The WTO Case Law of 2003 The American Law Institute Reporters 039 Studies The American Law Institute Reporters Studies on WTO Law

It ruled in favor of India on the second issue, however, concluding thatimports from producers not individually investigated could not bepresumed to be dumped for purposes of injury anal

THE AMERICAN LAW INSTITUTEREPORTERS’ STUDIES

This book is the third in a series of annual volumes that will be utilized

in the development of an American Law Institute (ALI) project onWorld Trade Organization (WTO) Law The volumes undertake a yearlyanalysis of the case law from the adjudicating bodies of the WTO.The Reporters’ Studies for 2003 cover a wide range of WTO law fromtrade in goods to trade in services Each case is jointly evaluated by

an economist and a lawyer, both well-known experts in the fields oftrade law and international economics The Reporters critically reviewthe jurisprudence of WTO adjudicating bodies and evaluate whetherthe ruling ‘‘makes sense’’ from an economic as well as legal point ofview, and if not, whether the problem lies in the interpretation of thelaw or the law itself The Studies do not always cover all issues discussed

in a case, but they seek to discuss both the procedural and thesubstantive issues that form, in the Reporters’ view, the ‘‘core’’ of thedispute

H E N R I K H O R N is Professor of International Economics at the Institutefor International Economic Studies, Stockholm University He is

a member of the Editorial Board of the World Trade Review and is

a member of the Centre of Economic Policy Research He has previouslyworked for the Economic Research and Analysis Division of the WorldTrade Organization, and has been a judge in the Swedish Market Court(supreme court for competition law)

P E T R O S C M A V R O I D I S is Edwin B Parker Professor of Law at ColumbiaLaw School, Professor at the University of Neuchaˆtel, and a member ofthe Centre for Economic Policy Research He was previously Chair ofCompetition Law, EUI, Florence, and a member of the Legal AffairsDivision of the World Trade Organization

THE WTO CASE LAW OF 2003

The American Law Institute

Reporters’ Studies

Edited byHENRIK HORNANDPETROS C MAVROIDIS

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo Cambridge University Press

The Edinburgh Building, Cambridge cb2 2ru, UK

First published in print format

isbn-13 978-0-521-83423-0

isbn-13 978-0-511-26636-2

© The American Law Institute 2005

2006

Information on this title: www.cambridg e.org /9780521834230

This publication is in copyright Subject to statutory exception and to the provision of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press.

isbn-10 0-511-26636-7

isbn-10 0-521-83423-6

Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org

hardback

eBook (EBL) eBook (EBL) hardback

Foreword page vii

A Note on the American Law Institute viii

American Law Institute Reporters ix

Henrik Horn and Petros C Mavroidis

2 European Communities
Anti-Dumping Duties on

Imports of Cotton-Type Bed Linen from India

(AB-2000-13, WT/DS141/AB/R:DSR 2003: III, 965):

Recourse to Article 21.5 of the DSU by India 11

Gene M Grossman and Alan O Sykes

3 United States
Sunset Review of Anti-Dumping Duties onCorrosion-Resistant Carbon Steel Flat Products From Japan(AB-2003-5, WT/DS244/AB/R): A Legal and EconomicAnalysis of the Appellate Body Ruling 31

Robert Howse and Robert W Staiger

4 United States
Continued Dumping and Subsidy

Offset Act of 2000 (WT/DS217/AB/R: DSR 2003:I,375) 52

Henrik Horn and Petros C Mavroidis

5 European Community
Antidumping Duties on MalleableCast Iron Tube or Pipe Fittings from Brazil

(WT/DS219/AB/R: DSR 2003:VI, 2613) 87

Henrik Horn and Petros C Mavroidis

6 United States
Final Determination with Respect to CertainSoftwood Lumber from Canada (AB-2003-6,

Henrik Horn and Petros C Mavroidis

v

7 United States
Definitive Safeguard Measures on Imports ofCertain Steel Products (WT/DS259; WT/DS252; WT/DS248;WT/DS249; WT/DS251; WT/DS258; WT/DS254; WT/DS253:DSR 2003:VII, 3117) 146

Gene M Grossman and Alan O Sykes

8 Mexico
Measures Affecting Telecommunications Services(WT/DS204/R): A Comment on ‘‘El mess in TELMEX’’ 188

Damien J Neven and Petros C Mavroidis

9 European Communities
Conditions for the Granting ofTariff Preferences to Developing Countries

Gene M Grossman and Alan O Sykes

10 United States
Anti-Dumping Act of 1916 (Original

Complaint by the European Communities)
Recourse toArbitration by the United States under 22.6 of the DSU,

