WPS5168 Policy Research Working Paper 5168 Developing Countries, Dispute Settlement, and the Advisory Centre on WTO Law Chad P. Bown Rachel McCulloch The World Bank Development Research Group Trade and Integration Team January 2010 Policy Research Working Paper 5168 Abstract Critical appraisals of the current and potential benefits as developed country markets, and the prevalence of from developing country engagement in the World disputes targeting highly observable causes of lost foreign Trade Organization (WTO) focus mainly on the Doha market access, such as antidumping, countervailing Round of negotiations. This paper examines developing duties, and safeguards. The paper also examines potential country participation in the WTO dispute settlement impacts of the Advisory Centre on WTO Law (ACWL) system to enforce foreign market access rights already into the WTO system in 2001. A close look at the data negotiated in earlier multilateral rounds. The dispute data reveals evidence on at least three channels through which from 1995 through 2008 reveal three notable trends: the ACWL may be enhancing developing countries' developing countries' sustained rate of self-enforcement ability to self-enforce foreign market access: increased actions despite declining use of the Dispute Settlement initiation of sole-complainant cases, more extensive Understanding (DSU) by developed countries, pursuit of the DSU legal process for any given case, and developing countries' increased use of the DSU to self- initiation of disputes over smaller values of lost trade. enforce their access to the markets of developing as well This paper--a product of the Trade and Integration Team, Development Research Group--is part of a larger effort in the department to evaluate the impact that international institutions have on the market access opportunities of developing countries. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be contacted at cbown@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Developing Countries, Dispute Settlement, and the Advisory Centre on WTO Law Chad P. Bown Rachel McCulloch The World Bank Brandeis University JEL No. F13 Keywords: WTO, dispute settlement, developing countries, ACWL ________________ Bown: Development Research Group, Trade and International Integration; The World Bank, 1818 H Street, NW, MSN MC3- 303, Washington, DC 20433 USA. tel: +1.202.473.9588, fax: +1.202.522.1159, email: cbown@worldbank.org, web: http://econ.worldbank.org/staff/cbown. McCulloch: Department of Economics and International Business School, MS021, Brandeis University, Waltham, MA 02454- 9110, USA tel: +1.781.736.2245, fax: +1.781.736.2269, email: mcculloch@brandeis.edu, web: www.brandeis.edu/~rmccullo/. Bown gratefully acknowledges the Hewlett Foundation and Brandeis University for financial support. For helpful comments and discussions, the authors thank Niall Meagher, Bernard Hoekman, Manfred Elsig, Håkan Nordström, Greg Shaffer, Robert Staiger, Kyle Bagwell, Alan Sykes, Judith Goldstein, Michael Tomz, Hunter Nottage, Thomas Sebastian, Fernando Piérola and seminar participants at Stanford University. Any opinions expressed in this paper are the authors' and should not be attributed to the World Bank. All remaining errors are our own. 1. Introduction The original 23 founding members (officially "Contracting Parties") of the General Agreement on Tariffs and Trade (GATT) had swelled to 91 by September 1986, when the Uruguay Round negotiations began. Notwithstanding the GATT's early reputation as a "rich man's club," by 1986 the majority of the members were poor countries, including many newly independent African nations. Still more poor nations joined the GATT during the protracted negotiations that produced the World Trade Organization (WTO). The WTO opened its doors on January 1, 1995 with 128 members. By July 2008, an overwhelming majority of the 153 WTO members were developing countries, with 32 of the poorest classified as least developed countries (LDCs). Yet many observers, and especially those representing the interests of poor countries, judge that participation in the Uruguay Round and in the WTO have so far yielded few benefits for these countries. Of the accomplishments from the Uruguay Round, the eagerly sought dismantling of the Multifibre Arrangement (MFA) has been a major disappointment, as quota rents dissipated and China's share of export markets exploded. The Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) is widely seen as causing, at least potentially, an adverse movement in the terms of trade of poorer countries, which are overwhelmingly importers of proprietary technologies created mainly in a few rich countries. Promised elimination of U.S. and European Community (EC) agricultural subsidies has stalled, disappointing middle-income developing countries with comparative advantage in sugar, rice, cotton, soybeans, and other agricultural products. And the Doha Development Round, aimed specifically at addressing concerns of poor countries, has been declared dead on several occasions. Almost completely overshadowed by laments regarding lack of progress in the Doha Round is the increasing benefit derived by developing countries from another achievement of the Uruguay Round: creation via the Dispute Settlement Understanding (DSU) of an enhanced process that allows members to self-enforce the market access to which their trading partners have agreed. This paper begins by presenting historical data on WTO dispute settlement that document self- enforcement activities of developing countries. Viewed from the perspective of developing country 2 participation, the data on WTO disputes over the 1995-2008 period reveal three interesting patterns. First, the data show a steady trend of WTO self-enforcement actions undertaken by developing countries throughout the WTO era, in strong contrast to the declining trend of self-enforcement actions undertaken by the developed countries over the same period. Second, while we show developing countries to be interested in enforcing the WTO commitments of their rich (especially U.S. and EC) trading partners, they are also increasingly interested in enforcing the WTO commitments of other developing countries. Finally, the data show that developing countries focus their self-enforcement actions on types of WTO violations that are more directly observable to exporting firms and their government policymaker representatives, especially antidumping. While some of this can be attributed to the global proliferation in this particular form of import protection, the pattern may also reflect the higher cost of identifying other types of violations that may also result in loss of market access, such as illegal subsidies to competing products and domestic regulatory barriers. Many developing countries point to the high cost of self-enforcement, and several have proposed that the WTO should bear all costs associated with the efforts of developing countries to enforce their market access rights. In recognition of the need for this type of assistance, a group of nations established the Advisory Centre on WTO Law (ACWL) in 2001 to help overcome one particular obstacle to developing country use of WTO dispute settlement ­ the high cost of litigation. The second half of the paper analyzes how the availability of ACWL assistance has affected the WTO enforcement efforts of developing countries. The paper thus contributes to a growing literature that examines the obstacles confronting developing countries as they use WTO dispute settlement to self-enforce the market-access commitments of their trading partners.1 1 Scholars who have studied WTO enforcement efforts of developing countries include economists, political scientists, and legal scholars, e.g., Bown (2005), Bown and Hoekman (2005, 2008), Horn, Mavroidis and Nordström (2005), Shaffer (2003, 2006), Nordström and Shaffer (2008), Busch and Reinhardt (2003), Busch, Reinhardt and Shaffer (2008), and Davis and Bermeo (forthcoming). See Bown (forthcoming) for a synthesis of the literature from an economic perspective. 3 We carry out a detailed examination of data on developing countries' use of the ACWL, and we also make comparisons to cases from countries that could have used the ACWL but chose not to do so. Our initial examination of the data on ACWL involvement in WTO dispute settlement indicates that the ACWL caseload largely mirrors the full WTO caseload involving developing countries. Notably, the ACWL has assisted developing countries in their self-enforcement efforts with respect to both developed country and developing country markets. As a second step, we use economic theory to identify ways that introduction of the ACWL might be expected to affect the developing country dispute initiation and prosecution caseload. We then examine the data for evidence that the ACWL may be having such effects. For example, introduction of the ACWL can lower the cost to a developing country of pursuing a WTO dispute, and this may affect the observed country-level DSU caseload pattern through changes at two different margins. The ACWL may affect the DSU caseload through the intensive margin, i.e., lower enforcement costs to countries using ACWL services may lead to more WTO disputes brought forward by the same countries that have brought disputes forward previously. The ACWL may also affect the pattern of DSU cases at the country-level extensive margin, i.e., new complainants without a history of using WTO dispute settlement may begin using the DSU for the first time. The limited evidence available from use of the ACWL from 2001 through 2008 suggests that the ACWL has affected the