other than Singapore, the dilemma is the high initial cost dency to tackle the symptom rather than the cause. New of a public-transport network. Virtually all large Euro- expressways are invariably accompanied by new sites for pean and Northeast Asian cities, in contrast, have sub- industrial estates and new towns, and they allow execu- way networks, while medium-sized European and some tives and managers of large factories to live in exclusive Japanese cities maintain tramway systems and discour- residential areas such as Pondok Indah in Jakarta or age vehicle access. These systems feed into commuter Forbes Park in Manila. Time has run out for these cities railways and link to rail terminals and often airports. to adopt a long-term transport plan. The high cost of investing in such systems stems not only from the need for new lines and stations but also How Cities of Different Size Can Tackle Transport from the high price of land in central cities. To minimize and Land Use such costs, Manila and Bangkok have begun to invest in Cities of different sizes often need to approach trans- overhead rail lines along main thoroughfares. Manila's 14- port, land use, and communications differently. One kilometer light-rail line, opened in 1984-85, is only now option for large cities-a technique that Japan has used SOUTHEAST ASIAN CITIES TACKLE TRANSPORT, COMMUNICATIONS, AND LAND USE I67 Table 19.1 grate major infrastructure projects with local needs and Urban Agglomerations by Gross Regional Domestic plans, especially in densely populated neighborhoods Product, 1995 with low motor vehicle ownership. Integrated planning for transportation and land use Current int'l GRP may be one way out of megacities' transportation crises. US$billion adjusted Urban corridors such as the Kelang Valley, which com- for Purchasing National Power Parity total bine infrastructure, manufacturing, and both middle- (PPP) (percent) class and worker housing, are likely to generate the SINGAPORE 68.3 100.0 greatest benefits, especially if seaport and airport are lo- cated along or adjacent to the corridor. The Malaysian THAILAND 438.8 100.0 Bangkok metropolis 172.5 39.9 government has explicitly pursued such rational deci- Bangkok and vicinity sionmaking in planning the new information corridor. (5 provinces) 226.0 51.5 In Thailand, by contrast, the siting of the airport (Don INDONESIA 734.5 100.0 Muang) well north and the new deep-sea container port Jakarta DKI 112.4 15.3 (Laem Chabang) well south of Bangkok has made it par- Jabotabek 148.0 20.0 ticularly difficult for companies to decide where to lo- PHILIPPINES 195.5 100.0 cate (Rimmer 1995). Resiting the international airport Manila national capital region 63.5 32.5 to a new location southeast of Bangkok will facilitate in- dustrial concentration along the corridor from Bangkok MALAYSIA 181.3 100.0 down the Eastern Seaboard. Jakarta's airport and con- Kuala Lumpur (Wilayah) 23.5 13.0 tamer port also sit on opposite sides of the city center, but in this case completion of ring-road connections VIETNAM 17.6 100.0 moderates travel time and eliminates the substantial ad- Ho Chi Minh City 3.0 17.2 vantage of any one axis. Source: World Development Report 1997: The State in a Changing C World. World Bank, Oxford University Press 1997, and national Cities of 1 mllion or less, m contrast, typically do not statistical yearbooks. suffer any serious transport crisis. They are compact enough to be served by slow-speed road and ferry links without significant peak-hour congestion, even in down- town areas. Some have modest private bus networks, but with success-is to encourage private companies to fi- paratransit can make journeys from the urban fringe to nance infrastructure by relying on earnings from related the central business district cheaply and in reasonable development. For example, Japanese railroad companies time. Motorcycles are more important than motor cars, build their own department stores, hotels, and leisure while bicycles and pedestrian traffic often account for a facilities at downtown terminals, and create new towns, substantial share of trips. housing, shopping, and parks or resorts at outlying des- In these cities prestige rather than need often drives tinations. In Bangkok, Jakarta, and Manila in the 1980s investment in large infrastructure projects such as traf- and 1990s, in contrast, governments financed express- fic arteries and transport terminals. Overinvestment way construction but developers reaped the gains (and based on grants from central governments and multilat- losses) from induced investment in towns, malls, and eral loans is particularly noticeable in nearby airports, leisure facilities. which are often poorly sited in terms of efficient access. In addressing their transport problems, large cities As in larger cities, these major projects often disadvan- need to pay particular attention to the needs of the poor. tage pedestrian and slow-moving traffic. Ambitious bu- Very often decisions designed to relieve vehicular bot- reaucrats in medium-sized and smaller cities also often tlenecks have ignored the needs of slow-speed and needlessly interfere with paratransit. Large projects to pedestrian traffic. These decisions have forced nonmo- modernize transport systems generate both political torized rather than motorized traffic to make longer de- kudos and slush funds for the officials in charge. They tours, notwithstanding the considerable human energy have little incentive to investigate low-cost investments required to travel extra distances in a tropical climate. in better and more direct walkways, all-weather foot- Poor siting of bus stops and rail stations and design of paths, and dedicated bicycle lanes. overhead pedestrian bridges impose considerable incon- Governments of these cities need to focus more on venience and discomfort, and schoolchildren can no specific projects rather than overall infrastructure. Cen- longer safely ride bicycles to school. To address these tral governments can help by monitoring local projects problems, international lending agencies should inte- and holding officials accountable, and by earmarking i68 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION grants or loans for construction of sewerage systems. A nomic crisis has highlighted the weakness of highly cen- modern sewerage network is desirable on both public tralized decisionmaking and provoked domestic pres- health and environmental grounds, and can be designed sures for reform, especially in Indonesia. Local energies and constructed more cheaply when cities are still rela- and initiatives are clearly one of Southeast Asia's great tively small. And planners should note that the design untapped resources. of a sewerage system is as fundamental to long-term urban planning as transport networks. Bibliography Most urban governments, whether large or small, Crooks Michell Peacock Stewart. 1971. The Urban Renewal and find themselves caught between rising public needs and Development Project, Singapore: For the United Nations Devel- aspirations and their limited powers: These governments opment Programme (Special Fund). Singapore, Crooks Michell do not administer functional urban areas, have little na- Peacock Stewart Graphics and Printing Division (5 parts and tional political clout, and retain only limited access to summary volume). their tax base. During the property boom of the 1980s Dick, H., and Rimmer, P. J. 1998. "Beyond the Third World and 1990s, for example, public revenue captured very City: The New Urban Geography of Southeast Asia." Urban littleof the escalation of land prices, even where they Studies 35(12): 2303-21. little of the escalation of land prices,evenwherethey.Dick, H., and Rimmer, P. J. 1999. "Privatising Climate: First stemmed from investments in public infrastructure. A World Cities in the Third World." In East-West Perspectives good example was the huge increase in land values adja- on 21st Century Urban Development. Aldershot, Ashgate/ cent to Jakarta's urban ring roads built on what formerly Vancouver: University of British Columbia: 305-24. was mainly kampung land. Dick, H., and Rimmer, P. J. Forthcoming. Cities, Transport, and Dealing with this situation will entail changes in pub- Communications: The Economic Integration of Southeast Asia lic finance and the application of good government as since 1850. London: Macmillan. much fnase tansprth englineriong ofirst urb govern -t a Ginsburg, N., Koppel, B., and McGee, T. G. (eds.). 