WT/DS136/ARB, 24 February 2004: A Legal and

Economic Analysis 254

Robert Howse and Robert W Staiger

11 Japan
Measures Affecting the Importation of Apples

(AB-2003-4): One Bad Apple? (DS245/AB/R):

Damien J Neven and Joseph H H Weiler

This is the third volume in the American Law Institute’s effort

to analyze decisions rendered in disputes before the World TradeOrganization

Trade Law is in its infancy as a body of legal doctrine In two priorvolumes, the ALI sponsored analyses of decisions issued in 2001 and

2002 This book presents an examination of decisions rendered in 2003

As before, the work has been accomplished by teams consisting of

a lawyer and an economist, each a distinguished expert on the worldtrading system Early drafts were criticized by the various participants,and then redrafts were presented to an international group of experts

at a meeting in April 2005 at the WTO headquarters in Geneva.Having studied three years of WTO decisions in this ‘‘bottom-up’’manner, we will now begin to draft the general principles of trade law

We also hope to continue with the analysis of individual decisions

We are immensely grateful to the two leaders of this project, HenrikHorn of Stockholm University and Petros Mavroidis of the University

of Neuchaˆtel and Columbia University We also appreciate the work

of the economists and lawyers who wrote the studies in this volume.And we appreciate the generous financial support for our project fromJan Wallander’s and Tom Hedelius’ Research Foundation, SvenskaHandelsbanken, Stockholm, and the Milton and Miriam HandlerFoundation

Lance LiebmanDirectorThe American Law Institute

vii

The American Law Institute was founded in 1923 and is based inPhiladelphia The Institute, through a careful and deliberative process,drafts and then publishes various restatements of the law, model codes,and other proposals for legal reform ‘‘to promote the clarification andsimplification of the law and its better adaptation to social needs,

to secure the better administration of justice, and to encourage andcarry on scholarly and scientific legal work.’’ Its membership consists

of judges, practicing lawyers, and legal scholars from all areas of theUnited States as well as some foreign countries, selected on the basis ofprofessional achievement and demonstrated interest in the improve-ment of the law The Institute’s incorporators included Chief Justiceand former President William Howard Taft, future Chief Justice CharlesEvans Hughes, and former Secretary of State Elihu Root JudgesBenjamin N Cardozo and Learned Hand were among its early leaders.The Institutes’s restatements, model codes, and legal studies are used

as references by the entire legal profession

The American Law Institute’s website ishttp://www.ali.org

viii

KYLE BAGWELL is Professor of Economics, Columbia University.

GENE M GROSSMAN is Jacob Viner Professor of International Economics,Princeton University

HENRIK HORN is Professor of International Economics, Institute forInternational Economic Studies, Stockholm University

ROBERT L HOWSEis Professor of Law, University of Michigan Law School

PETROS C MAVROIDIS is Edwin B Parker Professor of Foreign and parative Law, Columbia Law School, and Professor of Law, University

Com-of Neuchaˆtel

DAMIEN J NEVEN is Professor of Economics, Graduate Institute forInternational Studies, University of Geneva

ROBERT W STAIGERis Professor of Economics, University of Wisconsin

ALAN O SYKES is Frank and Bernice Greenberg Professor of Law,University of Chicago Law School

JOSEPH H H WEILER is Joseph Straus Professor of Law and Jean MonnetChair, New York University School of Law

ix

henrik hornpetros c mavroidis

1 The projectThis is the third volume of the Reporters’ Studies undertaken in thecontext of the American Law Institute (ALI) project Principles of WorldTrade Law: The World Trade Organization (WTO) The aim of theproject is to provide a systematic analysis of WTO law based in bothEconomics and Law This year’s focus has mainly been on disputesthat came to an administrative end during the year 2003, either becausethey were not appealed or because the appeal process concluded.Each dispute has been evaluated jointly by an economist and a lawyer.The general task of this two-person team is to evaluate whether theruling ‘‘makes sense’’ from an economic as well as a legal point of viewand, if it does not, whether the problem lies in the legal text or inthe interpretation thereof The authors do not always cover all issuesdiscussed in a case, but they seek to discuss both the procedural andthe substantive issues that form the ‘‘core’’ of the dispute, as they see it.The Reporters’ Studies have, this year, been drafted by the followingpersons, who have been appointed Reporters for the project by the ALI:Gene M Grossman, Jacob Viner Professor of International Economics,Princeton University, USA