volume of disputes almost entirely through the intensive margin: the same developing countries have been making greater use of the DSU process. Given this result, we examine the data on these DSU-using developing countries to determine how the establishment of the ACWL may have affected the way they use WTO enforcement. The lower cost of access to WTO enforcement could lead developing countries to pursue the same types of disputes as they have in the past, but with differences in the way the cases are pursued. Alternatively, they may pursue different types of cases. We present evidence from three channels through which the ACWL may affect developing country use of WTO self-enforcement. First, we provide evidence that, by lowering the cost to the developing countries of litigation, the ACWL allows countries to file more sole-complainant disputes on 4 behalf of their exporters. Second, we provide evidence that the ACWL allows developing countries to pursue the DSU legal process more fully in support of a given market access enforcement interest. Third, we provide preliminary evidence of a "scale" effect, i.e., that the impact of the ACWL may operate through the size of the market access at stake in enforcement actions. Specifically, the ACWL-backed disputes involve smaller amounts of trade relative to otherwise similar non-ACWL-backed cases. Based on these results, we highlight an additional need of developing countries for support of their WTO self-enforcement efforts. The data suggest that when developing countries have good information regarding a foreign market access violation, they are able to pursue it through the DSU process. Establishment of the ACWL has largely overcome the obstacle posed by the high cost of the legal assistance required for WTO litigation. But inadequate information about possible violations of their market access rights remains an important obstacle that exporters in developing countries face in their WTO enforcement efforts. There is a clear role for further assistance, whether from the private or public sector, through monitoring of policy changes that affect market access of developing country exporters. Informing firms and their governments when their WTO rights may have been violated can strengthen the WTO's self-enforcement mechanism and thus offer greater market-access security to smaller exporters in poor countries. 2. The Data on Developing Countries and WTO Dispute Settlement During the period from 1995 through 2008, WTO members initiated 388 formal disputes via requests for consultations.2 Viewed in historical perspective, WTO member countries initiated almost 50 per cent more disputes during the WTO's first fourteen years than the GATT Contracting Parties did over the 2 In this section we rely on two sources of data. The first is fundamental data on WTO dispute initiations through the end of 2008. These data are available on the WTO website. The second source is a comprehensive database on WTO disputes through 2006. These data, available on the World Bank website, were compiled in a project organized by Horn and Mavroidis (2008) with funding from the World Bank. 5 entire 48-year GATT era, 1947 through 1994.3 This increase may reflect perceived improvements in the dispute resolution process and thus a higher probability of getting results. To establish a consistent accounting unit for characterizing disputes, we follow one strand of the literature and break disputes with multiple complainants into bilateral country pairs. For example, restate the U.S. ­ Shrimp dispute (DS58), initiated by the combination of India, Malaysia, Pakistan, and Thailand and recorded in the WTO dispute-initiation data as a single dispute, as four bilateral disputes.4 The basic list of 388 requests for consultations over the 1995-2008 period can be characterized as 415 bilateral disputes between pairs of WTO member countries.5 Figure 1 illustrates WTO dispute initiations by year. In Figure 1 and our subsequent analysis, we break the 1995-2008 period into two subperiods, reflecting a conspicuous break in dispute activity between the two. While the 1995-2000 period immediately following conclusion of the Uruguay Round averaged 41 disputes initiated per year, the 2001-2008 period averaged only 21 newly initiated cases per year ­ half the pace of the years immediately after founding of the WTO. At least three factors may have contributed to the sharp decrease in use of WTO dispute settlement after 2000. The first is that the conclusion of the Uruguay Round left some negotiating issues unresolved. Heavy initial use of the dispute settlement process in the early WTO era may reflect members' efforts to address some of these lingering issues through a judicial 3 GATT contracting parties initiated a total of 254 disputes from 1947 through 1994 (Bown (2002, table 1). Unlike formal WTO disputes, which are governed by the Single Undertaking, GATT disputes were not centralized. Although most GATT disputes were initiated and reported under Article XXIII, some were brought forward under plurilateral codes signed by only a subset of GATT members, especially the Tokyo Round codes on subsidies and antidumping. Bown's data are based on WTO (1995, 1997) and Hudec (1993). Hudec includes some GATT disputes substantively equivalent to Article XXIII disputes but initiated outside the prescribed Article XXIII channels (e.g., at GATT ministerials). 4 Even when the record shows multiple complainants taking on a single respondent over the same issue, this does not mean that the countries coordinated their actions in advance of dispute initiation. One important example is the WTO dispute over the 2002 U.S. steel safeguard import restriction. In this case, the DSU initiation data actually include nine separate U.S. ­ Steel Safeguard requests for consultations. 5 Some bilateral disputes are, in fact, initiated more than once. For example, the first (DS16) WTO initiation of what ultimately turned into the EC ­ Banana III dispute brought forward by Guatemala, Honduras, Mexico and the U.S., was in September 1995; the second initiation (DS27) brought forward by those same four complainants along with Ecuador, took place in February 1996. Empirical research that focuses on the outcomes of WTO dispute settlement typically sorts disputes to eliminate redundant entries. Since we are focusing here only on the broad pattern of the initiation over time, we simply note the existence of some redundant disputes. 6 process. A second contributing factor may have been members' initial lack of familiarity with the new dispute settlement system. In some instances countries initiated cases but did not follow through, perhaps to "test" the system to see if they could get something for nothing. In other instances complainants were forced to (re)initiate the same dispute, again suggesting that part of the high volume may be attributable to learning of the new procedures. The final and perhaps most important factor is the acceleration of export growth that began around 2001. As Figure 2 shows, 2001-2006 was a period of substantial export growth across much of the WTO membership. Broad acceleration of exports would have undercut members' ability to establish a loss of expected market access attributable to a trading partner's imposition of a WTO-inconsistent policy, and thus reduced the likelihood of complainant success. In the remainder of section 2, we examine the WTO dispute-initiation data more closely, looking in turn at the distribution of dispute initiation across developing member countries, across sectors of commercial activity, and across types of trade barriers at issue. 2.1 DSU involvement of WTO members Table 1 documents the frequency with which individual WTO members have been involved in WTO disputes as complainants, respondents, and interested third parties during this period.6 Although the U.S. and EC were the most frequent litigants, other industrialized countries (Japan, Canada and Korea) and some developing countries (Argentina, Brazil, Chile, India, Mexico, and Thailand) each initiated ten or more disputes during this era. Figure 3 breaks down the data on dispute-settlement initiation by category of complainant country over the two time periods. Together the U.S. and EC initiated an average of 20 new disputes per year during the 1995-2000 period--as many as all other WTO members combined. However, the U.S. and EC together averaged fewer than six newly initiated disputes per year during the 2001-2008 period, which was less than a third of their yearly average for 1995-2000. Although dispute initiation also 6 Throughout this paper we sometimes choose to separate data associated with U.S. and EC use of WTO dispute settlement from that of "other industrialized countries." When it is not necessary to do so, we refer to group of countries including the U.S., EC, and other industrialized countries collectively as "developed countries." 