1991. The much as transport engineering. First, urban govern- Extededetr ) Extended Metropolis: Settlement Transition in Asia. Honolulu: ments, whether local or provincial, must oversee enough University of Hawaii Press: 3-25. area to allow sound planning and political accountabil- JUTS. 1997. Janes Urban Transport Systems 1997-98. Coulsdon, ity. Second, to finance infrastructure programs, urban Janes Information Group. governments need much greater leverage over their tax Malaysia. 1996. Seventh Malaysia Plan 1996-2000. Kuala base, especially growth in land values, and require sub- Lumpur: Economic Planning Unit. stantial borrowing powers. Rimmer, P. J. '1986. Rikisha to Rapid Transit: Urban Public Transport Systems and Policy in Southeast Asia. Sydney: Perga- Finally, local governments that receive matching mon Press. grants from the central government need to be account- Rimmer, P. J. 1995. "Urbanization Problems in Thailand's able to that government, but their relationship to local Rapidly Industrializing Economy." In Medhi Krongkaew, citizens and interest groups is even more important. Co- ed., Thailand's Industrialization and Its Consequences. Lon- operation with business makes decisionmaking more ef- don: Macmillan: 183-217. fective and promotes local investment and innovation, Sirat, M. 1998. "Producer Services and Growth Management of a while .. goorlaioshpitctienfciittMetropolitan Region: The Case of Kuala Lumpur, Malaysia." while a good relationship with citizens facilitates rev- Asia Pacific Viewpoint 39(2): 221-35. enue raising and public initiatives. In Southeast Asia World Bank. 1997. World Development Report 1997: The State such ideas may still seem radical, but the Asian eco- in a Changing World. New York: Oxford University Press. chapter Public Participation in Urban Environmental Management in Japan Ryo Fujikura Faculty of Economics Ritsumeikan University n Japan, farmers and fishermen have tailed heavy reliance on copper exports to obtain foreign traditionally managed natural re- currency. The country's large-scale copper refineries did sources, beginning with the cultiva- nothing to abate pollution, and the results around the tion of rice 2,000 to 3,000 years ago. Ashio mine were typical: Sulfur dioxide emissions killed Community rules strictly controlled trees on surrounding mountains while sludge and efflu- the amount of water introduced into ent polluted rice fields downstream. Farmers protested, each field, and farmers managed surrounding forests but the Imperial government transferred them to other and grasslands as common resources as well. Fishing co- areas (Ui 1985). Protests against air pollution from an- operatives similarly determined when people could fish, other copper refinery, which began operating in 1904 by what method, and for which species. on Shikoku Island and damaged rice and wheat crops, As Japan modernized during the late 19th and early ultimately yielded more results: The company moved to 20th centuries, farmers and fishermen protested eco- another island and finally solved its problems in 1939 nomic losses from industrial pollution, but the Imperial after experimenting with several flue gas treatment tech- government suppressed the movements. When Japan's nologies (Himi 1987). economy rapidly developed anew during the 1960s, In urban areas, such as Tokyo and Osaka, air pollution fishers and neighborhood associations again protested and noise from factories bothered citizens beginning in serious pollution, and victims of pollution-related dis- the prewar period, and their complaints prompted cen- eases filed suit against industrial enterprises. These pro- tral and local governments to take some action. However, tests, as well as those of other local groups, raised public governments regarded the problems individually-the awareness of environmental issues throughout Japan concept of environmental pollution did not exist-and and proved essential in turning local and central govern- citizens could not easily speak out against the Imperial ments away from a sole focus on economic development government. With the onset of the Japan-China war in at any cost. Public participation has since become the 1937, industrial production became a national priority, key not only to local urban planning and environmen- and the ability to tolerate pollution became a virtue. In- tal management, but also to the country's approach to dustry did adopt antipollution measures during the war global environmental sustainability. but only to increase productivity, avoid disaster, or en- hance military strength (Institute for Pollution Control The Birth of Local Protest Suis17) Studies 1970). The first industrial pollution in Japan stemmed from Citizens' Movements during the Postwar Period the Ashio copper mine and refinery in Tochigi Prefec- ture during the 1890s and the 1900s. To catch up to The primary object of the Japanese government after developed nations, the Imperial government had estab- World War II was reconstruction, and in 1960 the Cab- lished the Rich and Strong Empire Policy, which en- inet instituted the Income Doubling Plan, which aimed 169 170 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION to double per capita gross domestic product (GDP) in municipal assembly to oppose the project, which was 10 years. As iron and steel production, the leading in- abandoned. dustry, quadrupled, rapid industrialization caused seri- Central government officials had not expected that ous pollution in cities throughout Japan. But as ordi- local protests would prompt the city to abandon the proj- nary citizens began to afford TV sets, refrigerators, and ect. The government and big business worried that soon, washing machines-previously symbols of rich fami- if things went on this way, they would not be able to lies-most considered pollution an unavoidable side ef- build factories anywhere (Broadbent 1998). Thus, the fect of their new prosperity. central government established the Basic Law for Pollu- Fishing communities again became the first obvious tion Control in 1967, but it stipulated that pollution victims of this pollution, and when effluent from a control should be implemented "in harmony with sound paper factory killed a high percentage of fish at a fishery economic development." This halfhiearted law failed to in Tokyo Bay in 1958, fishermen asked for compensa- assign responsibility for conserving the environment to tion. When the company did not respond, angry fishers any national agency. But when the central government burst into the factory, prompting the government to hesitated to enforce substantive antipollution measures, pass two laws controlling water pollution in the same some local governments, often headed by socialists or year-although those could not prevent pollution.1 communists, known as "reformists," began to act. Fishing communities similarly protested water pollution As these reformists began to replace conservative gov- from large petrochemical complexes in Yokkaichi City ernors and mayors throughout Japan, Prime Minister when the catch began to smell of industrial effluent and Sato, the head of a conservative party, convened the plummeted from 17,000 tons in 1956 to 4,000 tons in "Pollution Control Diet" session in 1970. This session 