《国际政治经济学》文献资料（Political Economy of International Trade：International Influences）Overlapping institutions, forum shopping, and dispute settlement in international trade

Institutions,Forum Shopping,and Trade Disputes 739 Another key consideration would be differences in law.There clearly are differ- ences in law across the WTO and NAFTA,for example,but there is also a marked degree of convergence in many areas,not least because successive rounds of multi- lateral obligations have been ratified domestically.As a result,disputes are typi- cally "actionable"under either forum.Indeed,NAFTA directly incorporates various WTO provisions and has been called on to adjudicate these where the law over- laps.For example,in the safeguard dispute,Mexico not only argued that the main legal issues arose under both agreements,but that NAFTA"necessarily [had]juris- diction to dispose of all overlapping GATT issues involved in that dispute."5 A third factor would be the timeliness of proceedings.On this count there is little variation:both fora set out similar timelines,from the duration of consulta- tions to the issuance of an interim and final report.That said,if the timeliness of justice were a main determinant of the choice of forum,complainants might be expected to prefer arbitration to formal dispute settlement(that is,alternative dis- pute resolution mechanisms under NAFTA or DSU Article 25 arbitration under the WTO),and yet they almost never do.16 A fourth factor would be the adoption of panel reports,and appellate review.The FTA/NAFTA never permitted defendants to block panel reports,a feature that com- pared favorably to GATT,but has been woven into the WTO.Even under GATT, there was far less blocking of reports than of panel requests;but in any case,Can- ada and the United States had no history of denying the other a legal victory.Appel- late review is another feature that distinguishes the WTO from NAFTA,although the trade-off is that WTO verdicts are "binding,"whereas NAFTA rulings are not, such that this feature of the DSU is not as germane to regional dispute settlement. Following from this,a fifth factor would be the availability of remedies at the end of a dispute.The WTO and NAFTA outline similar procedures whereby a wronged complainant might seek redress,though NAFTA provisions,on occa- sion,allow for more direct compensation,notably the "cultural industries excep- tion"discussed in the case study on Canada-Periodicals.Otherwise,complainants at either forum must wait out the defendant's (prospective)compliance.If this is not forthcoming in a timely manner,the complainant can request authorization to suspend"equivalent"benefits to the defendant.Few such requests have been made at either the WTO or NAFTA,and fewer still have been acted upon,7 suggesting 15.NAFTA Doc.USA-97-2008-01,para.28.All NAFTA documents are available from the NAFTA Secretariat Web site:(http://www.nafta-sec-alena.org). 16.Three cases have gone for DSU 25 arbitration at the multilateral level,two under GATT and one under the WTO,although the latter was conducted as a DSU 22.6 proceeding on authorization to retaliate,such that it can hardly be counted as a case of arbitration,as per the spirit of this provision. Nonetheless,recourse to arbitration under NAFTA or DSU 25 still hinges on a prior decision to file regionally or multilaterally,bringing it within reach of my argument.Indeed,as in the case of DSU 25, these decisions are fully backed by DSU 21 (on compliance)and 22 (retaliation),such that concern for precedent should be no less pressing in these disputes. 17.Authorization to retaliate has been granted in only seven WTO disputes,and acted upon in four of these

Another key consideration would be differences in law+ There clearly are differ￾ences in law across the WTO and NAFTA, for example, but there is also a marked degree of convergence in many areas, not least because successive rounds of multi￾lateral obligations have been ratified domestically+ As a result, disputes are typi￾cally “actionable” under either forum+ Indeed, NAFTA directly incorporates various WTO provisions and has been called on to adjudicate these where the law over￾laps+ For example, in the safeguard dispute, Mexico not only argued that the main legal issues arose under both agreements, but that NAFTA “necessarily @had# juris￾diction to dispose of all overlapping GATT issues involved in that dispute+”15 A third factor would be the timeliness of proceedings+ On this count there is little variation: both fora set out similar timelines, from the duration of consulta￾tions to the issuance of an interim and final report+ That said, if the timeliness of justice were a main determinant of the choice of forum, complainants might