Obstacles Facing Developing Countries in Antidumping Cases: The Path From Initial Filing to WTO Dispute Settlement

Obstacles Facing Developing Countries in Antidumping Cases: The Path From Initial Filing to WTO Dispute Settlement Chad P. Bown * Bernard Hoekman ** Caglar Ozden *** May 14, 2003 VERY PRELIMINARY: PLEASE DO NOT QUOTE OR CITE Abstract This paper examines the experience of developing countries in the US antidumping process. We trace their experience at each stage of the investigation process and through follow-up activity in disputes initiated at the GATT/WTO. A first pass at the data suggests evidence that the legalization of the process for fighting contingent protection may be particularly onerous for developing countries. * Department of Economics, Brandeis University. Email: cbown@brandeis.edu. ** World Bank, Development Research Group *** Emory University and World Bank, Development Research Group We would like to thank Thomas Prusa and Eric Reinhardt for their generosity in sharing their antidumping and GATT/WTO dispute data. The views expressed here are those of the authors and do not necessarily reflect those of their affiliated institutions. 1

I. Introduction Legalization has become one of the most visible trends in the international trading system in the last two decades. The shift towards a rule-based regime where policy is frequently determined as a result of litigation or a legal process is evident both at the domestic and multilateral levels. The most noticeable example at the domestic level is the increased use of antidumping actions which involve rather complicated legal and administrative rules and procedures, even if the underling motivations are primarily political. The counterpart at the multilateral level is WTO s Dispute Settlement Understanding (DSU) with its ever evolving case law and legal principles and move away from diplomacy and negotiation. One fundamental question is how these developments influence participation and integration of developing countries into the global trade regime. On one hand, some supporters of legalization argue power politics that dominate the trade policy agenda put developing countries at a disadvantage. Their rights are better protected under a legal system and they are able to integrate into the global economy more easily. On the other hand, a sophisticated legal environment requires a country have access to substantial financial and legal resources to effectively protect its rights. Developing countries in particular have both a limited legal capacity and familiarity with the complex domestic and international legal regimes, thus potentially putting them at a disadvantage in a legalized system. Furthermore, developed countries, such as the US and the EU, frequently modify their domestic statutes so that their protectionist policies become more difficult to surmount, both through domestic and multilateral channels. The intention of this paper is to investigate the experience of developing countries that are forced to fight off acts of developed country protection through a legalized process of contingent protection (antidumping) and potential dispute resolution. To address these issues, we examine two sets of related data. We first look at all antidumping cases initiated in the United States between 1979-98 and then turn to the GATT and WTO 2

dispute settlement activity relating to the United States use of its antidumping policies and actions. We investigate each stage of the legal process in both the domestic investigations and the GATT/WTO disputes and compare the experiences of developed and developing country exporters. For example, an antidumping petition against a foreign exporter typically results in one of three final outcomes: a duty applied against foreign firms, a rejection of the case, or a termination of the investigation due to a withdrawal or a settlement between the US petitioning industry and the foreign exporters. A negative decision can be taken at different stages by US International Trade Commission (ITC) or Department of Commerce (DOC) for lack evidence on dumping or injury. The evidence suggests that the ex ante probability of a positive injury decision is not substantially different for developing countries once the investigation starts. However the final duties imposed on developing country exports are much higher. Furthermore cases involving developing countries are more likely to be resolved at the later stages of the investigation and less frequently through settlements. These suggest that developing countries firms can not defend themselves as effectively as the firms from industrial countries do and have to wait for the decisions of the investigating authorities. 1 We also present evidence that the differences between industrial and developing countries continue at the multilateral forum as well. First, although they comprise almost two thirds of the positive findings in the US antidumping cases, developing countries have filed only one third of the total disputes against the US involving antidumping measures. Once filed, the disputes at the GATT/WTO can also result in a positive or a negative finding or end through a settlement. 2 Again, developing countries are less likely to settle and to obtain significant concessions at the end. 1 One emerging pattern in the 1990s is that newly industrialize countries (NICs) such as Korea, Singapore, Hong Kong and Taiwan appear to be acting more like industrial countries than other developing countries. 2 Of course, the key difference between antidumping decisions by the US government and panel decisions in a GATT (or WTO) panel is the issue of enforcement. Positive antidumping findings and the resulting duties are easily enforced by the US government while there is no guarantee that decisions of a dispute panel will be implemented. 3