1964. Fifteen fishing cooperatives requested compensa- passed 14 pollution-related bills, deleted the controver- tion from industry and the Mie Prefecture government sial requirement that pollution control essentially take a for falling prices in 1960, while neighborhood associa- back seat to economic development, and established the tions began to complain of noise, vibration, and offen- Environment Agency. The central government had fi- sive odors (International Center of Environmental Tech- nally changed its attitude toward the environment. The nology Transfer 1992). movement also reversed the direction of the justice sys- Meanwhile, in the mid-i 950s Minamata City saw ev- tem, as courts uniformly reached verdicts in favor of vic- idence of a strange illness known as Minamata disease, tims of pollution-related illnesses such as Itai Itai dis- whose cause was unknown-although fish caught in ease, Yokkaichi asthma, and Minamata disease in the Minamata Bay were suspected. When they could not sell early 1970s. Faced with no other choice, Japanese com- their fish and a research group of Kumamoto University panies began to invest intensively in pollution control concluded that organic mercury in fish was a possible during the mid- 1970s, and environmental quality sig- cause for the disease, fishermen held demonstrations nificantly improved. requesting compensation. Many members of fishing Overall, local people had at first kept silent because families suffered and died from the disease during the they considered pollution unavoidable and also hesitated decade the government took to identify organic mercury to speak out against the government and big business, in factory effluent as the culprit (Ui 1968; Social Scien- much as they had before the war. However, once they tific Study Group on Minamata Disease 1999). recognized that they suffered from pollution, they started The first successful citizens' movement that changed to protest and used new political institutions, such as the government's plan probably occurred in Mishima local elections and courts, to push the government to- and Numazu City, where the central government ward a pro-environment policy (Broadbent 1998). planned to establish a petrochemical complex in the The experience in Kitakyushu City reveals that 1960s. The governments of Shizuoka Prefecture and women often played a key role in training the spotlight both cities at first welcomed the project, but local peo- on the need to clean up severe industrial pollution. ple feared that the complex would produce the kind Women's associations have formed in Japan since the of air pollution Yokkaichi City had experienced. The prewar period, and during the Japan-China War and mayor of Mishima City convened a study group con- World War II, the central government relied on Patri- sisting of experts and high-school teachers whose report otic Women's Associations or Nation Defense Women's confirmed that the project would cause serious damage. Associations to enforce wartime policy. The women sup- Some 25,000 citizens-corresponding to one-third of ported injured soldiers and bereaved families, and facil- the city's eligible voters-demonstrated against the itated food rationing and obligatory household savings. plan. This show of force convinced the.mayor and the In order to help democratize Japan after the war, occu- PUBLIC PARTICIPATION IN URBAN ENVIRONMENTAL MANAGEMENT IN JAPAN I7I pation forces educated leaders of women's associations, that defined remedial action. This agreement was one of who studied the problems they faced and learned how the few concluded before companies throughout Japan to solve them. Because of women's efforts, the govern- accepted responsibility for controlling pollution a decade ment of Kitakyushu succeeded in improving the envi- later (Kitakyushu City 1998). ronment during the 1970s in close cooperation with in- Dust decreased in the late 1960s as the companies dustry, and is now trying to transfer its experience and substituted liquid for solid fuel, but airborne concentra- technology to developing nations. tion of sulfur dioxide grew. This time the Sanroku Dis- trict Association focused on health effects and enlisted Kitakyushu City: Women Take the Lead the help of a professor of medicine. The group found a Kitakyushu City became one of four major industrial relationship between air pollution and the number of ab- cities during postwar reconstruction and development. sent pupils from 1961 to 1964, uncovered 562 asthma Its economic development was based on the production or bronchitis patients among 817 families, and found of iron and steel, chemical products, and cement. Be- that many children suffered respiratory diseases when cause of these industries Kitakyushu suffered the worst they moved from rural areas to Kitakyushu. In 1967 the air quality in Japan. However, its citizens remained quiet association measured air pollution by exposing plates even while those in other industrial cities such as outside for one month, finding that the weight of iron Yokkaichi, Osaka, and Kawasaki sued the government plates increased significantly while that of tin and stain- and private enterprises. They regarded Kitakyushu not less steel plates dropped. The group presented their re- as a place to live but as a place to work: Factory workers sults to other citizens while showing an original 8 mm returned to their rural hometowns when they retired, film. Widespread media reports of these activities helped and hesitated to speak out against Nippon Steel and raise concern about pollution in the city (Hayashi 1971). other large companies. Meanwhile, the women endured pressure from the As more laborers began to raise their children in Ki- enterprises that employed their husbands-the families takyushu and remain after retirement, pollution aware- lived in company dormitories, and janitors monitored ness grew. Women found that they had to wipe the in- their activities. The association's leader received threat- terior of their houses daily because of smoke from the ening telephone calls. Guards did not allow anyone to factories, and that wash hung out to dry turned gray photograph the factory, even from outside. But during from oily dust. Children drew black and gray pictures the 1971 municipal election the conservative mayor, of their surroundings and asked mothers why suburban who had not professed interest in pollution previously, skies were blue; many became sick and were often ab- proposed a policy of pollution control. His rival, sup- sent from school. ported by the Communist Party, criticized the mayor's Women of the Tobata District Association in Ki- policy as favoring the enterprises and appealed for more takyushu, in particular, discovered that laundry hung stringent control. An intensive campaign by the business outside grew dirtier with proximity to a power plant community reelected the incumbent mayor, but many using low-quality coal. Because many of their husbands people who were not supporters of the Communist Party worked at the plant, they pressured the mayor and the voted for his rival. These results prompted Kitakyushu city assembly instead, and the plant installed a costly officials to reverse course in the 1970s, and enterprises precipitator to clean its emissions at the mayor's request. had no alternative but to clean up their act. City officials Another women's group, the Sanroku District Asso- and business leaders alike feared that a communist would ciation, monitored smoke and offensive odors from a become the next mayor and impose harsher regulation, factory manufacturing carbon black2 and coke, alerting as reformist mayors had done elsewhere. the municipal government and the factory when they Public Participation Today observed serious pollution and showing the mayor dirty wallpaper and floor cloth. When the situation did not Today, most participation in