be expected to prefer arbitration to formal dispute settlement ~that is, alternative dis￾pute resolution mechanisms under NAFTA or DSU Article 25 arbitration under the WTO!, and yet they almost never do+ 16 A fourth factor would be the adoption of panel reports, and appellate review+ The FTA0NAFTA never permitted defendants to block panel reports, a feature that com￾pared favorably to GATT, but has been woven into the WTO+ Even under GATT, there was far less blocking of reports than of panel requests; but in any case, Can￾ada and the United States had no history of denying the other a legal victory+ Appel￾late review is another feature that distinguishes the WTO from NAFTA, although the trade-off is that WTO verdicts are “binding,” whereas NAFTA rulings are not, such that this feature of the DSU is not as germane to regional dispute settlement+ Following from this, a fifth factor would be the availability of remedies at the end of a dispute+ The WTO and NAFTA outline similar procedures whereby a wronged complainant might seek redress, though NAFTA provisions, on occa￾sion, allow for more direct compensation, notably the “cultural industries excep￾tion” discussed in the case study on Canada—Periodicals+ Otherwise, complainants at either forum must wait out the defendant’s ~prospective! compliance+ If this is not forthcoming in a timely manner, the complainant can request authorization to suspend “equivalent” benefits to the defendant+ Few such requests have been made at either the WTO or NAFTA, and fewer still have been acted upon, 17 suggesting 15+ NAFTA Doc+ USA-97-2008-01, para+ 28+ All NAFTA documents are available from the NAFTA Secretariat Web site: ^http:00www+nafta-sec-alena+org&+ 16+ Three cases have gone for DSU 25 arbitration at the multilateral level, two under GATT and one under the WTO, although the latter was conducted as a DSU 22+6 proceeding on authorization to retaliate, such that it can hardly be counted as a case of arbitration, as per the spirit of this provision+ Nonetheless, recourse to arbitration under NAFTA or DSU 25 still hinges on a prior decision to file regionally or multilaterally, bringing it within reach of my argument+ Indeed, as in the case of DSU 25, these decisions are fully backed by DSU 21 ~on compliance! and 22 ~retaliation!, such that concern for precedent should be no less pressing in these disputes+ 17+ Authorization to retaliate has been granted in only seven WTO disputes, and acted upon in four of these+ Institutions, Forum Shopping, and Trade Disputes 739

740 International Organization that remedies,per se,are doubtful to be the deciding issue in the complainant's choice of forum. This discussion suggests that there is ample opportunity for forum shopping: both NAFTA and the WTO recognize that this can (and will)occur,and their rules and procedures are rarely so different as to prejudge the complainant's choice of "court."This,in turn,begs the obvious question:why not file at both institutions? NAFTA,for one,explicitly discourages this:Article 2005,paragraph 6,says that "the forum selected shall be used to the exclusion of the other...."Still,some might argue that,in practice,multiple filings do happen,as in the most recent iteration of the Canada-U.S.softwood lumber dispute,and Canada-Agricultural Products.In fact,neither bears this out:in softwood,most of the disputes filed at NAFTA concerned antidumping and countervailing duties,which fall outside the institution's main dispute settlement mechanism (that is,the one comparable to the DSU):19 in Canada-Agricultural Products,the United States went to the WTO to challenge Canadian efforts to comply with NAFTA.20 While the article antici- pates the circumstances under which a complainant might prefer to file in multiple fora,the point is that,as Marceau explains,"s]tates do not pursue multiple dis- pute settlement proceedings needlessly,working instead towards ensuring that their grievances are brought before the most appropriately equipped fora for settling their disputes."2 How,then,does a complainant select the most appropriate forum? The Argument The complainant's choice of forum depends on whether it prefers to set a regional or multilateral precedent,or no precedent at all.By setting a precedent,I mean adding to an institution's body of case law that is followed by its judicial bodies when ruling on subsequent disputes.This definition is not controversial;observers widely subscribe to the view that,as Palmeter and Mavroidis explain,"parties will continue to cite prior reports to panels,and panels will continue to take them into account by adopting their reasoning-in effect,following precedent."22 18.The article further states that "unless a Party makes a request pursuant to paragraphs 3 or 4," which invoke the special claims of jurisdiction on the environment,sanitary and phytosanitary mea- sures,and standards,although these paragraphs also insist on pursuing the matter "solely under this Agreement." 