7 declined for other country groups, the change was less dramatic. Yearly average total initiations for other industrialized countries dropped from 7.5 to 4.6, while the yearly average for developing country initiations fell only from 13 to 10.8 new disputes per year. Thus, developing country use of WTO dispute settlement increased relative to the use by developed countries when comparing 2001-2008 versus 1995-2000. Table 2 allocates disputes based on which complainant country challenged which respondent during the two periods. In the 1995-2000 period, the U.S. and EC together initiated half (123 of 246) of all bilateral dispute pairs. Overall, 19.5% of the entire WTO caseload during the immediate post- Uruguay-Round era consisted of the U.S. and EC challenging each other. In comparison, only 9% (15 out of 169) of the 2001-2008 caseload of initiated cases were ones in which the U.S. or EC were challenging each other. Developing country complainants, on the other hand, tended to split their disputes to target either the U.S. or the EC, or another developing country (Table 2). In 1995-2000, almost 58% (45 of 78) of developing country disputes targeted either the U.S. or EC, while 40% (31 of 78) targeted another developing country. In 2001-2008, 49% (42 of 86) of developing country disputes targeted either the U.S. or EC, while 47% targeted another developing country. Only infrequently do developing country complainants target other industrialized countries apart from the U.S. or EC. When developing countries are named as respondents, which complainants self-enforce their WTO commitments? During 1995-2000, developing countries were challenged 90 times, and exactly 50% (45 out of 90) of these cases were initiated by either the U.S. or the EC, with another 34% (31 out of 90) initiated by other developing countries. But developing countries increased their role in WTO self- enforcement activities after 2001. Of 67 disputes against developing countries initiated by WTO members in 2001-2008, 60% (40 out of 67) were initiated by developing country complainants and fewer than 36% (24 out of 67) were initiated by the U.S. or EC. To underscore the increased role that developing countries have been taking in self-enforcement of other developing countries' market access commitments, it is noteworthy that most of the cases 8 initiated by the U.S. and EC to enforce developing country WTO commitments in 2001-2008 were focused on just two countries: India and China. Specifically, of the 24 cases that the U.S. or EC initiated against developing countries during this period, nine were initiated against China alone (even though China had not acceded to the WTO until late 2001), and five were initiated against India. 2.2 The commercial sectors under dispute Are countries disputing mainly over commitments made as part of the Uruguay Round "Grand Bargain" (Ostry, 2002), i.e., are developed countries self-enforcing their foreign market access over TRIPS and services and developing countries self-enforcing the foreign market access promised to them in textiles and apparel products and agriculture? To answer this question, Figure 4 assigns each bilateral dispute to one commercial sector. 7 Over 50% of all disputes initiated by developing countries (84 out of 164) involve the enforcement of market access in agriculture, beverages or seafood products. This category is also large for developed countries (in particular, major developed economy exporters of certain agricultural products, like the U.S. and Cairns group members Australia, Canada, and New Zealand), though it represents a smaller share of their overall dispute-initiation caseload (85 out of 251). Other sectors of importance for disputes involving developing countries include textiles and apparel, steel, and other manufacturing. As expected, the vast majority of disputes in R&D-intensive or intellectual property (IP)-intensive sectors ­ e.g., pharmaceuticals, information technology, telecommunications, and media ­ have been initiated by developed countries. Developed countries have also been initiators of disputes involving capital-intensive industries such as autos, aircraft, and shipbuilding. One surprising feature of the data is the absence of expected lumpiness over time in the distribution of disputes initiated by certain sectors (Figure 5). In particular, the back-loaded phase-in of Uruguay Round commitments on apparel and textiles under the Agreement on Textiles and Clothing 7 The "not classifiable" category consists of disputes not readily associated with any particular industry. Such disputes may challenge a country's broader law, and not necessarily how that law was applied to any particular industry. The challenge may involve many products and industries and thus cannot be assigned to a single category. 9 (ATC), culminating in the elimination of the Multi-Fibre Arrangement (MFA) in 2005, might have been expected to result in a clustering of disputes in this sector around 2005. However, this was not the case. In fact, most of the disputes over textiles and apparel products were initiated in the 1995-2000 period, e.g., challenges to the U.S. use of the transitional safeguard for clothing and apparel available under the ATC. A possible interpretation of the relative absence of disputes over apparel and textiles since 2005 is that the U.S. and EC have managed to live up to their import market access commitments vis-ā-vis developing countries, with the notable exception of China. The U.S. and EC both negotiated voluntary export restraints in textiles and apparel products that restricted Chinese export growth over 2005-2008, thus preserving some foreign market access that other developing countries might have anticipated losing. Similarly, in the case of agriculture, the Uruguay Round Agreement on Agriculture contained a negotiated "Peace Clause" (Steinberg and Josling, 2003) designed to limit formal dispute-settlement activity in the sector ­ provided certain economic conditions were met ­ until the end of 2003. Yet there is no evidence of a sharp increase in disputes over agriculture in 2004 following expiration of the Peace Clause. The TRIPS Agreement included different phase-in periods for different groups of countries, with developed countries receiving one year in which to come into compliance, developing and transition economies five years, and least developed countries 11 years, subsequently extended to 21 years for pharmaceutical patents. Yet there is not much evidence of the clustering of WTO enforcement activity associated with enforcement in IP-intensive sectors that might have been expected for the developing and least-developed countries in 2000 and 2006. In fact, the only clustering of IP-related cases occurred in 1996-1998, when developed countries were required to become TRIPS-compliant. These data suggest that developed countries have not given priority to using WTO dispute settlement to self-enforce their TRIPS-related interests with regard to developing-country market access. Although there is evidence of temporal clustering of disputes over the steel sector, this pattern is the result of underlying trade policy activity (use of antidumping and safeguards) in that market rather than the phase-in period for any commitments made in the Uruguay Round. Over 50% (22 of 37) of all 10 WTO disputes initiated over steel products took place in just two years ­ 2000 and 2002. The 2002 cluster of steel disputes came in response to the U.S. imposition of import safeguards in 2002. Similarly, the 2000 cluster is largely attributable to U.S. and EC trade restrictions on steel products imposed that year and subsequently challenged by other members. As the example of steel disputes illustrates, dispute- initiation activity tends to reflect underlying policy volatility, and the implementation of WTO-illegal measures is unlikely to be distributed uniformly over time and across countries. 2.3 The causes of lost foreign market access under dispute We have already suggested that lack of information may be a particular obstacle faced by developing countries in self-enforcing their market access rights. Here we develop an indicator of the "observability" to an exporting firm of the underlying cause of lost foreign market access. We then examine whether developing countries are more likely to initiate disputes over "obvious" causes of lost market access, i.e., cases where the information needed to identify a promising WTO case is most readily available. Figure 6 allocates the disputes into one of five "observability" categories.8 From an exporting firm's perspective, antidumping and countervailing duties are the most "obvious" causes of lost foreign market access because WTO rules require the foreign government to inform affected exporting firms directly of its actions. Many developing country complaints fall into this category of market access lost due to an "obvious" import restriction.9 Safeguards are the next most obvious new trade restriction. Although WTO rules do not require the foreign government to inform exporting firms, it must alert affected WTO member governments. Many complaints initiated by developing countries also fall into this category. 8 Disputes classified as "other" are ones not directly related to a loss in foreign market access for a particular industry. These disputes focus on WTO-inconsistent policies or procedures that are systemic rather than involving specific industries or products. 9 There are, of course, other contributors in addition to the observability of the measure that makes antidumping a frequent target for WTO dispute settlement. Antidumping import restrictions are firm-specific, which can eliminate or reduce free-riding difficulties of political organization to engage government policymakers. Moreover, the use of antidumping has been proliferating across the WTO membership, including its use by many developing countries (Bown 2007, 2008). 