urban environmental man- improve after the mayor visited factory headquarters in agement is based on action-oriented activities and edu- Tokyo accompanied by local citizens in 1961, the asso- cation. In fiscal year 1998, for example, the Environ- ciation members turned to a professor of engineering for ment Agency of the Japanese Government supported help. The group of 500 women then measured airborne about 4,000 eco-clubs-voluntary clubs consisting of pollution by leaving white cloth and paper boxes out- some 70,000 elementary and junior high school pupils side and photographing smoke-emitting facilities. In throughout Japan that pursue environmental educa- 1963 the local people and factory owners concluded an tion-mainly by providing information and organizing agreement-mediated by prefecture and city officials- meetings (Environment Agency 1999). Municipalities 172 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION rely heavily on citizens' groups to support recycling ac- and business associations, and collected opinions from tivities, as citizens become more aware of the country's more than 400 citizens to further develop the frame- solid waste problems. In Tokyo's Minato Ward, home to work (Fujikura 1995). This process convinced the plan- numerous businesses and foreign embassies, over 160 ning committee to choose "creation of the environmen- groups collected 2,500 tons of recyclable waste in 1995 tally sound municipality" as their basic goal-in marked (Minato Ward 1996). contrast to other municipalities, which usually focus on On the other hand, media reports on endocrine dis- population, the economy, and land use. rupters also have prompted 221 local disputes regarding Kamakura City also relied on an environmental com- planned and existing waste-treatment facilities, as citi- mittee to prepare its Basic Environmental Plan (BMP). zens demand more information disclosure and partici- In Japan, the members of such committees are usually pation in decisionmaking on such facilities (Environ- selected by the government, which tends to exclude crit- ment Agency 1998). ics. Kamakura City advertised for committee members The 1992 Earth Summit prompted more opportu- and selected 5 representatives of different age groups nities for local involvement in environmental policy- and neighborhoods from among 60 applicants. Busi- making. Japan's Basic Environmental Law, enacted in nesses and citizen groups prepared materials for the November 1993, represented the first attempt by any committee-in contrast to other municipalities, where industrial nations to codify the concept of sustainable the government usually prepares such materials. The development. The Basic Environmental Plan, passed a city held two workshops and one symposium, attended year later after public hearings throughout the country, by several hundred citizens, to provide information on specifies comprehensive environmental goals for the the committee's discussions and feed citizens' opinions next 50 years and lays out the responsibilities of central back to the committee before it produced its final doc- and local governments, businesses, and citizens in ful- ument. The city used the same approach to prepare filling those goals. In particular, the plan identifies par- its Traffic Plan and Women's Plan, which facilitates ticipation as one of its basic concepts, facilitating public women's participation in society. participation in formulating policy of both the central In 1998 Kamakura City established two centers to and local governments. support the activities of local nonprofit organizations In 1993, moreover, the Urban Planning Law was working on social welfare, environmental management, amended to allow municipal governments to establish international cooperation, and urban development. master urban plans defining long-term development These centers, which provide meeting rooms, copying goals and specifying how they will reach them.3 Since and printing equipment, and information on the activi- then, public participation has been regarded as an indis- ties of other nonprofit organizations, are designed by pensable step in preparing not only master urban plans local people, funded by the municipal government, and but other local plans: Mayors must organize committees administered by volunteers. Support for such activities that include local citizens to prepare the plans, and they is rare in Japan, which tends to lack organized environ- must also hold public hearings. Such participation often mental groups and other nongovernmental organiza- remains nominal-but Kamakura City, 50 kilometers tions (NGOs). southwest of Tokyo, is an exception.4 Nongovernmental Organizations in Japan Kamakura City: The Public Role in Environmental Although neighborhood associations and cooperatives Decisionmaking have been instrumental in changing Japanese attitudes The Japanese capital during the 13th century, today toward pollution control, the role of NGOs has been Kamakura City is home to many industry executives limited (Japan Environment Association 1996). Al- and high-ranking government officials, and numerous though the activities of NGOs are growing, most work heritage sites and marine sports attract over 15 million locally-the country has almost no nationally influen- tourists annually. In the mid-1990s the municipal gov- tial environmental organizations. ernment adopted an innovative approach to public par- The Japanese system of lifelong, seniority-based em- ticipation in decisionmaking. ployment is one factor hindering the emergence of na- The government relied on over 200 citizens partici- tionwide NGOs. Although attitudes are changing, most pating in two symposia to devise the basic framework university graduates, particularly men, seek jobs at large for its new master urban plan. The government then enterprises or government agencies that assure long held 25 public hearings and over two dozen meetings term employment. More women brave short-term con- with representatives of local communities, cooperatives, tracts and work in foreign companies and international PUBLIC PARTICIPATION IN URBAN ENVIRONMENTAL MANAGEMENT IN JAPAN I73 organizations, partly because they find it difficult to ob- The Law Concerning Access to Information Held by tain jobs in large Japanese enterprises. Changing jobs in Administrative Organs was also established in 1999 and Japan remains difficult and is doubly so for a full-time will become effective within two years. The purpose of NGO staff member: Unlike in the United States, gov- the law is to strive for greater disclosure of information ernments do not recruit NGO staff as high-ranking of- from administrative organs. It also stipulates that local ficials. Full-time NGO staff are poorly paid and the jobs governments establish ordinances for the same purpose. lack prestige. Almost all prefecture governments and many munic- According to a survey by the Environment Agency, ipal governments have already established such ordi- 57 percent of NGOs lack full-time staff-most cannot nances, and most documents of prefecture governments afford them. Few foundations provide financial support are accessible. But citizens are demanding more infor- to Japanese NGOs; most rely largely on membership mation. For example, many citizens concerned about fees for revenue. And more than half have fewer than hazardous emissions from industrial waste treatment fa- 200 members; the Wild Bird Society, the largest NGO cilities have not been able to get hard data. Information in Japan, has some 50,000 members (Japan Environ- regarding emission from private facilities is not subject ment Association 1996). Because of these constraints, to disclosure. Japanese environmental NGOs have great difficulty be- Japan's experience suggests that democracy is in- coming as influential as those in the United States. dispensable in pushing for such