Henrik Horn, Professor of International Economics, Institute forInternational Economic Studies, Stockholm University, Sweden.Robert L Howse, Alene and Allan F Smith Professor of Law, University

of Michigan Law School, USA

1

Petros C Mavroidis, Professor of Law, University of Neuchaˆtel,Switzerland, and Edwin B Parker Professor of Law at ColumbiaLaw School, USA.

Damien J Neven, Professor of Economics, Graduate Institute forInternational Studies, University of Geneva, Switzerland

Robert W Staiger, Professor of Economics, University of Wisconsin, USA.Alan O Sykes, Frank and Bernice Greenberg Professor of Law,University of Chicago Law School, USA

Joseph H.H Weiler, Joseph Straus Professor of Law and Jean MonnetChair, New York University School of Law, USA

The Reporters’ Studies were initially scrutinized in an October 2004meeting of all of the Reporters in Princeton After revisions, the Studieswere presented and discussed in a meeting held in Geneva on April 12,

2005, with the following group of lawyers and economists:

Richard Baldwin, Graduate Institute of International Studies, Geneva,Switzerland

Armin von Bogdandy, Max Planck Institut, Heidelberg, Germany.Claus-Dieter Ehlermann, Wilmer Cutler, Brussels, Belgium

Wilfred J Ethier, Department of Economics, University of Pennsylvania,Philadelphia, USA

Gary Horlick, Wilmer Cutler, Washington DC, USA

Gabrielle Marceau, World Trade Organization, Geneva, Switzerland.Mitsuo Matshushita, Seikei University School of Law, Tokyo, Japan.Niall Meagher, Advisory Centre for WTO law, Geneva, Switzerland.Ha˚kan Nordstro¨m, National Board of Trade, Stockholm, Sweden.Donald Regan, University of Michigan School of Law, Ann Arbor, USA.Delissa Ridgway, US Court of International Trade, New York, USA.Frieder Roessler, Advisory Centre for WTO law, Geneva, Switzerland.Andre´ Sapir, Universite´ Libre de Bruxelles, Brussels, Belgium

Arun Venkataraman, World Trade Organization, Geneva, Switzerland.Jasper Wauters, World Trade Organization, Geneva, Switzerland.Eric White, EC Commission, Brussels, Belgium

Claire Wright, Thomas Jefferson School of Law, San Diego, USA.Werner Zdouc, World Trade Organization, Geneva, Switzerland.The final versions, as published in this volume, have been subjected

to yet another round of revisions derived from the advisory meeting.Despite these collective efforts, each pair of authors remains solelyresponsible for the Studies it has authored

This project would not have existed had it not been for the effortsand commitment of Professor Lance Liebman, Director of the ALI.

We have also benefited greatly from the support of the President ofthe ALI, Michael Traynor, the ALI Deputy Director, Elena Cappella,and the former ALI Deputy Director, Michael Greenwald The ALIhas also provided excellent assistance with the administration of ourmeetings We are extremely grateful for the generous financial supportthe project has received from Jan Wallander’s and Tom Hedelius’Research Foundation, Svenska Handelsbanken, Stockholm, and theMilton and Miriam Handel Foundation Without their support, thisproject would not have materialized

2 The Reporters’ Studies on the WTO Case Law of 2003

We here provide a brief summary of this year’s Studies in the order

of their appearance in this volume

A main theme in this year’s Studies has been contingent protection.Grossman and Sykes discuss the compliance dispute between Indiaand the European Community (EC) in which India alleged lack ofcompliance by the EC with an earlier WTO ruling on antidumpingduties for cotton-type bed linen India raised issues relating to the