Developing countries are also targeted more frequently in antidumping investigations, including the cases where the initiator is another developing country. One important reason is the inability of the developing countries to retaliate to such policies through their own antidumping investigations or other trade measures (Blonigen and Bown, 2003). In this paper, we provide evidence that the inability of the developing countries to effectively participate in such legal settings and to defend themselves may be a second reason. In other words, increased legalization of the trade system both at the domestic and multilateral levels might be creating new challenges for developing countries. The rest of this paper proceeds as follows. Section II discusses the antidumping filings in the United States and our evidence regarding the developing country experience. Section III examines the outcomes of US antidumping cases, and Section IV investigates the GATT/WTO trade disputes involving allegations of US misuse of its antidumping process, and the experience of developing countries in these disputes as well. Finally, section V concludes with a discussion of additional questions and the need for further research. II. Antidumping Filings in the United States During the last two decades, there has been significant developments in the use of antidumping actions, the most popular protectionist policy in use today. First, there has been an explosion in the total number of anti-dumping investigations filed. While the average annual number of antidumping initiations was 144 during 1980-1985, it reached 318 in 2000-2001 [Finger & Zlate, 2003]. Second, while it was used almost exclusively by the US, EU, Canada and Australia until 1985, antidumping policies started to spread to other countries in the 1990s [Miranda et al, 1998]. During 1995-2002, developing countries accounted for around 60% of all anti-dumping initiations. Finally, the targeted countries are now predominantly developing countries, regardless of the initiator. Developing countries initiate 70% of their investigations against other developing and transition 4

economies while this number is around 75% for industrial countries. The bias against importers from developing countries becomes more evident once we take the levels of trade into account. According to calculations by Finger & Zlate [2003], developing countries are six times more likely to be targeted by industrial countries and three times by other developing countries per unit of imports. The pattern in the United States is similar to other developed countries although administrative procedures in antidumping investigations differ. The first step in an investigation is the filing of a petition by individual firms claiming injury and the first critical question is which foreign firms/countries to name in the petition. This can be based on several factors, including the likelihood of the investigation leading to a positive dumping and injury determination as well as a petition resulting in the highest level of applied duties. Of course, settlement is an option and there is a large literature arguing how antidumping law can facilitate collusion between domestic and foreign firms [see Prusa, 1992 and Staiger and Wolak, 1994]. Settlement is more likely if domestic and foreign firms have been interacting for a long time and both sides are well prepared for litigation. We use Blonigen and Bown [2003] dataset to examine whether there is evidence of a bias against developing countries at the naming decision by the US firms. Their dataset includes every antidumping petition filed in the manufacturing sector, thus comprising 84% of all US cases over the 1980-1998 period. Their data includes every exporting country who has a market share larger than 3% in the products 3 that were listed in the petitions. This leads to 2015 potential targets (countries) of which 638 were named in the petitions. 3 3% is the non-negligibility rule of the WTO s Antidumping Code. 5