19.At NAFTA,disputes over antidumping and countervailing duties are taken up under Chapter 19,which gives the complainant (including firms,who have private standing)"binational"tribunals as an alternative to seeking relief in domestic courts.In this sense,Chapter 19 is not considered compa- rable to WTO dispute settlement. 20.In other words,this was not an example of the same case being filed twice,but a second dispute arising over the outcome of the first dispute. 21.Marceau2001,1081. 22.Palmeter and Mavroidis 2004,56.See also Davey 1998,79;Jackson 1998,83;Komuro 1995, 37;Petersmann 1994,1175;Huntington 1993,435;and Van Bael 1988,69

that remedies, per se, are doubtful to be the deciding issue in the complainant’s choice of forum+ This discussion suggests that there is ample opportunity for forum shopping: both NAFTA and the WTO recognize that this can ~and will! occur, and their rules and procedures are rarely so different as to prejudge the complainant’s choice of “court+” This, in turn, begs the obvious question: why not file at both institutions? NAFTA, for one, explicitly discourages this: Article 2005, paragraph 6, says that “the forum selected shall be used to the exclusion of the other+++ +”18 Still, some might argue that, in practice, multiple filings do happen, as in the most recent iteration of the Canada–U+S+ softwood lumber dispute, and Canada—Agricultural Products+ In fact, neither bears this out: in softwood, most of the disputes filed at NAFTA concerned antidumping and countervailing duties, which fall outside the institution’s main dispute settlement mechanism ~that is, the one comparable to the DSU!; 19 in Canada—Agricultural Products, the United States went to the WTO to challenge Canadian efforts to comply with NAFTA+ 20 While the article antici￾pates the circumstances under which a complainant might prefer to file in multiple fora, the point is that, as Marceau explains, “@s#tates do not pursue multiple dis￾pute settlement proceedings needlessly, working instead towards ensuring that their grievances are brought before the most appropriately equipped fora for settling their disputes+”21 How, then, does a complainant select the most appropriate forum? The Argument The complainant’s choice of forum depends on whether it prefers to set a regional or multilateral precedent, or no precedent at all+ By setting a precedent, I mean adding to an institution’s body of case law that is followed by its judicial bodies when ruling on subsequent disputes+ This definition is not controversial; observers widely subscribe to the view that, as Palmeter and Mavroidis explain, “parties will continue to cite prior reports to panels, and panels will continue to take them into account by adopting their reasoning—in effect, following precedent+”22 18+ The article further states that “unless a Party makes a request pursuant to paragraphs 3 or 4,” which invoke the special claims of jurisdiction on the environment, sanitary and phytosanitary mea￾sures, and standards, although these paragraphs also insist on pursuing the matter “solely under this Agreement+” 19+ At NAFTA, disputes over antidumping and countervailing duties are taken up under Chapter 19, which gives the complainant ~including firms, who have private standing! “binational” tribunals as an alternative to seeking relief in domestic courts+ In this sense, Chapter 19 is not considered compa￾rable to WTO dispute settlement+ 20+ In other words, this was not an example of the same case being filed twice, but a second dispute arising over the outcome of the first dispute+ 21+ Marceau 2001, 1081+ 22+ Palmeter and Mavroidis 2004, 56+ See also Davey 1998, 79; Jackson 1998, 83; Komuro 1995, 37; Petersmann 1994, 1175; Huntington 1993, 435; and Van Bael 1988, 69+ 740 International Organization