11 Medium observability causes include "border measure" types of infractions that may be observable to an exporting firm but not necessarily to officials of its government. An affected firm may be able to identify a change in treatment at the border involving new costs or restricted access to the foreign market. Such actions by an importing country may include imposing a new quantitative restriction or higher duty, reclassifying a product's tariff category, changing the procedure for valuing imports in a way that results in higher duties, or raising the costs of acquiring the licenses needed to engage in trade. As Figure 6 shows, disputes involving medium observability causes are evenly balanced between developing and developed economies, though a significant number of such disputes do involve developing countries. Low observability causes of disputes reflect situations where lost foreign market access may be due to influences that do not directly affect the exporting firm at the border but instead through an induced change in importer or consumer behavior, e.g., lost foreign market access that results when consumers switch demand toward another supplier, whether a domestic firm in the foreign market or a competing exporter. There are many examples of WTO-illegal causes of consumer demand switch that may be difficult for the exporting firm to identify due to lack of information. For example, a competing producer may have been able to offer a lower price because of a WTO-illegal subsidy, an export restriction on a key input that reduces the domestic cost of an intermediate input, discriminatory domestic tax treatment, or failure to enforce intellectual property rights. The EC and the U.S. dominate disputes initiated over low observability measures, and such complaints are much less frequently initiated by developing countries. Figure 7 compares the observability of alleged cause of market access loss in disputes initiated in the two study periods. Of the set of WTO disputes initiated during 2001-2008, 43% (74 out of 169) concerned one of three forms of administered protection ­ antidumping, countervailing duties, or safeguards--which we have classified as "obvious" and "high observability" infractions. This is a much larger share than during the 1995-2000 period. Although the increasingly prominent role of disputes over these measures is consistent with the global proliferation of their use since the 1990s (Bown 2007), 12 especially in the case of antidumping, it may also signal the existence of an informational barrier that hinders developing countries from enforcing their market access rights in other situations where WTO- illegal measures are less observable. 3. The Advisory Centre on WTO Law and Its Involvement in WTO Disputes To help developing countries get maximum benefits from WTO membership, a group of nations established the Advisory Centre on WTO Law (ACWL) in 2001; developed countries in the group provide the bulk of the financing. The ACWL assists developing countries in their enforcement through the DSU of WTO market access rights by providing a variety of subsidized legal services to developing country governments.10 Although these services include free or low cost legal advice on WTO issues and training programs for officials who carry out WTO-related functions on behalf of their governments (capacity building), the ACWL's most prominent role is to supply low-cost legal support to developing countries when they act as complainants, respondents, or third parties in WTO dispute-settlement proceedings. Under its charter, the ACWL can provide its services to developing countries, customs territories, or economies in transition. Specifically, ACWL services are available to any developing country that is a member of the Centre as well as any WTO member designated by the United Nations as a least developed country (LDC)--i.e., any WTO member with very low per capita income. In this section we analyze data on ACWL activities from 2001 through 2008. We begin by examining how developing countries have actually used the ACWL litigation services: which countries the ACWL has represented in DSU cases, how frequently, and against whom. Then we use economic theory to identify various mechanisms through which introduction of the ACWL in 2001 may have affected the ways developing countries participate in WTO dispute settlement and the resulting DSU caseload. Finally, we look for patterns in the data that are consistent with these channels. 10 Earlier theoretical analyses of the ACWL include Van der Borght (1999), Hoekman and Mavroidis (2000), Jackson (2002), Bown and Hoekman (2005), and Shaffer (2006). 13 3.1 The ACWL Role in WTO Disputes The data in Table 3 document ACWL involvement in WTO dispute-settlement cases from its establishment in 2001 through 2008, roughly seven years. 11 Over this period, WTO members initiated 144 formal disputes against other members (again see Figure 1). The ACWL assisted developing countries in 23 of the 144 disputes initiated during this period, or over 16% of all disputes.12 The ACWL has typically represented the complainant that initiated a WTO case. Of the 23 disputes in which the ACWL provided assistance to a developing country, 19 were instances in which it assisted the developing country as a complainant. In the remaining four cases, it assisted a developing country as an interested third party.13 To provide some perspective on the quantitative importance of ACWL activities, we can compare the number of cases initiated during the period in which the ACWL provided legal services to a developing country complainant (19) to the number of other complaints initiated by WTO members. By this measure, the only members more active as a complainant than the ACWL were the U.S. (also 19 times) and the EC (21 times). But the ACWL has been less active than the U.S. and EC in overall dispute settlement. The U.S. was required to defend itself as a respondent in 46 cases and served as an interested third party in another 34 cases. The EC likewise was a respondent in 28 cases and served as an interested third party in another 47 cases. Table 4 shows that seventeen different developing countries used the ACWL's services for DSU litigation, with the most frequent clients being 11 The first initiated dispute in which the ACWL played a role was DS231 in 2001, and the last dispute for which it reports participation was DS374 in May 2008. Thus, we use all WTO disputes initiated over the same time period as our comparison sample. 12 During this period, the ACWL also participated in three disputes (DS141, DS146, and DS192) initiated prior to 2001. While the ACWL had not been available to assist developing countries in the initiation of these cases, it was asked to assist at a later phase of the multi-year dispute settlement process. 13 In one (DS246) of the 19 disputes in which it worked on behalf of the complainant, it also separately represented another country (Paraguay) as a third party after determining its interests were in alignment with the main complainant client (India). In DS146, the ACWL did assist India as a respondent country, but this dispute was one of the cases initiated before establishment of the ACWL. Respondent developing countries were given ACWL-like assistance (in terms of access to legal services at subsidized rates) in two additional cases ­ the Dominican Republic in DS302 and Colombia in DS366. Because the ACWL was already representing the complainant in each case, the Centre could not also provide the legal assistance to the respondent in the dispute. However, each respondent received legal assistance from private law firms that were part of the ACWL's "Roster of External Legal Counsel" program, described at http://www.acwl.ch/e/dispute/counsel_e.aspx . 14 Thailand (5 times) and India (4 times). Other repeat clients for the ACWL's DSU support services include Indonesia, Pakistan, Panama, and the Philippines. Table 3 also identifies the ACWL clients' targets for WTO enforcement. Much like the broader WTO membership (see Table 2), the ACWL represented complainant countries filing cases most frequently against the U.S. (3 times) and the EC (6 times). But the ACWL also assisted its developing- country clients in nine instances in challenging another developing country's failure to live up to its WTO commitments. During this period, the ACWL worked on behalf of the complainant country in 17% (9 out of 54) of all disputes that a developing country initiated against another developing country. This is a larger number of cases initiated against developing countries than were initiated by any other WTO member except the EC (10) and the U.S. (12). 14 The significant total number of developing-country WTO challenges initiated against another developing country highlights the importance to developing-country exporters of reduced trade barriers in developing as well as developed countries, and also the importance of WTO dispute settlement as a means of enforcing market access rights. Finally, the data confirm that most of the disputes concern the sectors expected to be of greatest interest to exporters in developing countries (see Figure 4). Table 3 indicates that ACWL clients used a WTO complaint to enforce foreign market access rights in either agricultural (including foodstuffs and fisheries) or textiles and apparel in 13 out of the 19 disputes they initiated. 3.2 How has the ACWL affected the DSU caseload? The rate of initiation of developing country cases since 2001 Did establishment of the ACWL in 2001 lead to a sharp increase in developing-country participation in WTO dispute settlement activity? Our earlier analysis (see Figures 1 and 3) can rule out an obvious effect in the data. The number of disputes initiated each year by developing countries has fallen from roughly 13 cases per year (1995-2000) to 11 cases per year (2001-2008). But as we previously noted, this drop in 14 Furthermore, as we noted above, the U.S. and EC disputes against developing countries during this period were increasingly focused on China and India (12 out of 22 during this particular time period). Interestingly the ACWL has only been involved in one case filed against India (DS306 brought by Bangladesh) and none against China. 