Table 1 lists some statistics from this dataset. Of the 2015 eligible targets (line 1, left hand side of the table), 1015 were developed countries and 1005 were developing 4. The probability of being named in a petition is 32% for developed country while it is 38% for a developing country. There is, however, a wide variance between different developing and developed countries. The EU and Japan are named 31% and 33% of the time, respectively, which are very close to the average for all countries. On the other hand, Canada has the lowest level with 14%. Among developing countries, China has the highest level with 66% while Mexico has only 22%. This bias against China and in favor of Mexico and Canada appear at every stage of the investigation. The rest of the table shows the ex ante probabilities of three categories of outcomes in a US AD petition: a negative finding, a withdrawal/settlement and positive finding. The probability of a positive finding is 44% for developing and 43% for developed countries. However, it is much lower for EU and Canada (at 34%) relative to Japan (58%). Among the developing countries, Mexico has the lowest probability with 32% and China has the highest with 63%. The settlement rates are also similar between developing and developed countries. However, the level for developing countries is inflated due to large number of settlements by Eastern Block countries during the 1980s on steel products. Another important feature of the US antidumping investigations is that more than half involve the steel sector. Since the unique characteristics of this sector may bias the numbers, the right hand side of the table also presents data on the non-steel petitions only. The qualitative pattern of our earlier results is quite similar; nevertheless, there are some interesting differences. First, the likelihood of settlement falls since cases involving the steel industry settle around 20% of the time. 5 The probability of positive duties is similar between developing (excluding China) and developed countries (excluding Japan) 4 Note the total share of developing countries in US imports is much lower than 50%. 5 The steel industry may be more likely to settle given the problem of overcapacity in the global steel industry and various US administration s attempts to address the overcapacity problem through managed trade negotiations. 6

at 38%. However, when steel is excluded, developed countries settle 9% of the time while developing countries do so only 2% of the time. Table 2 provides similar statistics for the 1990-1998 period. The reason we focus on this period separately is to identify potential new patterns in the data. New patterns in the data could be due to concerns for retaliation. Blonigen and Bown [2003] have noted that the proliferation of antidumping use by US trade partners in the early 1990s and filing of the first trade dispute involving antidumping against the US at the GATT in 1989 are potential explanations. First, the probability of being named in a petition increases for all countries, except for Canada and Mexico, which are probably receiving better treatment due to higher integration through NAFTA. Second, the ratio of settlements decline in the steel sector, but stays the same in non-steel cases. This is due to the fact that a larger portion of the countries named in the steel cases of this period are developing countries who are new in the markets and are less likely to settle. In non-steel cases, developed countries settle 11% of the time as opposed to 4% for developing countries. Third, the outcomes involving Asian NICs resemble more like the developed countries, probably due to their increased expertise in antidumping cases. III. The Stages and Outcomes of US Antidumping Investigations The antidumping statute in the US has been frequently modified over the last two decades to ease the use of antidumping for protectionist purposes and to comply with WTO s antidumping codes [Blonigen and Prusa, forthcoming]. Nevertheless, since 1980 there have been two agencies involved in the process. The Department of Commerce (DOC) makes the decision of whether the foreign firms accused in the petition have dumped, while the International Trade Commission (ITC) decides if the US petitioning industry has been injured. If the decision of both agencies is positive, antidumping duties can be 7

imposed. Following Murray [1991] and Prusa [1991], the typical steps in an investigation are: A petition is filed Department of Commerce (DOC) summary investigation to determine if there is adequate information in the petition to support the allegation of dumping International Trade Commission (ITC) preliminary investigation to see if there is reasonable indication of injury DOC investigation to determine if there is evidence of dumping and calculation of the dumping margin ITC final injury determination A case is dismissed if the DOC or the ITC reach a negative decision at any point in the process. The DOC investigation of dumping has actually two stages. The preliminary investigation aims to determine if there is reasonable likelihood of dumping. If this is positive, imports are cleared through customs under a suspension of liquidation, requiring the importer post a bond to enter the US market. If the preliminary decision is negative, the case continues but there no duties are imposed. Furthermore, the US petitioning industry can withdraw the case at any point; for example, if they reach an agreement with the exporter. Prusa [1992] has noted that such agreements are frequently exempt from antitrust investigations, which can lead to collusive effects of antidumping actions. We divided all of the antidumping petitions into the following categories, depending on at what stage they were resolved: 1 Terminated before ITC preliminary or Negative ITC preliminary 2 Terminated between ITC preliminary and DOC final 3 Negative DOC final no dumping 4 Terminated between DOC final and ITC final 8