Institutions,Forum Shopping,and Trade Disputes 741 Steinberg concurs,arguing that the importance of a ruling "lies not only in its implications for the national measures that are the subject of the decision,but also in its precedential value."2 Bhala,in a series of seminal articles on the subject,4 calls this de facto stare decisis,which,in contrast to de jure stare decisis,is fol- lowed for extra-and quasi-legal factors,including custom and habit,rather than as a matter of legal requirement.25 Importantly,Bhala argues that de facto and de jure stare decisis are both "binding"sources of law,in that they establish a pre- sumption that precedents will be followed in the future.6 The Appellate Body (AB)bears out this view,arguing in a highly important ruling that"[t]he reason- ing in our Report in United States-Shrimp on which the Panel relied was not dicta;it was essential to our ruling.The Panel was right to use it,and right to rely on it.Nor are we surprised that the panel made frequent references to our Report in United States-Shrimp.Indeed,we would have expected the Panel to do so."27 NAFTA panels similarly follow a de facto stare decisis,as in the safeguards dis- pute over Mexican broom corn brooms,where the institution's prior ruling in Canada-Agricultural Tariffs was central in this case.28 Equally telling,these bod- ies of jurisprudence are well known and,at the multilateral level,meticulously catalogued in the WTO Analytical Index,which identifies the key precedents per- taining to each agreement,setting out,for example,that the AB's ruling in Canada- Autos is the relevant case law in distinguishing a de facto from a de jure subsidy. In short,as Bhala concludes,it is only "[o]ur intellectual rigidity [that]precludes us from admitting openly that the holdings of the Appellate Body-and,for that matter,panel-reports actually are a source of international law."30 Since de facto stare decisis is followed for extra-and quasi-legal reasons,includ- ing custom and habit,and not as a matter of legal requirement,it could be argued that precedents from one institution might constrain or shape the rulings of another, making it hard to disentangle different bodies of jurisprudence,and thus to engage in forum shopping.This concern is unwarranted,both in theory and in practice. For its part,the WTO has not cited a single ruling by NAFTA(or any other regional institution),3!not least because not all WTO members are NAFTA members,and thus are not bound by the precedents of this institution.This logic comes to the fore in Argentina-Poultry Antidumping Duties,for example,where the panel found Mercosur's ruling irrelevant to its deliberations.Building on the AB's ruling in 23.Steinberg 2004,254. 24.Bhala2001,1999b,and1999a. 25.Bhala1999b,3-4. 26.Ibid,9. 27.WTO Doc.WT/DS58/AB/RW,para.107.All WTO documents are available at(http://www. wto.org).Accessed 12 June 2007. 28.NAFTA Doc.USA-97-2008-01,para.44. 29.WT02003. 30.Bhala1999a,850-51. 31.In my effort to survey WTO reports and trade law experts on citations to NAFTA case law, I was unable to turn up a single example

Steinberg concurs, arguing that the importance of a ruling “lies not only in its implications for the national measures that are the subject of the decision, but also in its precedential value+”23 Bhala, in a series of seminal articles on the subject, 24 calls this de facto stare decisis, which, in contrast to de jure stare decisis, is fol￾lowed for extra- and quasi-legal factors, including custom and habit, rather than as a matter of legal requirement+ 25 Importantly, Bhala argues that de facto and de jure stare decisis are both “binding” sources of law, in that they establish a pre￾sumption that precedents will be followed in the future+ 26 The Appellate Body ~AB! bears out this view, arguing in a highly important ruling that “@t#he reason￾ing in our Report in United States—Shrimp on which the Panel relied was not dicta; it was essential to our ruling+ The Panel was right to use it, and right to rely on it+ Nor are we surprised that the panel made frequent references to our Report in United States—Shrimp+ Indeed, we would have expected the Panel to do so+”27 NAFTA panels similarly follow a de facto stare decisis, as in the safeguards dis￾pute over Mexican broom corn brooms, where the institution’s prior ruling in Canada—Agricultural Tariffs was central in this case+ 28 Equally telling, these bod￾ies of jurisprudence are well known and, at the multilateral level, meticulously catalogued in the WTO Analytical Index, which identifies the key precedents per￾taining to each agreement, setting out, for example, that the AB’s ruling in Canada— Autos is the relevant case law in distinguishing a de facto from a de jure subsidy+ 29 In short, as Bhala concludes, it is only “@o#ur intellectual rigidity @that# precludes us from admitting openly that the holdings of the Appellate Body—and, for that matter, panel—reports actually are a source of international law+”30 Since de facto stare decisis is followed for extra- and quasi-legal reasons, includ￾ing custom and habit, and not as a matter of legal requirement, it could be argued that precedents from one institution might constrain or shape the rulings of another, making it hard to disentangle different bodies of jurisprudence, and thus to engage in forum shopping+ This concern is unwarranted, both in theory and in practice+ For its part, the WTO has not cited a single