15 activity was much less dramatic than for other WTO members: a larger share of all WTO disputes was initiated by developing countries after establishment of the ACWL in July 2001. Between January 1, 1995 and June 30, 2001, developed countries initiated 172 disputes compared to 90 disputes initiated by developing countries, almost a 2:1 ratio. Between July 1, 2001 and December 2008, developed countries initiated only 80 disputes compared to the 73 initiations by developing country complainants, almost a 1:1 ratio. But while the ACWL may have played some role in preventing the number of cases initiated by developing countries from following the sharp downward trend of developed countries post-2000, and while there are reasons why the rate of dispute settlement initiation might have been expected to fall in the later period (see Figure 2 on good times and export growth), below we look at detailed data to identify evidence of specific channels through which availability of ACWL services has affected the WTO caseload. 3.3 How has the ACWL affected the DSU caseload? More cases by the same countries and not introduction of new countries The ACWL lowers the cost to an eligible country of enforcing another country's WTO commitments. This could affect the DSU enforcement caseload by inducing new complainants with no prior history of using WTO dispute settlement to self-enforce foreign market access commitments for the first time (the extensive margin) or by encouraging the same developing countries that are historical users of WTO enforcement to initiate more cases (the intensive margin). Table 4 presents data on ACWL client use of WTO enforcement both before and after its first ACWL experience. For each ACWL client, the table shows the year the country first used the ACWL for DSU services, how many times it had used the DSU prior to becoming an ACWL client (broken down by instances as complainant, respondent, and interested third party), and how frequently it was involved in DSU enforcement cases after using the ACWL for the first time. To interpret the information in Table 4, consider what it shows for Thailand. Thailand first used the ACWL for a WTO dispute initiated in 2003. Prior to the 2003 dispute, Thailand had been involved in 16 32 other WTO cases: eight as a complainant, one as a respondent, and 23 as a third party. Overall, it has used the ACWL services five times. Since its first involvement with the ACWL in 2003, Thailand has been involved in twenty additional WTO disputes ­ four times as a complainant, twice as a respondent, and fourteen times as a third party. The data for Thailand are quite typical. Specifically, most ACWL clients had substantial prior experience in WTO enforcement before their first use of ACWL services. Thus, almost all use of the ACWL over the 2001-2008 period is at the intensive margin, i.e., countries with past experience in the DSU used the ACWL to represent them in additional disputes. The only example of a country that had never previously been involved in the DSU before using ACWL support was Chad, which was a third party in the U.S. ­ Upland Cotton dispute initiated by Brazil.15 While these data reflect only the first seven years of ACWL operation, there is almost no evidence that the ACWL has had the effect of introducing new countries (the extensive margin) to formal WTO self-enforcement. Apart from the single instance of Chad, all other ACWL clients had some prior DSU litigation experience. However, this finding is not surprising given the ACWL's mandate and role. Because it cannot "ambulance chase" or even have direct (non-authorized) contact with firms with foreign market access concerns but only with their governments, the ACWL is really only a mechanism to assist countries that are already relatively knowledgeable about the WTO and its enforcement process--countries that have other ways to identify a potential WTO case worth litigating if the cost of doing so is not too high relative to the likely benefits from enforcing another country's WTO commitments. 3.4 How has the ACWL affected the DSU caseload? The same countries using the DSU differently Even if the ACWL has not been effective in encouraging new countries to participate in WTO dispute settlement, it could have provided current developing-country users with resources that allowed them to 15 Even in this instance, the ACWL assisted Chad in the dispute's latter phase of an "Article 21.5" Compliance Panel (ACWL, 2008). Chad received assistance during the Panel and Appellate Body phases of the dispute from the private law firm White & Case. 17 pursue additional cases, or to pursue cases differently, than they would have done without ACWL assistance. Here we examine two channels through which access to ACWL resources can benefit developing countries. The first possibility is that the ACWL has allowed countries to file sole- complainant disputes on behalf of their exporters, rather than waiting for cases in which they have common interests with other WTO members and are thus able to pursue joint disputes and share costs. The second possibility is the ACWL allows countries to pursue the DSU legal process more extensively in support of any given market access enforcement interest. Prior to establishment of the ACWL, the high cost of litigation may have caused developing countries to drop even a strong case, or to accept a disadvantageous early settlement, rather than committing the additional resources needed to get a legal ruling in their favor.16 Table 5 compares ACWL clients' use of the DSU process to the way they used the DSU prior to establishment of the ACWL. The table provides evidence that access to ACWL resources has affected developing countries' litigation behavior along the lines predicted by theory. The table lists the thirteen developing countries that used the ACWL between 2001 and 2008 to initiate at least one sole- complainant WTO dispute. Eleven of these thirteen countries had filed at least one complaint at the WTO previously, and the other two had participated in the DSU process as either a respondent or third party, so none of these countries were being introduced to WTO enforcement. But eight out of the thirteen countries had never previously litigated a sole-complainant case. Rather, these countries' previous experience in dispute settlement was mainly in instances in which they joined a more powerful WTO member in a dispute or pooled with other developing countries affected by the same policy. For example, Latin American banana-exporting countries followed the U.S. lead in EC ­ Banana III (DS27), and Southeast Asian shrimp-exporting countries acted collectively in U.S. ­ Shrimp (DS58). Relative to the way the same countries had used the DSU historically, their use of ACWL services to initiate their own 16 Busch and Reinhardt (2003) provide evidence from pre-2001 disputes that developing country complainants were more likely to settle cases early and less likely to receive significant concessions from respondent countries than developed country complainants. 18 disputes on behalf of their exporters suggests that availability of ACWL services has enhanced developing-country use of the DSU to enforce their market-access rights.17 A second possible mechanism through which access to ACWL resources may affect developing- country litigation behavior goes beyond what types of disputes are initiated to the way actual cases are prosecuted and litigated. Table 5 also provides evidence that the ACWL's sole-complainant clients were able to pursue the DSU process further once a case was initiated. The middle column of Table 5 documents that only India and the Philippines out of these thirteen ACWL sole-complainant client countries had ever previously pushed a sole-complainant DSU proceeding far enough to obtain a Panel ruling. The last column of Table 5 indicates that in ACWL-backed cases, seven "new" countries prosecuted their sole-complainant cases to a Panel ruling for the first time. In the other four cases involving sole-complainant ACWL clients that had not previously obtained a Panel Report, two countries (Bangladesh, Nicaragua) settled their disputes through "mutually agreed upon understanding" notifications to the WTO; the ACWL-backed disputes of the remaining two countries (Colombia, Panama) were initiated too recently to have yet been resolved. Thus, while there is almost no evidence that the ACWL has introduced completely new countries without prior DSU experience to WTO enforcement, the data on the ACWL's first seven years suggest that the ACWL has empowered developing countries with prior, though sometimes minimal, DSU experience to "do more" ­ i.e., to initiate cases on their own and to prosecute those cases further through the legal process to obtain (politically neutral) WTO legal rulings. Such rulings can be useful to the respondent country in mobilizing domestic support for the policy changes needed to comply with its WTO obligations. 