5 Negative ITC final no injury 6 Affirmative ITC final positive injury A case is generally resolved at Outcome 1 if the petitioner does not provide a strong case or if the US industry withdraws the petition. Some of the withdrawals may also involve settlements, but it is less likely at this stage. Next, Outcome 2 is interpreted as an early settlement between the petitioner and the defendant. Outcome 3 is the ITA s dumping decision and this is almost always positive given the broad way the fair value and dumping margin calculations are carried out. Outcome 4 is a late settlement and may be more likely if the DOC estimates a high dumping margin. If the case is not dismissed or settled, it proceeds to the ITC for injury decision. If ITC finds in the affirmative, duties are imposed Outcome 6. Otherwise, no duties are imposed at Outcome 5. Table 3 provides data on all antidumping investigations initiated in the United States between 1979-98. 6 Of the 811 petitions, 585 cases involved a US industry filing multiple petitions against different foreign countries over production over the same products (thus a common injury determination). This table includes non-manufacturing cases as well, compared to the previous ones and provides a better sense of the overall antidumping panorama. Among the 811 cases filed, 179 (22%) ended in Outcome 1. The level likelihood of this outcome is 27% for developed countries but increases to 30% when Japan is excluded. There is a bias against developing countries, for which the likelihood is only 17%. This early termination likelihood further decreases to 15% when we exclude Mexico. Only 13% of the cases that survive the first stage are settled early on (Stage 2 in the table). This level is 16% for all developed countries and 21% when Japan is excluded. The European firms, especially in steel, comprise the biggest group among the early settlers. 6 We thank Tom Prusa for his generosity in providing this data. 9

The conditional probability of an early settlement for developing countries is much lower at 11%, even when Mexico is included. The next outcome to consider is a negative dumping finding by the DOC which is very low at 6%. As we stated above, the criteria on dumping are quite relaxed, especially in the last decade. It is possible to determine there is dumping even if foreign firms charge higher prices at home and their exports are profitable [Blonigen and Prusa, forthcoming]. Settlements at a later stage are more rare with 6% of the cases that survived the DOC investigation. Developed countries settle 7% of the time compared to 4% for developing countries. The final stage is injury determination by ITC which is positive for two thirds of the cases that reach this stage. The levels for developed and developing countries are quite close but there is again strong bias against Japan and China and in favor of Canada. The importance of settlements becomes clearer when we investigate the ratio of positive findings to total filings. In 45% of the cases (185 out of 411) developing countries face duties while developed countries do so only 36% (144 out of 400) of the time. The right hand side of the table presents data for the non-steel cases which comprise 46% of the total. Patterns are quite similar, with a few notable exceptions. First, the ratio of settlements is much lower. This is due to the fact that steel firms settle with higher frequency, especially in the 1980s, a fact we identified earlier. Also, dismissals in stage 1 is much higher for all countries indicating that steel firms generally file much better prepared petitions. However, the differences between developed and developing countries in stage 1 outcomes as well as settlements are present here. The biases against Japan and China extend to non-steel cases. Finally, we analyze cases for the period between 1990 and 1998 in Table 4 to detect new patterns. Several observations merit a discussion. First, the DOC rarely issues negative findings; therefore the ITC becomes the sole determinant of the success of a case, in 10