laws and other aspects of On the other hand, volunteer activities are becoming environmental protection; the Imperial government suc- more popular and important. Neighborhood associa- cessfully quashed protests against pollution. International tions have traditionally cleaned towns, assisted vulnera- efforts to promote democratization thus are essential. ble and aged people, and reconstructed communities The Kitakyushu experience shows the importance of after disasters. Although many Japanese today do not public awareness and environmental education in build- have the time or inclination to join formal groups, a ing support for environmental protection. Many devel- number of young volunteers helped victims of the 1995 oping countries attempting to promote environmental earthquake disaster in Kobe City, and volunteers largely education lack trainers, teaching materials, and meth- removed oil spilled by a Russian tanker in the Sea of ods; international sharing of methods and materials Japan in 1997. According to a government survey, over would also support these efforts. half of Japanese are interested in volunteer activities Japan's experience reveals the importance of local gov- (Saotome 1995). More Japanese will undoubtedly par- ernments in controlling industrial pollution: Local lead- ticipate, but it is unclear whether influential NGOs will ers understand their communities better than the cen- grow from these largely project-oriented activities. tral government, and careful advice from local officials Many Japanese fear that NGOs will become too po- to individual polluters has solved many problems. Un- litical. Conversely, politicians often try to coopt envi- like the central government, local people elect local lead- ronmental NGOs, pretending to pay attention to them. ers directly, and they must therefore be sensitive to con- For their part, Japanese NGOs are not good at cooper- stituents' opinions. Local people can also more easily ating: Small differences often erode alliances. During participate in decisionmaking on a local rather than a the Kyoto Conference on Global Climate Change in national issue. December 1997, Japanese NGOs working on climate However, in many developing countries local gov- change did organize a nationwide network and appeal ernments exert little control over environmental man- for support from the nation and the media, but such ac- agement, and their capacity is weak compared with that tivities have been limited. of the central government and industrial enterprises. What's more, the central government often appoints Challxperienges forJapanandLessonsfrommunicipal leaders who do not understand the local situ- ation. Their staffs have little prestige and low incomes, The Japanese government has lagged behind other making it difficult to attract engineers and scientists developed nations and international organizations in knowledgeable about controlling pollution. transparency and accountability. However, the situation In countries that do not mandate local governments, has improved remarkably over the past few years. In international cooperation could help build the ability of 1999, the Cabinet adopted a procedure on public com- the central government to make sound environmental de- ment. Since then, all central government agencies invite cisions. Where independent local governments do exist, comments from the public and respond to them when such aid could greatly enhance their ability to incorpo- they want to introduce, amend, or abandon regulations. rate public participation into development decisions. I74 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION Bibliography Shoji, H. 1975. Nihon No Kogai (Pollution in Japan). Tokyo: Broadbent, J. 1998. Environmental Politics in Japan. Cambridge: Iwanami Shoten. Cambridge University Press. Soaal Scientfic Study Group on Minamata Disease. 1999. Mi- Environment Agency. 1998. Heisei 9 Nendo Kankyo Hakusho namatabyo No Higeikiwo Kurikaesanai Tameni (Not to Repeat (The Quality ofthe Environment in Japan, FY1997). Tokyo: Tragedy ofMinamata Disease). Minamata: National Institute Okurasho Insatsukyoku. for Minamata Disease. Environment Agency. 1999. Heisei 10 Nendo Kankyo Hakusho Ui, J. 1968. Kogai No Seijigaku (Politics of Pollution). Tokyo: (The Quality of the Environment in Japan, FY1998). Tokyo: Sanseido. Okurasho Insatsukyoku. Ui, J. 1985. Gijutsu To Sangyo Kogai (Technology and Industrial Fujikura, M. 1995. "Kankyo Kihon Keikaku Sakutei De Kuhu- Pollution). Tokyo: Tokyo Daigaku Shuppankai. sareta Atarashii Shiminsankaku No Katachi" (An Innovative Yokohama City. 1999. Press release and interview with an offi- Approach of Public Participation for Establishment of a Local cer of the municipal government of Yokohama City, July. Environmental Plan). Global Net, pp. 18-19. Tokyo: Global Environmental Forum. Notes Hayashi, E. 1971. Yahata No Kogai (Pollution in Yahata). Tokyo: Asahishimbun. 1. The two laws for water pollution control and the Soot and Himi, K. 1987. "Showa 30 Nenndai Kara 40 Nendai Ni Itaru Smoke Control Law enacted in 1962 did not prevent pollution Madeno Wagakuni Ni Okeru Taikiosen Taisakushi No because the government had not changed its pro-industry policy, Gaiyo" (An Outline of Air Pollution Control Measures Taken and the laws established effluent and emission standards only from the Mid-1950s until the Mid-I 960s). Kenko Eikyo Ni within "designated areas." Almost no regulation was enforced out- Tuiteno Kakono Taikiosen Deta No Seirini Kansuru Kenkyu side of the areas. The two laws and the Soot and Smoke Control (Study Report on Past Data Regarding Health Effects Caused by Law were replaced by the Water Quality Conservation Law and Air Pollution), pp. 1-30. Tokyo: Taikiosen Kenkyu Kyokai. the Air Pollution Control Law in 1970 and 1968, respectively. International Center of Environmental Technology Transfer 2. Carbon black is manufactured by heat decomposition of (ICETT). 1992. Yokkaichishi Kogai Kankyokaizen Heno Atumi oil and used as material for pigment, ink, and rubber products. (History of Pollution and Environment Improvement in 3. Before the amendment, only prefecture governments could Yokkaichi). Yokkaichi: ICETT. prepare their master urban plans. Institute for Pollution Control Studies, Tokyo Metropolitan 4. Yokohama City also took an innovative approach to pub- Government. 1970. Kogai To Toukyoto (Pollution and Tokyo). lic participation in planning a road project. After a questionnaire Tokyo Metropolitan Government. survey to 40,000 families regarding participation to decision- Japan Environment Association. 1996. Kannkyo NGO Soshiki making, municipal government presented citizens with several UneiJittai To Anketo Chosa Kekka (Study Report on the Status alternatives early in the process-including the option of "no Quo of the Management of Environmental NGOs). Tokyo: development"-and information on the advantages and the dis- Japan Environment Association. advantages of each plan. It organized a committee consisting of Kitakyushu City. 1998. Kitakyushushu Kogai Taisakushi (History 2 local experts, 7 representatives of local communities, and 12 ofPollution Control in Kitakyushu). Kitakyushu: Kitakyushu citizens selected from applicants in order to identify which alter- Municipal Government. native should be adopted. The committee made another ques- Minato Ward. 1996. Minato Eco-Life Plan. Tokyo: Minato tionnaire survey to 10,000 families regarding the alternatives. In Ward Government. 