17 Although the evidence is consistent with the hypothesis that availability of AWCL services has allowed more developing countries to file sole-complainant cases, other factors for which we are not controlling here may also have affected countries' decisions to pursue sole-complainant disputes. An alternative hypothesis also consistent with the data is that these countries used their prior experience in co-complainant disputes to learn about the DSU process until they had acquired the expertise needed to initiate a dispute on their own. 19 3.5 How has the ACWL affected the DSU caseload? Evidence of a scale effect In addition to providing resources that may affect the DSU caseload by allowing developing countries to pursue more sole-complainant disputes and to pursue them more intensively, which may be important in obtaining the required policy reform in respondent countries, the ACWL may also affect the scale of cases that are initiated. Consider a theoretical model in which an exporting firm and its government each face certain costs of using the WTO self-enforcement process to enforce the foreign market access to which a trading partner has committed. The self-enforcement process includes a number of steps, beginning with an exporting firm's identification of a possible WTO complaint (a possibly WTO-illegal measure that has reduced the firm's export market access), the firm's calculation of potential costs and benefits from enforcing its foreign market access, the firm's efforts to convince its government to bring the case, development and prosecution of the legal case at the WTO, calculation of retaliation threats, and finally efforts to generate public and political support in the respondent country for the policy reform required to bring the respondent into WTO compliance.18 By offering subsidized legal assistance to developing countries at the litigation stage, the ACWL lowers the overall costs of the self-enforcement process. If this cost reduction is too small to affect the composition of cases that are brought to the WTO, i.e., if only the same cases are litigated, then the existence of the ACWL merely transfers resources from rich to poor countries without affecting the volume or composition of WTO enforcement activity. But if the ACWL lowers the total cost of enforcement faced by developing countries by enough to change the scale of disputes countries choose to initiate at the WTO, "more" foreign market access would be enforceable at the WTO, i.e., trade disputes involving a smaller dollar value would be brought for DSU enforcement. To find evidence in support of this kind of scale effect, we examine developing-country initiation and ACWL involvement in one specific category of WTO dispute ­ those in which the respondent country has imposed a potentially 18 Bown and Hoekman (2005) refer to this six-step sequence as the WTO's "extended litigation process." 20 WTO-inconsistent antidumping measure. We focus on WTO disputes initiated during the ACWL period by developing countries that were either ACWL clients or which could have become ACWL members in order to use ACWL services. 19 One indicator of a scale effect consistent with the underlying theory would be that the ACWL represented clients in disputes where the economic stakes involved were smaller than in disputes initiated without ACWL representation.20 Table 6 provides information on the scale of 11 disputes initiated by developing countries over respondent use of antidumping during the ACWL period. In the cases shown in the upper part of the table, the developing country complainant did not use the ACWL; in the cases in the lower part of the table, the complainant did use the ACWL. To interpret the table, consider first a WTO dispute like the case that Turkey brought without ACWL assistance against South Africa (South Africa -- Definitive Anti-Dumping Measures on Blanketing from Turkey). In the three years prior to the South African import restriction, Turkish exporting firms had averaged $5.9 million in sales to South Africa per year. A mere two years after the imposition of the South African AD restriction, the Turkish exporters had lost $5.8 million in sales to that market. Turkey did not use the ACWL in this dispute, although it had become a member of the ACWL and was therefore eligible to do so.21 Compare this case with the similar WTO challenge to AD involving matches that Pakistan brought against Egypt (Egypt -- Matches), in which Pakistan did use ACWL services. In this case the market access at stake, which is likely related to the size of the profits that firms would need to make to cover the cost of litigation if they were to pursue a dispute without ACWL assistance, was much smaller. Pakistan's exporting firms had averaged only $2.6 million in sales to Egypt per year in the three years prior to the new trade restriction. 19 The list of disputes initiated without ACWL participation include disputes initiated by developing countries that had not acceded to the ACWL (Argentina, Brazil) but also some initiated by ACWL members that chose not to use ACWL services in a particular case (India, Taiwan, Turkey). 20 A question raised by our comparison is why countries with access to the ACWL's subsidized legal services choose to use them mainly in smaller cases. 21 This dispute was initiated in April 2003, and Turkey became a member of the ACWL in August 2003. As Table 3 indicates, countries such as India and Pakistan used ACWL services for later phases of the dispute-settlement process (e.g., the Panel process) if they became members after initiation of a particular dispute. 21 While such a comparison of just two cases provides only anecdotal evidence, the same pattern is reflected in the averages for the two sets of cases presented in Table 6. Both the value of trade prior to the new trade restriction and the lost value of trade due to the new antidumping measure (i.e., both of the proxies for the size of market access at stake) are typically much smaller in the WTO disputes pursued with ACWL assistance than in the non-ACWL disputes. Prior to imposition of the contested trade restrictions, four out of six non-ACWL disputes had market access valued at more than $10 million per year, while in four out of five ACWL-backed disputes prior market access was valued at less than $3 million per year. When we examine averages across the two sets of disputes, prior exports for non-ACWL cases was larger ($28.7 million versus $9.7 million), and the value of lost exports for non-ACWL cases was also larger ($20.0 million versus $1.9 million). The evidence from the ACWL's early cases in Table 6 is thus consistent with the hypothesis that the ACWL may have a scale effect by allowing developing countries to enforce foreign market access commitments involving a lower total value of export sales.22 These results are mostly suggestive, since the total number of observations is so small. While we are able to control for some factors by focusing on one particular type of trade dispute, the limited data do not allow us to employ a full regression framework and thus control for other factors that may also affect whether a country chooses to use the ACWL. However, our approach illustrates the sorts of comparisons that can be made more comprehensively once time passes and there exists additional data. 4 Conclusions and Policy Implications Our data show that developing country use of the DSU to self-enforce foreign market access commitments under the WTO has continued at a relatively constant rate throughout the 1995-2008 period, 22 An unanswered question is why eligible countries chose not to use the ACWL's low-cost services even in cases where the economic stakes are large. One possible explanation is that the firms in these larger cases already have established relationships with their own private law firms, so that proceeding with these firms instead of beginning a new relationship with ACWL lawyers may be faster and more efficient. Alternatively, firms with larger and more complex business operations may prefer to limit the amount of proprietary business information that must be shared with government officials. A final possibility is that some private firms "ambulance chase" by identifying potential cases with stakes large enough to be profitable for clients to pursue even when paying for their legal services at market rates. 22 in contrast to a marked decline in the use of the DSU by developed countries. The data show that developing countries have used the DSU to self-enforce their foreign market access rights in other developing countries as well as in developed countries. We also show that developing countries are more likely to bring WTO disputes when the cause of lost market access is readily apparent to exporting firms and government officials, i.e., foreign governments' implementation of antidumping policies, countervailing duties, or safeguards. Our efforts to detect the influence of the ACWL on self-enforcement actions of developing countries reveals that at least so far, the availability of low-cost ACWL services has not been enough to expand the set of developing countries that undertake litigation under the DSU to enforce their foreign market access rights. Rather, access to assistance from the ACWL has allowed countries that were prior users of DSU enforcement to act for the first time as sole complainants, to pursue disputes undertaken more fully, and initiate some cases with smaller stakes than those that are profitable to undertake even when paying the market rate for the required legal services. Several developing-country groups have recently called for additional funding within the WTO system to cover their costs of DSU litigation. 