EC treatment of ‘‘other factors,’’ that is, issues regarding the manner

in which an investigating authority should treat factors, other thanthose mentioned in the body of the Agreement, that cause injury Indiahad briefly noted such factors in the original proceedings but hadnot argued them fully Although, in their view, Grossman and Sykescan imagine cases where the failure by a panel to investigate factualaspects of a case might in and of itself warrant a reversal of its eventualfinding, they decline to pronounce on the existence of such grounds

in this dispute, lacking an effective demonstration by India that thePanel’s omission led to abuse or a biased outcome The authors goone step further, however, and suggest that this issue raises interestingquestions regarding the proper scope of res judicata, issue preclusion,and waiver in WTO jurisprudence To highlight these questions, theydevelop an analytical model to compare expected litigation costs in

a judicial system with a rule of waiver to those in a system in whichlitigants can bring multiple claims They show that a rule of waiver neednot minimize litigation costs, because some disputes can be resolved

at lower costs if a claimant is not forced to bring all of its arguments

at once India also claimed that the EC violated WTO rules when

conducting its revised injury analysis, by treating all imports fromfirms not individually investigated as if they had been dumping,despite the fact that some of the firms individually investigated werefound not to be dumping On this score, the authors conclude that theAppellate Body’s (AB) decision has some economic appeal, but rests

on shaky legal foundations

In their paper on Corrosion Steel, Howse and Staiger deal with thelegal benchmark to be applied by an investigating authority whenevaluating whether to retain antidumping duties beyond the originalfive-year period (the sunset review) The current law is quite open-ended and requires WTO Members to demonstrate the likelihood ofcontinued or renewed dumping before agreeing on the extension ofduties in place In the authors’ view, a meaningful assessment of thelikelihood of continued or renewed dumping requires an understand-ing of the conditions that led to dumping in the first place, and

a determination of whether these conditions have changed in a waythat removes the original reason for dumping In their opinion,neither of these two elements appears to have played any real role in the

US investigating authority’s methodology for determining likelihood, or

in the AB’s assessment of the legitimacy of this methodology.The authors note that Art 9.1 of the Antidumping Agreement (AD)does not require an assessment of the conditions that have led thecompanies named in an antidumping order to dump; the authorsmaintain, however, that a different criterion should be applied inArt 9.1 AD compared to that applied in Art 11.3 AD They see nonecessary inconsistency in this: while a requirement could have beenincluded in the AD that the particular reasons for dumping must

be articulated as a condition for imposing antidumping duties underArt 9.1 AD, there is no logical necessity that such a requirement must

be included By contrast, Howse and Staiger argue that an standing and articulation of the conditions that led to dumping islogically necessary to assess the likelihood of continued or reneweddumping, and therefore, is an implied condition for extending definitiveantidumping duties beyond the five-year termination date that Art 11.3

under-AD otherwise dictates In light of their preferred benchmark for theadjudicating of such cases, the authors conclude that the AB erred bynot requiring the United States to provide such an assessment as part

of its sunset review

The ‘‘Byrd ’’ litigation between the United States and a number ofWTO Members over the US Continued Dumping and Subsidy Offset Act

of 2000 (CDSOA) is discussed by Horn and Mavroidis ThisUnited States law reserves disbursement of collected antidumping andcountervailing duties exclusively to those economic operators that havebacked a petition to open a dumping (or countervailing) investigation.The AB, in part modifying the Panel’s findings, concluded that thelaw was WTO-inconsistent in that it constituted an impermissibleaction against dumping (and subsidization) In the authors’ view, the

AB erred both in terms of its reasoning and in terms of its findings:the AB’s treatment of the reasons advanced by the Panel to supportits finding that the Byrd legislation violated Art 18.1 AD and Art 32.1 ofthe Agreement on Subsidies and Countervailing Measures (SCM) wasinconsistent; the AB, in striking down the legislation, used an economictheory that was inadequately motivated and of doubtful validity, ifmeant to describe the legislation’s typical impact on various industries,and it failed to explain how general it believed the theory to be; the ABstruck down the legislation on much weaker grounds than those it hadpreviously established as a requirement in its case law; and finally, the

AB should have undertaken a more comprehensive discussion of theclaims under Art 5.4 AD and 11.4 SCM, since an illegality could morenaturally have been established under these provisions