1999, it prepared a draft report on the best route. The munici- Oya, K. 1998. "Sinrin Rinya No Chiiki Shakai Kanri" (For- pal government will start to prepare its road plan based on the estry Management by Regional Societies). Asia No Kannkyo draft report in 2000. It took already seven years from the first Mondai (Environmental Problems in Asia) pp. 265-78. Tokyo: survey, and will take another some years until the completion of Toyokeizai. the projects. However, this is a test case: the municipal govern- Saotome, M. 1995. The Volunteer. Tokyo: International Devel- ment has not decided to use this approach for all development opment Journal. projects (Yokohama City 1999). chapter Coping with Municipal Solid Waste in Developing Countries David N. Beede U.S. Department of Commerce David E. Bloom School of Public Health Harvard University lthough recent environmental issues estimate on data from the World Resources Institute such as global warming and strato- and our own correlations between municipal solid spheric ozone depletion garner the waste, population size, and income per capita. Our fig- lion's share of public attention, mu- ures show that on average, each person generates two- nicipal solid waste-more colloqui- thirds of 1 kilogram of municipal solid waste each day- ally known as garbage-has been a finding broadly consistent with estimates using other with us for centuries. The implications of such waste for data and methods. the quality of life of people living in many developing Whether 1.5 billion metric tons is a lot depends on countries are also more urgent. one's perspective. That amount exceeds the annual Municipal solid waste typically encompasses all solid global production of wheat and rice. Yet if the waste oc- wastes except industrial, agricultural, and human wastes. cupied a 100-meter-deep landfill with the same density Local definitions may or may not include material such as U.S. garbage, it would cover only 28 square kilome- as tires, construction debris, and sludge and ashes from ters. Filling an area the size of Rhode Island-the small- waste treatment plants. Reliable data on the generation, est U.S. state and only 0.002 percent of the world's land recovery, and disposal of municipal solid waste are some- area-would take 100 years. what hard to come by, as most countries do not keep good records. And cross-country comparisons suffer T from a lack of common definitions and techniques for Not surprisingly, the amount of municipal solid waste estimating amounts. Most estimates are based on either is very sensitive to population size, growing at a one-to- a sampling technique or a materials-balance analysis. one relationship. Municipal solid waste also rises with Sampling techniques involve measuring municipal solid per capita income, although less than proportionately. waste generated by samples of households, and using Thus developing countries account for a disproportion- those to estimate the generation rate for the entire pop- ately high share of municipal solid waste relative to their ulation. Materials-balance analyses indirectly estimate income, while rich countries account for a dispropor- municipal solid waste generation by estimating annual tionately high share relative to their population. rates (by weight) of the consumption of nondurable The problem stems not so much from the amount of goods and the disposal of durable goods. Neither tech- waste but from its spatial distribution and growth rate. nique is very precise. Although nature has an inherent ability to assimilate Despite these challenges, we have estimated that the waste-depending on temperature, precipitation, the world generates some 1.5 billion metric tons of munici- physical characteristics of soil, the height of the water pal solid waste annually, and that this amount is grow- table, and the extent of forest cover-the amount of ing by a substantial 2.7 percent per year. We base this waste in urban areas exceeds nature's absorptive and re- 175 176 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION cycling capacities. And since urban population is grow- today's management practices will affect their health and ing rapidly-far faster than world population-the environmental quality. concentration of waste creates enormous health and en- vironmental problems. Poor public management and From Concern to Action transportation infrastructure compound these problems. People in developing countries are acutely aware of the Many cities in the developing world dump great municipal solid waste problem. In a 1992 Gallup sur- quantities of waste haphazardly on land or in uncovered vey conducted in 12 developing countries, including and unlined landfills atop porous soils. Inefficient col- Russia, Hungary, and Poland, 45 percent of respondents lection techniques exacerbate this problem. For exam- (weighted by population) considered inadequate sanita- ple, the practice in some cities of discarding food wastes tion and garbage disposal a "very serious" problem. That in the streets, from which it is collected by hand each was higher than the percentage of respondents who con- day, appears to be much less efficient than household sidered water, air, or soil problems, overcrowding, or too disposal in tightly closed bins combined with collection much noise "very serious." By contrast, only 16 percent two or three times a week. Using uncovered collection of respondents in 12 industrial countries rated solid trucks leads to significant spillage back onto streets. waste as a very serious problem. Such activities make public cleaning of streets and open However, translating concern into action in the face areas critically important. Given that the per ton cost of of enormous financial constraints is difficult. This is street cleaning in developing countries is estimated to compounded by the fact that unlike, say, specific health be between two and three times that of curbside collec- problems, where new vaccines, diagnostic tools, or drug tion (Cointreau-Levine 1994), using more costly collec- therapies can have powerful beneficial impacts, the mu- tion equipment that reduces spillage, such as covered nicipal solid waste problem has no technical magic bul- trucks, would likely be more efficient. let. Moreover, judging which municipal solid waste Because developing countries rarely separate long- management options are cost-beneficial is hampered by lived hazardous waste such as insecticides, solvents, bat- poor, incomplete, or inconsistent data (see, for example, teries, used car oil, and medical waste from the munici- Spencer 1994). pal waste stream, these substances commonly leach into The valuation of some costs and benefits for manag- the water table. Disposing of hazardous waste according ing municipal solid waste is relatively straightforward, to the standards set by developed countries may seem to including out-of-pocket collection and transport ex- be beyond the financial capacity of governments or the penses and revenues from the sale of recyclable materi- private sector in developing countries. Suppose, however, als, compost, and energy. However, other costs and ben- that codumping of municipal solid waste and hazardous efits are less obvious but may be key, such as the waste means that within 15 to 25 years commingled opportunity costs of land (for transfer stations, process- wastes must be handled at a facility that isolates or treats ing facilities, and landfills) and of household labor (es- the waste more thoroughly, and at much greater cost than pecially if households are expected to sort their waste or for municipal solid waste alone. Our calculations show transport it to a central collection point). Savings from that it would be highly cost-beneficial to segregate and avoided disposal costs when new technologies are im- treat or dispose of the hazardous waste at the more ex- plemented are also sometimes overlooked. pensive facility from the start (Beede and Bloom 1995). Other costs and benefits may be extremely difficult to Many communities also burn much municipal waste, value, such as changes in public health and in the aes- without emission controls, and make no attempt to cap- thetic quality of air, water, and land. Furthermore, valu- ture the resulting energy. Other communities dispose of ations of