ruling by NAFTA ~or any other regional institution!, 31 not least because not all WTO members are NAFTA members, and thus are not bound by the precedents of this institution+ This logic comes to the fore in Argentina—Poultry Antidumping Duties, for example, where the panel found Mercosur’s ruling irrelevant to its deliberations+ Building on the AB’s ruling in 23+ Steinberg 2004, 254+ 24+ Bhala 2001, 1999b, and 1999a+ 25+ Bhala 1999b, 3–4+ 26+ Ibid+, 9+ 27+ WTO Doc+ WT0DS580AB0RW, para+ 107+ All WTO documents are available at ^http:00www+ wto+org&+ Accessed 12 June 2007+ 28+ NAFTA Doc+ USA-97-2008-01, para+ 44+ 29+ WTO 2003+ 30+ Bhala 1999a, 850–51+ 31+ In my effort to survey WTO reports and trade law experts on citations to NAFTA case law, I was unable to turn up a single example+ Institutions, Forum Shopping, and Trade Disputes 741

742 International Organization Japan-Alcoholic Beverages,which reasons that only adopted reports should carry weight because they are "endorsed"by all the members (which is not true of unadopted WTO reports,much less NAFTA reports),2 and the requirement in DSU 3.2 for WTO bodies to interpret the institution's agreements for themselves,33 the panel explained that "we see no reason at all why we should be bound by the rulings of non-WTO dispute settlement bodies."34 More generally,Pauwelyn points out that,even when hearing cases covering claims similar to those before NAFTA (that is,softwood lumber and sweeteners),WTO bodies have not even mentioned these regional proceedings,much less been guided by the decisions that result.35 In a few NAFTA disputes,on the other hand,GATT/WTO decisions have been referenced by panels,not least because the agreement incorporates certain WTO articles.Even here,however,these references have simply been of suggestive value to the panel,much like at the U.S.Supreme Court,where,as in Lawrence v.Texas, decisions by the European Court of Human Rights were cited,despite the fact that these have no bearing on the high court's de jure stare decisis.36 Also like at the U.S.Supreme Court,moreover,NAFTA's references to the WTO have been ad hoc and controversial,rather than setting expectations about the institution's de facto stare decisis.Indeed,even NAFTA's body of jurisprudence is far from porous: in broom corn broom,for example,the panel explained that,since the rule in ques- tion was similar across the two institutions,"the Panel chose to rest its decision entirely on NAFTA Annex 803.3(12),without relying on Article 3.1 of the WTO Safeguards Code"or its attending case law.37 The WTO panel in Mexico-Soft Drinks could not agree more that NAFTA should take this tack,reasoning that "any findings made by this Panel,as well as its conclusions and recommendations in the present case,only relate to Mexico's rights and obligations under the WTO covered agreements,and not to its rights and obligations under other international agreements,such as the NAFTA,or other rules of international law."38 In short,one can speak meaningfully about separate bodies of de facto stare decisis at the WTO and NAFTA.39 Other scholars also argue that precedent matters 32.WTO Doc.WT/DS8/AB/R.14. 33.See Mavroidis,Howse,and Bermann forthcoming,63. 34.WTO Doc.WT/DS241/R,para.7.41.Likewise,in EC-Biotech Products,the panel explains that both the Convention on Biological Diversity and the Biosafety Protocol are "not applicable"for this same reason.WTO Doc.WT/DS291/R,para.7.75. 35.Pauwelyn2006,202. 36.See Joan Biskupic,"Supreme Court Citing More Foreign Cases,"USA Today.7 July 2003,avail- able at (http://www.usatoday.com/news/washington/2003-07-07-foreign-usat_x.htm),accessed 18 April 2007;and Jeffrey Toobin,"Swing Shift:How Anthony Kennedy's Passion for Foreign Law Could Change the Supreme Court,"New Yorker,12 September 2005,available at (http://www.newyorker.com/ archive/2005/09/12/050912fa_fact),accessed 18 April 2007. 37.NAFTA Doc.USA-97-2008-01,para.50. 38.WTO Doc.WT/DS308/R,para.7.15. 39.In discussing the model,and Figure 1,in particular,I will show that,even if there were some unanticipated spillover of precedent from the WTO to NAFTA,the complainant would still engage in forum shopping to discriminate among memberships,and would be more inclined to file at NAFTA as a result