Horn and Mavroidis also discuss the Pipe Fittings litigation betweenBrazil and the EC In this dispute, Brazil raised a series of issues con-cerning the lawfulness of the EC imposition of antidumping dutiesagainst imports of pipe fittings originating in Brazil Chief amongthe issues raised were questions relating to the treatment of a devaluedcurrency, the conditions under which low-volume imports should

be considered imports in the ordinary course of trade, and the scope

of information that should go into a final notice advising interestedparties as to the extent of an antidumping imposition In the authors’view, the AB’s conclusions with respect to the first two issues are hardlysupported by a contextual understanding of the various obligationslaid down in the AD Disregarding the implication of a devaluation (andits effect on pricing decisions) would lead the investigating authorities

to practice antidumping against its very purpose: antidumping sures are not available to provide injured parties with a lump-sumpayment, but, instead, to offset dumping occurring in future transac-tions (i.e after the conclusion of the investigation) Additionally, theAB’s understanding of the discretion conferred upon the investigatingauthorities when constructing the normal value would lead, as it did inthe present case, to logically incoherent outcomes Finally, the authors

mea-argue that the transparency obligations should be read in light of theirpurpose, which is to inform uninformed parties the basis of decisionsmade during the investigation and thus to ensure that due processhas been adhered to throughout the whole administrative procedure,otherwise these obligations will be reduced to mere procedural require-ments that do not perform their assigned function.

The Softwood Lumber IV litigation between Canada and the UnitedStates is also examined by Horn and Mavroidis This dispute is part

of the long-standing conflict between Canada and the United Statesconcerning the forestry sector A main issue in the present disputeconcerns the benchmark that a WTO Member can use in order tocalculate the amount of subsidization In this dispute, the Panelconcluded that there was a problem with the current text of theSCM Agreement in that it does not include sufficient flexibility toallow WTO Members to use alternative benchmarks when confronting

a factual situation not envisaged by the law itself; accordingly, thePanel indirectly argued for some form of legislative amendment The

AB did not agree with the Panel’s reading, however, concluding thatthe existing text reflects sufficient flexibility to allow WTO Members

to deal with situations such as the one in the present dispute Theauthors, however, find it hard to interpret Art 14(d) SCM so as toallow for alternative benchmarks, such as the benchmark proposed

by the United States In their view, the AB here effectively took onthe role of legislator, thus contravening the discipline laid down inArt 3.2 of the Understanding on Dispute Settlement

Grossman and Sykes discuss the steel disputes between the UnitedStates and a host of complainants that challenged the consistency ofsafeguards in the steel sector imposed by the United States againstimports from a variety of sources The authors take the position thatthe AB’s decision in this dispute is one more link in a line of unsatis-factory decisions in the safeguards area In their view, the underlyingproblem stems from the fact that the treaty text regarding the pre-conditions for the use of safeguard measures is seriously deficient The

AB, with its usual emphasis on textual interpretation, has done little

to resolve the puzzles that the text creates As a result, WTO membersare still left with little guidance concerning the proper use of safeguardsbeyond some confusing and sometimes incoherent standards, notwith-standing the sizeable amount of jurisprudence in this field The authorsreview these issues as they have arisen, not only in the instant decision,but in prior decisions as well, and then discuss the details of the steel

dispute They place more emphasis on the Panel report than the ABreport, as the latter breaks little new ground In the authors’ opinion,although it is probably difficult to quarrel with the outcome in thisparticular case (in light of the procedural deficiencies they observe),

it is high time to reverse the tide in the safeguards area by breakingwith the current line of jurisprudence

Some of the reports discussed in this year’s volume deal with issuesother than contingent protection Neven and Mavroidis discuss thefirst ‘‘genuine’’ services litigation in the WTO context, the disputebetween the United States and Mexico concerning the rates charged

by the Mexican monopolist, Telmex, for terminating calls originating

in the United States (Mexico
Telecoms) The Panel, whose findingswere never appealed, found that Telmex was charging rates unrelated

to its cost structure, thus violating its obligations under the GeneralAgreement of Trade in Services (GATS) The authors critically distancethemselves from both the Panel’s reasoning and its findings: in theauthors’ opinion, the Panel mischaracterized the facts before it, and,moreover, applied the wrong law In their view, Mexico had made

no commitment regarding the factual situation as presented by thecomplaining party (and was hence not bound by the relevant legalframework) Furthermore, even if it had made such commitments,the relevant framework did not address the situation before the Panel(cross-border supply of termination services), but rather addressedonly the mode-3 supply of telecoms services (commercial presence ofthe supplier in the market of the country terminating incoming calls).The authors conclude that the Panel’s findings constitute an imper-missible extension of the existing legal framework to transactions thatthe founding fathers did not intend to cover They also take a criticalstance with respect to the pure competition-law component of thedispute, where they argue that it was an overly bold move for the Panel

to pronounce on whether the rates charged were cost-oriented, in light

of the information presented before it The Panel’s approach is, fromthe authors’ perspective, at odds with the approaches of numerouscompetition authorities on the same score