terms of the affirmative or negative decision. Of course, the DOC does still influence the final outcomes that result in duties, as they determine the level of applied duties. Second, the settlement ratio falls considerably, again mainly due to decline in the settlements in steel cases. Third, petitions against Canadian firms decline and most of these are dismissed. There are no antidumping duties imposed on Canadian firms in this period. Petitions against Mexico increase but are mostly dropped in stage 1. Fourth, almost the same ratio of cases are dismissed against developing and developed countries in stage 1. However, the increase for developing countries is driven by increases for the Asian NICs and Mexico. Otherwise the overall gap between other developing countries and developed countries remain, especially in non-steel cases. Another indicator of the obstacles faced by developing countries is the average tariffs imposed when the ITC finds injury. The duty is based on calculations of the DOC. There is a literature on how these calculations are performed, how biased they are against the foreign exporters, and Blonigen [2003] further discusses how the DOC s discretionary practices have changed over time. The point we want to emphasize is the information on which the determination is based. The DOC shall make a determination, based upon the best information available (Murray [1991]). For this purpose, DOC requests information from the foreign companies. If this is not available or not usable (i.e. in the format required etc.), DOC relies on the information provided by the petitioner, which needless to say, is likely to be biased. Murray [1991] explains the process in the following manner: Consider the problem facing the foreign respondent who receives a request for information from the DOC. It arrives in the form of a questionnaire, some 100 pages long, in English, requesting specific accounting data on individual sales in the home market (and possibly to third countries), data on sales to the United States, data needed to adjust arm s-length market prices to net ex-factory prices,,,,, adjustments for taxes and duties on imported inputs, adjustments for exporters sales prices, international shipping costs, distribution costs in the US, and a host of other details. There must be enough information for the DOC to investigate nearly every US sale (that is very transaction) for a period of six months. All this information must be identified, retrieved, recorded, and then transmitted to 11

the DOC in English on hard copy and in computer readable format (compatible with their system) within the short deadline stipulated under the US statutes. Many firms from developing countries have difficulty complying with these requirements. As a result, the DOC bases its decisions on the best information available which, in most cases, is strongly biased against the defendant. This is evident in table 5 which lists the average duties imposed in cases that are resolved in stage 6. For all cases and all countries, the average duty was 46%. We should note this is a simple numerical average which does not take affected import volume into account. The average tariffs are 42% and 49%, respectively for developed and developing countries. For non-steel cases, they are 52% and 54% respectively. However, when we exclude Japan, average tariff for developed countries falls to 26% for all cases and 23% in non-steel cases. Furthermore, when we exclude Asian NICs and Mexico, the average tariffs on developing countries increase to 67% for all cases and 77% for non-steel cases. The same pattern exists for the 1990s which is presented in the second row of the table. The average tariffs increase to 58% in the whole sample and 68% in non-steel cases, evidence of the increased role of antidumping as a protectionist device. The gaps between developing and developed countries continue although still somewhat masked by the bias against Japanese and Chinese firms. The average tariff in non-steel cases against developed countries, excluding Japan, is 23%. The same number for developing countries, except Asian NICs, Mexico and China (which seems to face the highest duties by a wide margin) is 69%. In short, the inability of the developing countries to comply with the administrative rules manifests itself very clearly in the antidumping margin calculations. IV. Antidumping disputes in the GATT/WTO 12

Parallel to this growing popularity of antidumping as a domestic trade policy measure, there has been an increased attention paid to antidumping in the GATT/WTO. Major amendments were introduced in the Tokyo Round which made the dumping allegations easier to prove [Blonigen and Prusa, forthcoming]. However, the efforts to reign in the spread of antidumping through WTO codes did not seem to be very effective. The main recourse to the targets of antidumping duties has been to file a dispute. Since 1979, total of 560 disputes were filed (at the GATT and WTO) and 105 of them (19%) involved an antidumping issue. 7 The disputes where the United States is the defendant provide a more interesting pattern. Under the GATT regime, there were 86 disputes filed against the US and 25 of them (29%) involved an antidumping issue. However, under the WTO, there had been 80 disputes filed and 33 of them (41%) were related to antidumping. Table 6 provides data on disputes filed against the US from 1979 until 2002 that involve antidumping policies and actions. A total of 58 cases were filed, of which 34 were by developed countries. Of these cases, 41 concerned specific antidumping actions of the US while the rest were about general US laws and statutes. The most infamous dispute was filed against the US by 11 countries over the Byrd Amendment which required tariff revenue in US antidumping cases be refunded to the US petitioning industry. 8 Overall, only one third of the 41 disputes were filed by developing countries although they faced two thirds of all antidumping duties imposed by the US. And these cases were typically filed by only a handful of developing countries Brazil, Chile, India, Korea and Mexico. Because it was not a WTO member until recently, China has never filed a dispute against the US although it has targeted more frequently than any other developing country. Similarly Japan has filed few cases although it has also faced many antidumping investigations. On the other hand, Canada has been very aggressive in filing disputes. The reluctance of developing countries to continue to pursue their rights in the multilateral 7 Based on the dataset used by Busch and Reinhardt (forthcoming). 8 See WTO disputes DS217 and DS234, United States: Continued Dumping & Subsidy Offset Act of 2000. 13