Japan—Alcoholic Beverages, which reasons that only adopted reports should carry weight because they are “endorsed” by all the members ~which is not true of unadopted WTO reports, much less NAFTA reports!, 32 and the requirement in DSU 3+2 for WTO bodies to interpret the institution’s agreements for themselves, 33 the panel explained that “we see no reason at all why we should be bound by the rulings of non-WTO dispute settlement bodies+”34 More generally, Pauwelyn points out that, even when hearing cases covering claims similar to those before NAFTA ~that is, softwood lumber and sweeteners!, WTO bodies have not even mentioned these regional proceedings, much less been guided by the decisions that result+ 35 In a few NAFTA disputes, on the other hand, GATT0WTO decisions have been referenced by panels, not least because the agreement incorporates certain WTO articles+ Even here, however, these references have simply been of suggestive value to the panel, much like at the U+S+ Supreme Court, where, as in Lawrence v. Texas, decisions by the European Court of Human Rights were cited, despite the fact that these have no bearing on the high court’s de jure stare decisis+ 36 Also like at the U+S+ Supreme Court, moreover, NAFTA’s references to the WTO have been ad hoc and controversial, rather than setting expectations about the institution’s de facto stare decisis+ Indeed, even NAFTA’s body of jurisprudence is far from porous: in broom corn broom, for example, the panel explained that, since the rule in ques￾tion was similar across the two institutions, “the Panel chose to rest its decision entirely on NAFTA Annex 803+3~12!, without relying on Article 3+1 of the WTO Safeguards Code” or its attending case law+ 37 The WTO panel in Mexico—Soft Drinks could not agree more that NAFTA should take this tack, reasoning that “any findings made by this Panel, as well as its conclusions and recommendations in the present case, only relate to Mexico’s rights and obligations under the WTO covered agreements, and not to its rights and obligations under other international agreements, such as the NAFTA, or other rules of international law+”38 In short, one can speak meaningfully about separate bodies of de facto stare decisis at the WTO and NAFTA+ 39 Other scholars also argue that precedent matters 32+ WTO Doc+ WT0DS80AB0R, 14+ 33+ See Mavroidis, Howse, and Bermann forthcoming, 63+ 34+ WTO Doc+ WT0DS2410R, para+ 7+41+ Likewise, in EC—Biotech Products, the panel explains that both the Convention on Biological Diversity and the Biosafety Protocol are “not applicable” for this same reason+ WTO Doc+ WT0DS2910R, para+ 7+75+ 35+ Pauwelyn 2006, 202+ 36+ See Joan Biskupic, “Supreme Court Citing More Foreign Cases,” USA Today+ 7 July 2003, avail￾able at ^http:00www+usatoday+com0news0washington02003-07-07-foreign-usat_x+htm&, accessed 18 April 2007; and Jeffrey Toobin, “Swing Shift: How Anthony Kennedy’s Passion for Foreign Law Could Change the Supreme Court,” New Yorker, 12 September 2005, available at ^http:00www+newyorker+com0 archive020050090120050912fa_fact&, accessed 18 April 2007+ 37+ NAFTA Doc+ USA-97-2008-01, para+ 50+ 38+ WTO Doc+ WT0DS3080R, para+ 7+15+ 39+ In discussing the model, and Figure 1, in particular, I will show that, even if there were some unanticipated spillover of precedent from the WTO to NAFTA, the complainant would still engage in forum shopping to discriminate among memberships, and would be more inclined to file at NAFTA as a result+ 742 International Organization

Institutions,Forum Shopping,and Trade Disputes 743 in trade law and policy,including in studies of forum shopping.4 The novel con- tribution of this article is that I specify the precise circumstances under which precedent will lead a complainant to file its dispute in one or the other forum,or to avoid filing at all. It is logical that the complainant will file in whichever court is most likely to come closest to its "ideal"ruling.Indeed,if the complainant cared only about winning a legal victory over the defendant,this would be the end of the story.The problem,however,is that precedent ensures that a legal victory does not just apply to the complainant's trade with the defendant,but also to its trade with other mem- bers of the institution in which the resulting case law was set.Here,precedent can be a double-edged sword:it can make it easier for the complainant to bring litiga- tion against these other members of the institution,but it can also help these other members bring litigation against the complainant.Thus,forum shopping involves discriminating among institutional memberships,the logic being that,on a given measure(s),some complainants will prefer a precedent that bears only on a subset of their trade relations,others a precedent that bears on all their trade relations, while still others will prefer not to set a precedent at all.To generate specific hypoth- eses,I focus on two variables:(1)the complainant's preference concerning the outcome of the dispute,which I define as being more or less liberal,or free- trade oriented,than the status quo policies of the defendant,other trade partners, and the likely verdicts of the regional and multilateral institutions;and (2)the complainant's expectation concerning the future value of the precedent set,by which I mean the likelihood that the complainant will use the resulting case law in future litigation against other members more than other members will use it against the complainant. First,to say that the complainant is liberal (illiberal)is simply to say that it prefers a ruling on the disputed measure that is more free-trade oriented (protec- tionist).For example,the complainant might favor a ruling that advances a fuller interpretation of a member's rights,a stricter understanding of its obligations,or one that otherwise clarifies principles of nondiscrimination in a given issue area. Since the complainant has no incentive to file if it prefers a less liberal outcome than the defendant's status quo policy,the real question is how liberal the com- plainant is in relation to other trade partners,the anticipated rulings of NAFTA and the WTO,and the expected value of future regional or multilateral litigation on similar matters.I thus model two sets of disputes that are the most generaliz- able:the first involving an illiberal complainant,the second a liberal one.The model predicts that both"types"of complainants will sometimes choose the regional forum,and at other times choose the multilateral forum.It also predicts that an illiberal complainant,who is nonetheless more liberal than the defendant,may prefer not to file at all. 40.See Acheson and Maule 1999:Gantz 1999:and Goldstein 1996.553

in trade law and policy, including in studies of forum shopping+ 40 The novel con￾tribution of this article is that I specify the precise circumstances under which precedent will lead a complainant to file its dispute in one or the other forum, or to avoid filing at all+ It is logical that the complainant will file in whichever court is most likely to come closest to its “ideal” ruling+ Indeed, if the complainant cared only about winning a legal victory over the defendant, this would be the end of the story+ The problem, however, is that precedent ensures that a legal victory does not just apply to the complainant’s trade with the defendant, but also to its trade with other mem￾bers of the institution in which the resulting case law was set+ Here, precedent can be a double-edged sword: it can make it easier for the complainant to bring litiga￾tion against these other members of the institution, but it can also help these other members bring litigation against the complainant+ Thus, forum shopping involves discriminating among institutional memberships, the logic being that, on a given measure~s!, some complainants will prefer a precedent that bears only on a subset of their trade relations, others a precedent that bears on all their trade relations, while still others will prefer not to set a precedent at all+ To generate specific hypoth￾eses, I focus on two variables: ~1! the complainant’s preference concerning the outcome of the dispute, which I define as being more or less liberal, or free￾trade oriented, than the status quo policies of the defendant, other trade partners, and the likely verdicts of the regional and multilateral institutions; and ~2! the complainant’s expectation concerning the future value of the precedent set, by which I mean the likelihood that the complainant will use the resulting case law in future litigation against other members more than other members will use it against the complainant+ First, to say that the complainant is liberal ~illiberal! is simply to say that it prefers a ruling on the disputed measure that is more free-trade oriented ~protec￾tionist!+ For example, the complainant might favor a ruling that advances a fuller interpretation of a member’s rights, a stricter understanding of its obligations, or one that otherwise clarifies principles of nondiscrimination in a given issue area+ Since the complainant has no incentive to file if it prefers a less liberal outcome than the defendant’s status quo policy, the real question is how liberal the com￾plainant is in relation to other trade partners, the anticipated rulings of NAFTA and the WTO, and the expected value of future regional or multilateral litigation on similar matters+ I thus model two sets of disputes that are the most generaliz￾able: the first involving an illiberal complainant, the second a liberal one+ The model predicts that both “types” of complainants will sometimes choose the regional forum, and at other times choose the multilateral forum+ It also predicts that an illiberal complainant, who is nonetheless more liberal than the defendant, may prefer not to file at all+ 40+ See Acheson and Maule 1999; Gantz 1999; and Goldstein 1996, 553+ Institutions, Forum Shopping, and Trade Disputes 743