Grossman and Sykes deal with the Generalized System of Preferences(GSP) dispute between India and the EC, where, for the first time,the legality of practices by donor countries under the terms of theEnabling Clause was being discussed In this case, India complainedagainst the EC practice of making distinctions among develop-ing countries, and, based on such distinctions, granting additional

preferences to some developing countries (those that had engaged incombating drug production and trafficking) The AB, overturning inpart the Panel findings in this respect, condemned the specific ECpractice (because the list of beneficiaries was closed), but accepted theprinciple that the making of distinctions among developing countrieswas perfectly compatible with the WTO legal order, to the extent thatobjective criteria had been established ex ante The authors disagreewith both the reasoning and the outcome of this litigation In theirview, the EC scheme could, at least theoretically, lead to trade diversion,with less efficient developing countries increasing their exports atthe expense of the exports of more efficient developing countries.Moreover, in the authors’ opinion, beyond the original distinctionbetween developing and least developed countries that is enshrined

in the WTO legal framework, and the graduation from developing

to developed country status, there is no justification for schemes such

as the one contested in this case Finally, from a policy perspective,there is no reason to believe that such schemes are to the benefit ofthe countries receiving preferences

Howse and Staiger discuss the arbitration on the 1916 AntidumpingAct, which addresses the meaning of equivalence between a violation

of international law and the countermeasures applied to respond tothe violation In this case, the EC proposed countermeasures that

‘‘mirrored’’ the US WTO-inconsistent behavior (the availability of

a private right of action for treble damages against a foreign firmengaging in certain kinds of dumping) Based largely on previousrulings, the Arbitrator held that, while the central purpose of counter-measures under DSU Art 22 was to induce compliance and excludepunitive purposes, it could not endorse the proposed countermeasures

of the EC, unless it had evidence that the trade or economic effect of the

EC mirror-reaction on the United States would not exceed the trade

or economic effect of the United States’ original action on the EC.The authors reject the Arbitrator’s approach for several reasons: first,

in their view, the Arbitrator focused on the word ‘‘level’’ withoutalso considering its immediate context
the notion of ‘‘equivalence.’’The word ‘‘equivalence’’ implies proportionality between two thingsthat are not entirely commensurate or reducible in value to a commonmetric Second, ‘‘normative countermeasures,’’ where equivalence orproportionality is achieved by suspending a symmetrical obligation,have often been recognized as consistent with the principles of stateresponsibility Third, the Arbitrator confused the issue of the limits

on countermeasures with the issue of whether the purpose ofcountermeasures was to re-establish the pre-breach equilibrium betweenthe parties, to achieve compliance, or to punish The authors alsoadvance in their paper some normative thoughts on the placement ofefficient breach of contract inside the current WTO legal regime Theyargue that the Arbitrator’s focus on trade effects has some meritfrom an economic point of view, and they also propose how the tradeforegone should be measured.

Neven and Weiler review the decision by the AB in the Japan
Apples dispute, which concerns measures affecting the importation

of apples in Japan The authors emphasize that a crucial aspect of theSanitary and Phytosanitary (SPS) agreement is that it imposes adiscipline on risk-reducing measures even in the absence of discrimina-tion in favor of domestic products Their discussion on the evaluation

of risk-reducing measures focuses on two issues, namely the scope

of the mandate given to the adjudicators and the standard of reviewthat they should apply Neven and Weiler emphasize the difficulty thatadjudicators face in distinguishing between the level of risk that acountry will find it optimal to support (which cannot be challenged),and the question of whether the risk-reducing measures are necessary toachieve the chosen level of risk They further observe that the commonmethodology used by panels, evaluating the existence of risk in theabsence of risk-reducing measures, has limited applicability The authorsalso discuss how the Panel’s approach can be abused, leading theadjudicators to slip from the evaluation of whether a measure isnecessary to achieve a given level of risk to an implicit challenge of thelevel of risk itself (which should remain the preserve of the Members).Regarding the standard of review, they argue that a lower standardshould be applied to measures that do not threaten fundamentalprinciples like nondiscrimination Finally, the authors also consider thePrecautionary Principle in the context of the SPS Agreement In theirview, the provisions of the SPS Agreement reflect the distinctionbetween risk and ambiguity: whereas the former would cover theprovisions that concern science-based SPS measures, the latter wouldextend to cover precautionary measures The authors consequently callinto question the Panel’s and the AB’s unwillingness to apply thePrecautionary Principle in the context of this dispute