forum is one reason why they might be more targeted for antidumping investigations in the first place. The GATT and WTO dispute settlement procedures are as legally complex as the domestic antidumping investigations. Furthermore, many procedures were modified and new ones were added under the WTO, especially with respect to appeals and adoption of panel decisions. 9 In general, a case can result in a withdrawal, settlement, or a panel decision which can also be appealed. Finally, under the WTO, an arbitration process can authorize retaliation in cases in which the defendant country has refused to bring its policies into conformity with a panel request and thus its obligations under the WTO. Nevertheless, there is no guarantee that a decision that passes the appeal will actually be implemented, as the self-enforcing nature of the GATT/WTO system requires plaintiff have the capacity to retaliate. Table 6 notes that 25 of the 41 cases had concluded as of the end of 2002. Among these, 14 had a panel decide on them while 11 were settled without a panel decision. A majority of the disputes filed by developed countries were settled without a panel (8 out of 15) while the settlement ratio is much lower for developing countries (4 out of 10). This is similar to the outcomes in domestic antidumping cases where settlement rate is much lower for developing countries. However, settlement may be more important in the GATT/WTO context, as Busch and Reinhardt [2000] argue that early settlement generally leads to highest likelihood of concessions from a defendant. One argument is that defendants are more likely to abide by their responsibilities if it comes out of a voluntary settlement as opposed to a panel decision. On the other hand, a WTO panel decision may be necessary to provide a domestic government with the commitment 9 See Jackson [1997] for detailed description of the WTO s Dispute Settlement Understanding. 14

power that it would otherwise lack to convince its private sector of the seriousness of the consequences of the failure to give concessions and engage in policy reform. 10 The critical question is why developing countries settle less frequently? A plaintiff and a defendant are more likely to settle if they can credibly threaten to complete the case and enforce a punishment. The increased legal complexity of the environment, especially under the WTO, imposes relatively higher burdens on developing countries compared to industrial ones. As a result US settles with developing countries in fewer cases and offers fewer concessions. In the subsample from the Busch and Reinhardt data that applies to US cases, 7 out 11 settled cases lead to significant concessions by the United States while the same number is 5 out of 14 for cases where there was a panel decision. As a result, developing countries received concessions in fewer cases then the developed countries. If we view WTO disputes as the next step after domestic antidumping investigations (for example, kind of an appeal process), we see developing countries are generally less successful. This is due to fewer settlements and obtaining of fewer concessions as a consequence. And naturally, the lower prospects for them at this stage leads the US to target them more frequently in the first place. This is consistent with the evidence provided in Blonigen and Bown [2003], who found that the capacity of a foreign country to engage in a GATT/WTO trade dispute against the US and potentially retaliate affected the decision of whether to give antidumping protection at all. 11 10 Bown (2003) interprets dispute settlement activity as an area where defendant governments appeal to the GATT/WTO for the commitment power necessary to engage their domestic constituencies in reform. Trade theorists such as Staiger and Tabellini (1987) and Maggi and Rodriguez-Clare (1998) have suggested that one role for the GATT/WTO is to provide domestic governments with such commitment power. 11 This is also consistent with the results of Bown (forthcoming), who studies the choice of governments giving protection through the GATT safeguards provisions as opposed to through some GATT-illegal policy, such as a dubious antidumping measure. In this study, the protection-imposing country was less likely to impose a policy such as a GATT-illegal AD measure when the affected trading partner had a substantial capacity to retaliate. 15