23 However, the potential benefits of direct WTO funding over the current system in which low-cost legal assistance is provided by the ACWL, are not apparent. The ACWL already provides a substantial transfer from rich countries to fund the poor countries' dispute settlement assistance needs. Repeat use of ACWL services by many of its developing-country clients suggests that the quality of the services provided is at least adequate. If underfunding were preventing the ACWL from carrying out its mission, there would be an argument for expanding its budget and staff. But it is unclear what establishing a potentially redundant framework within the WTO would achieve. Moreover, the status of the ACWL as an independent organization outside of the formal WTO framework may have the important advantage of making ACWL operations on behalf of its clients relatively immune 23 Examples include proposals from Cuba, Egypt, India, Malaysia and Pakistan (the "Like Minded Group") "Dispute Settlement Understanding Proposals: Legal Text; Revisions In Some Of The Proposals In TN/DS/W/47" WTO document JOB 06/222, 10 July 2006; as well as one from the African Group, see "Text for the African Group Proposals on Dispute Settlement Understanding Negotiations," WTO document TN/DS/W/92, 5 March 2008. 23 from the internal politics surrounding other WTO business, especially multilateral negotiations. An alternative arrangement in which similar services were provided internally could also limit the perceived freedom of action of developing countries in their involvement with the WTO Secretariat. Our analysis does suggest a remaining obstacle currently limiting developing countries' access to WTO self-enforcement that the ACWL, despite its current successes, is not able to overcome--lack of the information required to identify potential cases for WTO dispute settlement. Our data confirm that when developing country firms and their governments can readily observe the cause of the lost market access (antidumping, countervailing measures, safeguards), they challenge it at the WTO, often with the help of the ACWL. However, developing countries need additional monitoring resources to help them identify less-observable causes of lost market access and thus situations in which they also may have foreign market-access enforcement interests to pursue. Furthermore, arrangements that subsidize the cost of WTO litigation may actually worsen the problem of inadequate information to the extent that they discourage private law firms from ambulance chasing--i.e., generating information on behalf of prospective developing-country clients. Given the existence of the ACWL, such potential clients can now bring disputes to the ACWL for litigation at lower cost, so private law firms are less likely to recoup their investment in information. To the extent that private law firms are in fact discouraged from generating information as a means to attract new developing-country clients, measures to encourage developing-country use of the WTO self-enforcement process by subsidizing the cost of WTO litigation, whether through the ACWL or an alternative WTO- operated facility, may thus increase the need for public provision of information on potential WTO cases. 24 References ACWL (2008) "Assistance in WTO Dispute Settlement Proceedings Since July 2001," available at http://www.acwl.ch/e/dispute/wto_e.aspx, last accessed on 5 December 2008. ACWL (2007) "How to Use the Services of the ACWL: A Guide for Developing Countries and LDCs," Geneva, ACWL, December 2007. Available at http://www.acwl.ch/pdf/how_to.pdf . Bown, Chad P. (2009) Self Enforcing Trade: Developing Countries and WTO Dispute Settlement. Washington, DC: Brookings Institution Press, forthcoming. Bown, Chad P. (2008) "The WTO and Antidumping in Developing Countries," Economics and Politics 20(2): 255-288. Bown, Chad P. (2007) "Global Antidumping Database," (Current Version 3.0), Brandeis University working paper, June. Available at http://people.brandeis.edu/~cbown/global_ad/ . Bown, Chad P. (2005) "Participation in WTO Dispute Settlement: Complainants, Interested Parties and Free Riders," World Bank Economic Review 19(2): 287-310. Bown, Chad P. (2002) "The Economics of Trade Disputes, the GATT's Article XXIII and the WTO's Dispute Settlement Understanding," Economics and Politics 14(3): 283-323. Bown, Chad P. and Bernard M. Hoekman (2008) "Developing Countries and Enforcement of Trade Agreements: Why Dispute Settlement Is Not Enough," Journal of World Trade 42(1): 177-203. Bown, Chad P. and Bernard M. Hoekman (2005) "WTO Dispute Settlement and the Missing Developing Country Cases: Engaging the Private Sector," Journal of International Economic Law 8(4): 861- 890. Busch, Marc L. and Eric Reinhardt (2003) "Developing Countries and GATT/WTO Dispute Settlement," Journal of World Trade 37(4): 719-735. Busch, Marc L., Eric Reinhardt, and Gregory Shaffer (2008) "Does Legal Capacity Matter? Explaining Patterns of Protectionism in the Shadow of WTO Litigation," University of Minnesota unpublished manuscript, August. Davis, Christina L. and Sarah Blodgett Bermeo (forthcoming) "Who Files? Developing Country Participation in WTO Adjudication," Journal of Politics. Hoekman, Bernard M. and Petros C. Mavroidis (2000) "WTO Dispute Settlement, Transparency and Surveillance," The World Economy 23(4): 527 ­ 542. Horn, Henrik and Petros C. Mavroidis (2008) "WTO Dispute Settlement Database," available at http://go.worldbank.org/X5EZPHXJY0 , last accessed 29 December 2008. Horn, Henrik, Petros C. Mavroidis and Håkan Nordström (2005) "Is the Use of the WTO Dispute Settlement System Biased?" in Petros C. Mavroidis and Alan Sykes (eds.), The WTO and International Trade Law/Dispute Settlement, Cheltenham, UK: Edward Elgar. 25 Hudec, Robert E. (1993) Enforcing International Trade Law: the Evolution of the Modern GATT Legal System. New Hampshire: Butterworth Legal Publishers. Jackson, John H. (2002) "Perceptions about the WTO Trade Institutions," World Trade Review 1(1): 101- 114. Nordström, Håkan and Gregory Shaffer (2008) "Access to Justice in the World Trade Organization: A Case for a Small Claims Procedure?" World Trade Review 7(4): 587-640. Ostry, Sylvia (2002) "The Uruguay Round North-South Grand Bargain: Implications for Future Negotiations," in Daniel L. M. Kennedy and James D. Southwick, eds., The Political Economy of International Trade Law: Essays in Honor of Robert E. Hudec. New York: Cambridge University Press. Shaffer, Gregory (2006) "The Challenges of WTO Law: Strategies for Developing Country Adaptation," World Trade Review 5(2): 177-198. Shaffer, Gregory (2003) Defending Interests: Public-Private Partnerships in WTO Litigation. Washington, DC: Brookings Institution Press. Steinberg, Richard H. and Timothy E. Josling (2003) "When the Peace Ends: The Vulnerability of EC and US Agricultural Subsidies to WTO Legal Challenge," Journal of International Economic Law 6(2): 369-417. Van der Borght, Kim (1999) "The Advisory Center on the WTO Law: Advancing Fairness and Equality," Journal of International Economic Law 2(4): 723-728. WTO (1995) Analytical Index: Guide to GATT Law and Practice, Vols. 1 and 2. Geneva: WTO, Geneva. WTO (1997) Panel Reports under the MTN Agreements and Arrangements (Tokyo Round Codes) of 1979. WTO unpublished document. 26 Figure 1. WTO Dispute Initiations, 1995-2000 and 2001-2008 WTO 60 dispute initiations by year 50 40 30 20 10 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Source: Compiled by the authors from WTO (2008). Disputes are broken down into bilateral (complainant/ respondent) pairs. Because some disputes involved more than one complainant, the 388 requests for consultations initiated over the 1995-2008 period yielded 415 bilaterally paired disputes. 27 Figure 2. World Exports, 1995-2000 and 2001-2006 $ trillions 8.00 7.00 Developed 6.00 Countries 5.00 Developing Countries 4.00 (except China) China 3.00 2.00 1.00 0.00 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Source: Compiled by the authors from WITS. 28 Figure 3. Average WTO Disputes per Year by Category of Complainant, 1995-2000 and 2001-2008 Average number 14 of WTO disputes initiated per year 12 during period 10 8 1995-2000 2001-2008 6 4 2 0 U.S. EC Other industrialized Developing complainants complainants Source: Compiled by the authors from WTO (2008). Disputes are broken down into bilateral (complainant/ respondent) pairs. See source notes for Figure 1. 29 Figure 4. WTO Disputes by Industrial Sector and Category of Complainant Agriculture, Beverages, Seafood Apparel and Textiles Steel Developing Other Manufacturing complainants Developed Aircraft, Shipbuilding complainants Auto Industry Pharmaceuticals, IT, Media Not Classifiable 0 10 20 30 40 50 60 70 80 90 Total Number of Disputes, 1995-2008 Source: Compiled by the authors from WTO (2008) and Horn and Mavroidis (2008). Disputes are broken down into bilateral (complainant/ respondent) pairs. See source notes for Figure 1. 30 Figure 5. WTO Disputes by Industrial Sector, 1995-2008 25 Disputes initiated per year 20 15 Agriculture, Beverages, Seafood Steel 10 Apparel and Textiles Pharmaceuticals, IT, Media 5 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Source: Compiled by the authors from WTO (2008) and Horn and Mavroidis (2008). 