the costs and benefits of changing one aspect of municipal waste in waterways used for drinking and municipal solid waste management must reflect any im- cooking. The result is growing air and water pollution pacts on the costs and benefits of other aspects of man- and breeding grounds for disease-carrying rodents and aging municipal solid waste. For example, the cost of mosquitoes. Poorly managed municipal solid waste also producing agricultural-quality compost may fall sharply generates offensive and sometimes overwhelming odors. if households separate their compostable and noncom- Overall, free markets clearly do not work very well in postable wastes. Finally, cost and benefit valuations must promoting socially efficient controls on solid waste, as account for the value of resources over time, requiring delays seem easier and less costly, and dumpers and the choice of a discount rate, which is often less straight- burners do not usually have to pay a direct price for their forward than would appear at first. actions. And children and future generations do not The fact that much economic activity in developing have a vote or a say in the marketplace, even though countries is effectively outside the realm of state control COPING WITH MUNICIPAL SOLID WASTE IN DEVELOPING COUNTRIES 177 and tax authorities further complicates the ability of pol- use blowers to separate light materials, such as plastics, icymakers to make effective choices. In addition, munic- by weight), eddy-current separators to extract aluminum ipal managers in developing countries often lack the in- (magnets above a conveyor belt that induce an oppos- formation they need to make well-informed decisions.1 ing magnetic field in aluminum below and push it off Overall, developing countries face challenges in ad- into a separate bin), and "low-tech" conveyor belts that dressing the four main components of the problem of transport recyclable materials past workers who sort managing municipal solid waste: collecting the waste; them. transporting the waste; processing the waste to make it The least capital-intensive, and hence most labor- more disposable, to reduce its health threat, and possi- intensive, recycling systems in industrial countries re- bly even to draw value from it; and disposing of the quire that households sort and separate each type of re- residue. On the first two counts, many developing coun- cyclable material (paper, aluminum, steel cans, different tries lack the transport infrastructure needed to collect types of plastic, and glass by color). Workers may also sort and transport waste. Narrow and unpaved roads limit commingled recyclable materials as they collect them, the use of large and heavily loaded trucks, thus under- placing each type in its own compartment in the truck. mining economies of scale. Households may also transport separated recyclable ma- Three main techniques are available to process and terials to containers scattered throughout a community dispose of waste: landfills, incineration, and recycling or to staffed drop-off centers. Households may also take and composting. Landfills are expensive to design and beverage containers to buy-back centers, in the case of operate. In addition to liners and covers, they also re- deposit systems for bottles. Finally, even the least capital- quire systems for collecting and managing leachate, re- intensive recycling systems in the industrial countries use moving hazardous waste, and flaring or capturing the mechanical shredders, balers, and pulverizers to make methane that results from anaerobic decomposition. materials denser, thereby reducing transport costs-and Incineration is also very expensive. In many develop- rendering these systems relatively capital-intensive com- ing countries, the waste stream includes a high propor- pared with systems in developing countries. tion of organic matter, such as food trimmings. The water More labor-intensive collection and processing of re- content in such waste means that burning it properly re- cyclable materials are found throughout the developing quires adding fuel, raising the cost of processing the world. These efforts can entail households bringing their waste. Most developing countries cannot afford to pur- recyclables to redemption centers, and small-scale entre- chase and operate the capital-intensive (in both human preneurs going door-to-door to purchase recyclables. and physical terms) incinerators that industrial countries Labor-intensive recycling can also involve municipal have devised to minimize the emissions from burning workers and informal sector scavengers rummaging municipal solid waste. Recycling and composting may be through household wastes put out for collection, or better alternatives in most developing countries, as they scavengers sifting through municipal solid waste at also capitalize on the resource value of the waste. transfer stations and final dumpsites. Often the privately Efforts to collect and process recyclable materials run businesses that purchase, clean, sort, and sell recy- in industrial countries are considerably more capital- clables in bulk to other middlemen or directly to facto- intensive than in most developing countries.2 The most ries are also highly labor-intensive.3 capital-intensive method is mixed collection in which Composting will not fulfill its promise in developing municipal solid waste is collected and delivered to a fa- countries until both quality control and demand im- cility for mechanical extraction of recyclable materi- prove significantly, since farmers, landscapers, and other als, with the remainder often used to make fuels for potential users of compost typically insist on a product electricity-generating incinerators. A somewhat less that is free from bits of glass, plastics, metals, and toxic capital-intensive system is the separate collection of old materials. Eliminating subsidies for chemical fertilizers- newspapers and commingled glass, metal, and plastic common throughout the developing world-would help materials, plus other waste. Generally this method re- make compost more competitive and thus raise demand. quires special trucks that have two compartments, one for newspapers and the other for other recyclable mate- Paying for Proper Waste Management rials. Households and firms perform the initial separa- Many economic schemes can help municipal officials tion of recyclable materials, and the process is refined at pay for proper management of solid waste and cut the materials-recovery facilities of varying capital intensity. amount illicitly dumped into waterways and onto road- These can include highly automated systems that use sides and open fields. However, these schemes often in- magnets to extract ferrous metals, air classifiers (which volve tradeoffs.4 178 LOCAL DYNAMICS IN AN ERA OF GLOBALIZATION For example, governments can provide municipal likelihood that private municipal solid waste manage- solid waste collection, transport, processing, and dis- ment will be more efficient than public management. posal services, charging either a flat fee per household Competition under the vigilant eyes of government or a fee per unit volume of waste collected. Besides fi- regulators and private citizens can also achieve appropri- nancing such services, a fee per unit volume creates in- ate economies of scale in solid waste management. In centives for households to reduce their waste by encour- particular, governments can achieve operational effi- aging recycling and altering consumption. Setting a ciency by allowing free entry into and exit from relatively per-unit fee equal to the net marginal social cost of man- labor-intensive activities (such as curbside collection of aging municipal solid waste may lead to a more efficient solid waste), and by holding open competitions for the outcome than charging a flat fee that effectively sets