V. Conclusion In both domestic antidumping investigations and GATT/WTO disputes filed against the US on antidumping related issues, we see similar patterns with respect to developing countries. They are targeted more frequently in domestic investigations and much more likely to face duties. The relative success of developed countries, in general, does not occur at the injury determination stage but in their ability to settle the case early on. Furthermore, the average duties imposed on exports from developing countries are significantly higher. We argue that these are generally due to the inability of the developing countries firms to comply with the legal and administrative complexities of the investigation. In other words, legalization of the process imposes larger burdens on developing countries. Similar problems arise in the GATT/WTO dispute settlement cases against the US involving antidumping related issues. Developing countries file fewer cases (even though they may have more reasons to do so), settle them less frequently and receive smaller concessions. These are all parallel to the outcomes in domestic antidumping investigations indicating that the same legal capacity problems exist in this forum as well. 16

References Blonigen, Bruce and Thomas Prusa [forthcoming] Antidumping. in James Harrigan, ed. Handbook of International Trade, Basil Blackwell. Blonigen, Bruce A. [2003] Evolving Discretionary Practices of U.S. Antidumping Activity NBER Working paper, No. 9625, April. Blonigen, Bruce A. and Chad P. Bown [2003] Antidumping and Retaliation Threats Journal of International Economics, 60 (2): 249-273. Boltuck, Richard and Robert Litan (eds.) [1991] Down in the Dumps: Administration of the Unfair Trade Laws, The Brookings Institution, Washington DC. Bown, Chad P. [forthcoming] Trade Disputes and the Implementation of Protection under the GATT: An Empirical Assessment, Journal of International Economics. Bown, Chad P. [2003] On the Economic Success of GATT/WTO Dispute Settlement, Brandeis University manuscript, January. Busch, Marc and Eric Reinhardt [2000], Bargaining in the Shadow of Law: Early Settlement in the GATT/WTO Disputes, Fordham International Law Journal, vol.24, p.158-72. Busch, Marc and Eric Reinhardt [forthcoming], Developing Countries and GATT/WTO Dispute Settlement, Journal of World Trade Finger and Zlate [2003] Jackson, John H. [1997] The World Trading System: Law and Policy of International Economic Relations. (2 nd edition) Cambridge, MA: MIT Press. Maggi, Giovanni and Andrés Rodríguez-Clare. [1998] The Value of Trade Agreements in the Presence of Political Pressures, Journal of Political Economy, 106(3): 574-601. Murray, Tracy [1991], The Administration of the Antidumping Duty Law by the Department of Commerce in Richard Boltuck and Robert Litan (eds.) Down in the Dumps: Administration of the Unfair Trade Laws, The Brookings Institution, Washington DC. 17

Miranda, Jorge, Raul A. Torres and Mario Ruiz [1998]. The International Use of Antidumping: 1987-1997. Journal of World Trade, 32(5): 5-71. Prusa, Thomas J. [1991] The Selection of Antidumping Cases for ITC Determination in R.E. Baldwin, ed. Empirical Studies of Commercial Policy. Chicago: University of Chicago Press. Prusa, Thomas J. [1992] Why are So Many Antidumping Petitions Withdrawn? Journal of International Economics, 33(1): 1-20. Staiger, Robert W. and Guido Tabellini. [1987] Discretionary Trade Policy and Excessive Protection, American Economic Review, 77 (5): 823-37. Staiger, Robert W. and Frank Wolak. [1994] Measuring Industry Specific Protection: Antidumping in the United States. Brookings Papers on Economic Activity: Microeconomics, 51-118. 18