This year, we will not attempt to classify the papers according towhether the authors see the rationale provided by the adjudicatingbodies, or the outcome, as correct Nevertheless, we observe that,

in basically all the disputes discussed, the authors raise serious cerns about the adjudicators’ reasoning in support of the final outcome.Whether the outcomes of the disputes make sense from an economicpoint of view is more difficult to judge this year, since the Studies

con-to a large extent address antidumping issues, and, as is well known,

it is difficult to reconcile antidumping measures with standard nomic reasoning

eco-European Communities
Anti-Dumping Duties

on Imports of Cotton-Type Bed Linen from India (AB-2000-13, WT/DS141/AB/R:DSR 2003: III, 965): Recourse to Article 21.5 of the DSU by India

gene m grossman*

Princeton Universityalan o sykes*

University of Chicago

1 IntroductionThis chapter addresses the dispute brought to the World TradeOrganization (WTO) by India concerning anti-dumping duties imposed

by the European Communities (EC) on cotton-type bed linen Anearlier complaint brought by India challenged the anti-dumping duties

on a number of points, including the EC practice of ‘‘zeroing’’ for thecomputation of dumping margins (which had the effect of assigning

a negative dumping margin a weight of zero when computing

a weighted average dumping margin).1India prevailed in that dispute,2and the EC responded with Council Regulation (EC) No 1644/2001,amending the original anti-dumping measure on bed linen from India.India was of the view that the amended measure did not comply with

EC obligations under the WTO Anti-dumping Agreement, and brought

1 The decision in the earlier proceeding is the subject of an earlier chapter in this series See Janow and Staiger ( 2003 ).

2

See European Communities
Anti-Dumping Duties on Imports of Cotton-Type Bed Linen from India, WT/DS141/AB/R, adopted March 12, 2001.

* This chapter was prepared for the American Law Institute project on ‘‘The Principles

of WTO Law.’’ We thank Kathy Spier for thoughtful assistance.

11

the proceeding under Art 21.5 of the DSU that is the subject ofthis chapter.

Several issues were raised before the Art 21.5 Panel, but only threeissues reached the Appellate Body First, India argued that although the

EC had corrected the ‘‘zeroing’’ problem, it had failed to ensure thatinjury attributable to ‘‘other factors’’ had not wrongly been attributed todumped imports, in violation of Art 3 of the Anti-dumping Agreement.Second, India argued that in conducting its revised injury analysis,the EC violated Art 3 when it presumed that all imports from exportersnot individually investigated were ‘‘dumped,’’ even though 53% of theimports from exporters that were individually investigated were foundnot to have been dumped once the ‘‘zeroing’’ method of calculationwas abandoned Finally, India argued that the EC had not properlyconsidered certain factors bearing on injury that it was required toconsider under Art 3

The Appellate Body ruled in favor of the EC on the first issue, holdingthat it had been resolved definitively in the original proceeding

It ruled in favor of India on the second issue, however, concluding thatimports from producers not individually investigated could not bepresumed to be dumped for purposes of injury analysis when some ofthe individually investigated exporters were not dumping On the thirdissue, the Appellate Body upheld the finding against India by the Panel,deferring to its resolution of what the Appellate Body considered anessentially factual issue

From a legal perspective, the Appellate Body’s decision on the firstissue raises some interesting questions about the proper scope ofres judicata, issue preclusion, and waiver in WTO jurisprudence, butprovides few answers The case breaks new ground with respect to thesecond issue noted above as well, and we quibble somewhat with theAppellate Body’s legal and logical reasoning there Finally, the AppellateBody’s deference to the Panel on the third issue seems appropriate, asbest we can determine