the exclusive rights to operate large-scale, capital-intensive marginal cost to households for managing municipal activities (such as landfill and incineration facilities). solid waste at zero. However, per-unit fees may also encourage house- Conclusion holds to deal with their municipal solid wastes by dump- There are no sure-fire solutions to the municipal solid ing or burning them illegally, or by placing them into the waste problems of developing countries. Nevertheless, dumpsters of other households and businesses. Because municipal solid waste is more a quality-of-life problem of such undesirable results, efficiency may be higher if than a major impediment to economic growth,5 and the per-unit fees are set lower than the marginal social cost one-two-three punch of economic growth, poverty re- of managing municipal solid wastes-as low as zero, as duction, and good governance will naturally enhance in the flat-fee option, particularly if a per-unit fee pro- countries' capacity to address it. Income growth can gram costs more to administer than a flat-fee program. help because garbage grows less than proportionately For items such as batteries, paints, and motor oil with income, and countries with rising per capita in- whose burning or dumping are especially damaging to comes can afford, and will typically desire, to spend the environment and whose codisposal with nonhaz- more to address waste. The ratio of services to goods also ardous municipal solid waste would drive up the cost of tends to rise with income, and services tend not to be managing the entire solid waste stream, a deposit-refund very waste-intensive. Finally, good governance is impor- mechanism may be more efficient than per-unit fees. tant because waste management is generally provided Under a deposit-refund mechanism, households and and financed publicly, or at least publicly regulated. firms pay a fee when they buy certain goods, which is re- funded when they dispose of or recycle the associated Bibliography waste acceptably. The advantage of deposit-refund mech- Beede, David N., and David E. Bloom. 1995. "The Economics anisms is that they provide households and firms with an of Municipal Solid Waste." The World Bank Research Ob- incentive not to illegally dump or burn certain parts serverAugust: 113-50. Bloom, David E. 1995. "International Public Opinion on the of the municipal solid waste stream, and they may be Environment." Science269: 354-58. cheaper to manage than monitoring the behavior of Bennett, Christopher P. A., David E. Bloom, Lakshmi Reddy many small-scale illegal dumpers. Bloom, Marguerite S. Robinson, and Michael H. Simpson. Some governments use competitive bidding to con- 1993. EnterprisesfortheRecyclingand CompostingofMunici- tract out waste management and then monitor the qual- pal Solid Waste. Jakarta, Indonesia: Center for Policy and Im- plementation Studies. ity of service, penalizing or canceling the contract of any Cointreau-Levine, Sandra J. 1994. Private Sector Participation in trash hauler or processor that does not perform to stan- Municipal Solid Waste Services in Developing Countries, vol. 1, dard. Profit-seeking firms generally have both greater The Formal Sector. Washington, D.C.: World Bank. flexibility and incentives than government bureaucra- Kinnaman, Thomas C., and Don Fullerton. 1999. "The Eco- cies to redeploy workers and physical capital quickly in nomics of Residential Solid Waste Management." National Bureau of Economic Research Working Paper 7326 (August). response to changing circumstances, and to design and Kreith, Frank, ed. 1994. Handbook of Solid Waste Management. implement cost-cutting innovations. The better able the New York: McGraw-Hill. government is to specify the tasks it seeks to accomplish, Rosenberg, Larry, ed. 1996. International Source Book on Envi- including setting standards for environmentally sound ronmentally Sound Technologies for Municipal Solid Waste practices, the more easily it can evaluate contractor per- Management. Osaka: United Nations Environment Pro- gramme International Environmental Technology Centre. formance. The more effective the government is at Spencer, David B. 1994. "Recycling, Part E." In Frank Kreith, implementing a mechanism for penalizing poor per- ed., Handbook of Solid Waste Management. New York: formance and monopolistic behavior, the greater the McGraw-Hill. COPING WITH MUNICIPAL SOLID WASTE IN DEVELOPING COUNTRIES 179 World Bank. 1997. "Turkey-Antalya Solid Waste Manage- 1. The U.N. Environment Programme's recent source book ment." World Bank Project Information Document for Proj- provides a good summary of current practices in both industrial ect Number TR-PA-40865 (March). and developing countries, along with recommendations. See World Bank. 1999a. "Dominican Republic-Solid Waste Rosenberg 1996. Management in Tourism Centers." World Bank Project 2. See Spencer 1994 for a description of recycling systems in Information Document for Project Number DOPE59511 industrial countries, and Kreith 1994 for an excellent reference (anuary 6). on the technology and economics of capital-intensive recycling World Bank. 1999b. "West Bank and Gaza-Solid Waste and systems. Environmental Management Project." World Bank Project 3. For example, see Bennett and others, 1992, for descriptions Information Document for Project XOPE54051 (uly 26). of the recycling industry in Jakarta. Notes 4. Kinnaman and Fullerton 1999 review economic models and empirical evidence on the effectiveness of a wide range of in- Larry Rosenberg provided helpful comments on an earlier draft centive schemes for optimal municipal solid waster management. of this chapter. The views expressed in this article do not neces- 5. Poor management of municipal solid waster management sarily reflect the views of the U.S. Department of Commerce or projects in the Dominican Republic, Turkey, and the West Bank the United States Government. and Gaza (see World Bank 1997, 1999a, and 1999b). ven as globalization annihilates distance and brings countries closer together, pressure for local autonomy and a shift toward urbanization are diffusing economic and political power within countries. These changes are transforming the role of governments and enlarging the involvement of private and nongovernmental participants. The number of critical decisionmakers will multiply, calling for reform and the creation of new modes of governance, including institutions to harness efforts aimed at raising welfare. One possible outcome of this dynamic is a shift away from a central state- led view of development toward a more complex, and potentially untidy, coalition of actors who will typically not confine themselves to operating in a single arena. A second implication is that the combination of local dynamics and global forces will catalyze the development of regions and cities. To the extent that such development does or does not occur, it may well widen the gap between the richer and poorer regions of the world. An outgrowth of the World Development Report 1999/2000, Local Dynamics in an Era of Globalization is a collection of papers by leading authorities discussing some of the major aspects of decentralization and urban change in the context of globalization. These scholars have drawn upon the experiences of developing and industrial countries from every continent. An essential reference for all those interested in one of the most important issues of the new millennium, this volume will be of special interest to governments, nongovernmental organizations, the private sector, economists, and those in the international development community. The World Bank 1818 H Street N.W. Washington, D.C. 20433 USA Telephone: (202) 477-1234 Facsimile: (202) 477-6391 Telex: MCI 64145 WORLDBANK MCI 248423 WORLDBANK Internet: www.worldbank.org E-mail: books@worldbank.org * .l Oxford University Press ISBN 0-19-521597-4