31 Figure 6. WTO Disputes by Degree of Observability of Cause of Alleged Lost Market Access and Complainant Category "Obvious" (AD, CVD) Developing complainant High (safeguard) Other Medium industrialized (border measure, complainant customs, valuation, etc.) EC or U.S. Low (subsidy, other complainant domestic measure, export restriction, etc.) Other 0 10 20 30 40 50 60 70 80 Total Number of Disputes, 1995-2008 Source: Compiled by the authors from WTO (2008) and Horn and Mavroidis (2008). Disputes are broken down into bilateral (complainant/ respondent) pairs. See source notes for Figure 1. AD is antidumping; CVD is countervailing duty. 32 Figure 7. WTO Disputes by Observability of Alleged Policy Cause of Lost Market Access, 1995-2000 and 2001-2008 Share of 0.45 total disputes initiated during 0.4 period 0.35 0.3 0.25 1995-2000 2001-2008 0.2 0.15 0.1 0.05 0 "Obvious" High Medium Low Other (AD, CVD) (safeguard) (border measure, (subsidy, other customs, valuation, domestic measure, etc.) export restriction, etc.) Source: Compiled by the authors from WTO (2008) and Horn and Mavroidis (2008). Disputes are broken down into bilateral (complainant/ respondent) pairs. See source notes for Figure 1. AD is antidumping; CVD is countervailing duty. 33 Table 1. WTO Member Dispute Participation as Complainant, Respondent, and Third Party, 1995-2008 Country Number of Times Number of Times Number of Times Complainant Respondent Third Party* EC 78 89 82 U.S. 91 116 73 Other industrialized countries Australia 7 10 47 Canada 31 15 64 Japan 13 15 90 Korea 13 13 43 New Zealand 7 0 27 Norway 3 0 27 Singapore 1 0 4 Switzerland 4 0 8 Taiwan 3 0 39 Developing countries Antigua and Barbuda 1 0 0 Argentina 14 16 20 Bangladesh 1 0 1 Brazil 24 14 49 Chile 10 12 22 China 3 13 62 Colombia 5 3 16 Costa Rica 4 0 9 Croatia 0 1 0 Czech Republic 1 2 0 Dominican Republic 0 3 3 Ecuador 3 3 9 Egypt 0 4 4 Guatemala 6 2 11 Honduras 6 0 12 Hong Kong 1 0 0 Hungary 5 7 2 India 18 20 51 Indonesia 4 4 4 Malaysia 1 1 2 Mexico 20 14 45 Nicaragua 1 2 6 Pakistan 3 2 9 Panama 5 1 2 Peru 2 4 8 Philippines 5 4 5 Poland 3 1 1 Romania 0 2 0 Slovak Republic 0 3 0 South Africa 0 3 0 Sri Lanka 1 0 3 Thailand 13 3 37 Trinidad and Tobago 0 2 3 Turkey 2 8 18 Uruguay 1 1 5 Venezuela 1 2 15 Total 415 415 938 Source: Compiled by the authors from WTO (2008). Disputes are broken down into bilateral (complainant/ respondent) pairs. See source notes for Figure 1. Data on participation as third party is available only through dispute DS367, initiated 31 August 2007, available at WTO website, last accessed 5 January 2009. *Does not include WTO members that participated in DSU activities only as third party. This list omits 32 WTO members that have participated collectively as third parties 117 times in addition to those reported in the table. Most of those 32 WTO members participated three times or fewer, and most are developing or least developed countries. 34 Table 2. WTO Disputes by Category of Complainant and Respondent Countries Respondent 1995-2000 2001-2008 Complainant U.S./EC Other Developing Total U.S./EC Other Developing Total Ind. Ind. U.S./EC 48 30 45 123 15 7 24 46 Other Ind. 24 7 14 45 31 3 3 37 Developing 45 2 31 78 42 4 40 86 Total 117 39 90 246 88 14 67 169 Source: Compiled by the authors from WTO (2008). Disputes are broken down into bilateral (complainant/ respondent) pairs. See source notes for Figure 1. 35 Table 3. ACWL Participation in WTO Trade Disputes, 2001-2008* WTO Dispute, year initiated ACWL Client, Role in Dispute DS141: EC -- Bed Linen, 1998 India, potential appellant DS146: India -- Autos, 1998 India, respondent DS192: U.S. -- Cotton Yarn, 2000 Pakistan, complainant DS231: EC -- Sardines, 2001 Peru, complainant DS237: Turkey -- Fresh Fruit Import Procedures, 2001 Ecuador, complainant DS243: U.S. -- Textiles Rules of Origin, 2002 India, complainant DS246: EC -- Tariff Preferences, 2002 India, complainant; Paraguay, Colombia**, Ecuador**, Peru**, and Venezuela** as third parties DS264: U.S. -- Softwood Lumber V, 2002 Thailand, third party DS267: U.S. -- Upland Cotton, 2002 Chad, third party DS270: Australia -- Fresh Fruit and Vegetables, 2002 Philippines, complainant DS271: Australia -- Certain Measures Affecting the Importation of Philippines, complainant Fresh Pineapple, 2002 DS283: EC -- Export Subsidies on Sugar, 2003 Thailand, complainant DS284: Mexico -- Certain Measures Preventing the Importation of Nicaragua, complainant Black Beans from Nicaragua, 2003 DS286: EC -- Chicken Cuts, 2003 Thailand, complainant DS302: Dominican Republic -- Import and Sale of Cigarettes, 2003 Honduras, complainant; Dominican Republic**, respondent DS306: India -- AD Measure on Batteries from Bangladesh, 2004 Bangladesh, complainant DS312: Korea -- Certain Paper, 2004 Indonesia, complainant DS322: U.S. -- Zeroing (Japan), 2004 Thailand, third party DS327: Egypt -- Matches, 2005 Pakistan, complainant DS331: Mexico -- Steel Pipes and Tubes, 2005 Guatemala, complainant DS334: Turkey ­ Rice, 2006 Turkey**, respondent DS343: U.S. -- Shrimp (Thailand), 2006 Thailand, complainant DS348: Colombia -- Customs Measures on Importation of Certain Panama, complainant Goods from Panama, 2006 DS361: EC -- Regime for the Importation of Bananas, 2007 Colombia, complainant DS366: Colombia -- Ports of Entry, 2007 Panama, complainant; Colombia**, respondent DS374: South Africa -- AD Measures on Uncoated Woodfree Paper, Indonesia, complainant 2008 Sources: ACWL "Assistance in WTO dispute settlement proceedings since July 2001," http://www.acwl.ch/e/dispute/wto_e.aspx *Through May 2008. ** Legal assistance provided not by the ACWL but through hiring from its "Roster of External Legal Counsel" program. Dispute initiated prior to ACWL 2001 establishment; the ACWL assisted at a later phase of the multi-year dispute-settlement process, such as the appeal. 36 Table 4. ACWL Clients in WTO Disputes before and after First ACWL Experience, 2001-2008* First No. of time times No. of times WTO disputant No. of times WTO disputant Country ACWL ACWL prior to first ACWL experience after first ACWL experience client client as... as... ...Third ...Third ...Comp. ...Resp. ...Comp. ...Resp. Party Party Bangladesh 2004 1 0 0 1 0 0 0 Chad 2007 1 0 0 0 0 0 0 Colombia 2007 2 4 2 16 0 1 0 Dominican Republic 2003 1 0 1 3 0 1 0 Ecuador 2001 2 1 2 6 1 1 3 Guatemala 2005 1 5 2 9 0 0 2 Honduras 2003 1 5 0 10 0 0 2 India 2001 4 13 13 31 4 7 20 Indonesia 2004 2 2 4 4 1 0 0 Nicaragua 2003 1 0 2 5 0 0 1 Pakistan 2001 2 1 2 4 1 0 5 Panama 2006 2 2 1 2 2 0 0 Paraguay 2002 1 0 0 5 0 0 9 Peru 2001 2 1 2 4 0 2 4 Philippines 2002 2 2 4 4 1 0 1 Thailand 2003 5 8 1 23 4 2 14 Venezuela 2001 1 1 1 4 0 1 10 Source: Data compiled by the authors from matching public records from ACWL website with public information on WTO website. "Comp." indicates complainant and "Resp." indicates respondent. *Through May 2008 and dispute DS374. 37 Table 5. Countries Using the ACWL to Initiate Sole-Complainant WTO Disputes, 2001-2008* Number of Prior Number of ACWL- Disputes as Sole Backed Disputes as Number of Prior Number of Prior Complainant that Sole Complainant that WTO Disputes Disputes as Sole Resulted in at Least Resulted in at Least a Country as Complainant Complainant a Panel Report Panel Report Bangladesh 0 0 0 0** Colombia 4 3 0 0 Ecuador 1 0 0 1** Guatemala 5 2 0 1 Honduras 4 0 0 1 India 11 8 3 2 Indonesia 2 0 0 1 Nicaragua 0 0 0 0** Pakistan 1 0 0 1** Panama 2 0 0 0 Peru 1 0 0 1 Philippines 2 1 1 0 Thailand 8 2 0 1 Source: Data compiled by the authors by matching public records from ACWL website with public information on WTO website. *Through May 2008 and dispute DS374. "Prior" is prior to country use of ACWL services as a complainant in DSU proceedings. **Indicates at least one additional sole-complainant dispute that resulted in a settlement notified to the WTO as a "mutually agreed upon solution." Sole complainant disputes tied into larger disputes pursued by other WTO members (for India: Turkey ­ Textiles; for Indonesia: Argentina -- Safeguard Measures on Imports of Footwear; for Panama: EC -- Regime for the Importation of Bananas; for Philippines: U.S. -- Import Prohibition of Certain Shrimp and Shrimp Products; for Thailand: EC -- Duties on Imports of Rice, Turkey -- Restrictions on Imports of Textile and Clothing Products, EC ­ Tariff Preferences in disputes prior to ACWL experience; and for Thailand: EC ­ Export Subsidies on Sugar and EC ­ Chicken Cuts in ACWL-backed disputes). 38 Table 6. Value of Market Access at Stake in ACWL versus non-ACWL WTO Disputes, AD Cases Involving Developing Country Complainants, 2001-2008 Average Value of Estimated Value of Complainant Exports Lost Exports due to WTO Dispute (Developing Country Complainant) in Three Years Prior to AD** AD* Non-ACWL Client Cases DS241: Argentina -- Poultry Anti-Dumping Duties (Brazil) $41,464,128 -$25,128,358 DS272: Peru -- Provisional Anti-Dumping Duties on Vegetable $11,000,726 -$9,720,227 Oils from Argentina (Argentina) DS288: South Africa -- Definitive Anti-Dumping Measures on $5,906,750 -$5,766,517 Blanketing from Turkey (Turkey) DS313: EC -- Anti-Dumping Duties on Certain Flat Rolled Iron $39,868,190 -$8,481,772 or Non-Alloy Steel Products from India (India) DS318: India -- Anti-Dumping Measures on Certain Products $3,072,471 -$1,432,583 from the Separate Customs Territory of Taiwan, Penghu, Kinmen and Matsu (Taiwan) DS355: Brazil -- Anti-Dumping Measures on Resins (Argentina) $71,215,545 -$69,672,704 Mean value in non-ACWL Cases: $28,754,635 -$20,033,693 ACWL Client Cases DS306: India -- Anti-Dumping Measure on Batteries from $315,430 -$315,430 Bangladesh (Bangladesh) DS312: Korea --Certain Paper (Indonesia) $42,136,886 -$3,853,435 DS327: Egypt -- Matches (Pakistan) $2,608,283 -$2,453,799 DS331: Mexico -- Steel Pipes and Tubes (Guatemala) $2,693,535 -$2,242,200 DS374: South Africa -- Anti-Dumping Measures on Uncoated $844,778 -$802,930 Woodfree Paper (Indonesia) Mean value in ACWL Cases: $9,719,782 -$1,933,559 Source: Data compiled by the authors. To make samples comparable, all disputes are over recently imposed AD measures against developing countries eligible for membership in the ACWL. Complainant exports are of 6-digit Harmonized System (HS) product subject to the AD import restriction; HS export data taken from WITS. *Average annual value in the three years prior to the AD investigation. Value of lost exports calculated as value of exports two years after the AD investigation minus the average annual exports in the three years prior to the AD. 39