From an economic perspective, we find the procedural issue to be aninteresting one Little analytical work has been done by economists onthe proper scope of res judicata and the related notions of issue pre-clusion and waiver We develop some simple points about these issuesbelow, which provide some basis for questioning the refusal of thecompliance Panel to entertain India’s arguments on ‘‘non-attribution.’’Regarding the second issue, the Anti-dumping laws make so littleeconomic sense in general that it is difficult to offer any guidance as to

their ‘‘proper’’ administration The ruling in favor of India on thepresumption of dumping issue seems reasonable from a statisticalstandpoint, however, although it is less clear that it is right as a legalmatter Finally, the Appellate Body’s deference to the factual conclusion

of the Panel on the third issue raises no economic issues of note

We lay out the legal issues and their resolution by the Panel and theAppellate Body in Section2 Section3offers a critical analysis of the casefrom a law and economics perspective

2 Factual and legal issues and their disposition

2.1 Non-attribution of injury caused by ‘‘other factors’’Article 3.5 of the Anti-dumping Agreement provides that investigatingauthorities must ‘‘examine any known factors other than the dumpedimports which at the same time are injuring the domestic industry,and the injuries caused by these other factors must not be attributed

to the dumped imports.’’ In the original proceeding, India challengedthe EC’s duties, inter alia, on the grounds that the EC had failed toensure that injury attributable to ‘‘other factors’’ was not attributed

to dumped imports from India, although it did not pursue the issuevery actively The original Panel dismissed the one substantive pointraised by India under this rubric, and otherwise said that Indiahad failed to make out a prima facie case on the issue That finding wasnot appealed

The EC did not conduct a new analysis of ‘‘other factors’’ as part of itsrevised injury analysis when it promulgated Regulation No 1644/2001,and had simply relied on its previous discussion of the matter Indiathen argued again that the EC had failed to ensure that injury caused by

‘‘other factors’’ was not attributed to dumped imports In particular,

it pointed to various ‘‘other factors’’ that had not been a subject ofdiscussion before the original Panel, including rising input costs forEuropean firms and the failure of output prices in the EC to keep upwith inflation

The EC requested a preliminary ruling from the Panel to the effectthat such matters could not be raised in an Art 21.5 proceeding, andthe Panel agreed: ‘‘To rule on this aspect of India’s claim underArticle 3.5 in this proceeding would be to allow India a second chance

to prevail on a claim which it raised, but did not pursue, in the originalproceeding We cannot conclude that such a result is required by

Article 21.5 of the DSU, or any other provision The possibility formanipulative or abusive litigation tactics that would be opened byallowing Members an opportunity to obtain a ruling in an Article 21.5proceeding that they could have sought and obtained in the originaldispute would, in our view, be inestimably harmful to the effectiveoperation of the dispute settlement system.’’3 Although the Panel didnot use these terms, its reasoning invokes notions of res judicata,issue preclusion, and waiver.

The Appellate Body affirmed the Panel’s ruling on this issue In doing

so, it emphasized that new claims can at times be raised before anArt 21.5 Panel It noted that the purpose of such Panels is to review theWTO consistency of measures taken to comply with prior rulings, andthat many such measures will differ significantly from the measuresoriginally challenged, and may be inconsistent with WTO obligations

in ways that the original measures were not Although the AppellateBody made clear that new inconsistencies of this sort were the propersubject of discussion before an Art 21.5 Panel, it stated: ‘‘[h]ere, Indiadid not raise a new claim before the Art 21.5 panel; rather, Indiareasserted in the Art 21.5 proceedings the same claim that it had raisedbefore the original panel in respect of a component of the implementa-tion measure which was the same as the original measure The sameclaim was dismissed by the original Panel, and India did not appeal thatfinding.’’4 The Appellate Body went on to hold that when the originalPanel report was adopted by the Dispute Settlement Body (DSB),

it became a final resolution of the dispute on the ‘‘other factors’’ issue.5Like the Panel, the Appellate Body relied for its ruling not so much

on any treaty text that addressed the issue, but on policy considerationsand on its earlier decision reviewing a similar issue that had arisenbefore the Shrimp
Turtle compliance Panel The Appellate Body empha-sized that India had raised the ‘‘same’’ claim earlier and lost, and put lessemphasis than the Panel had on the notion that the particular issuesraised by India could have been raised before but were not

2.2 Injury due to exporters not individually investigated

In the second investigation, as in the first, the EC did not investigateevery Indian exporter of cotton-type bed linen Article 6.10 of the Anti-dumping Agreement allows importing nations to investigate only

3

Panel Rep ô6.43. 4 AB Rep ô80. 5 AB Rep ô99.

a sample of all exporters in cases where an individual investigation

of all of them would be ‘‘impracticable.’’ Accordingly, the EC conductedindividual investigations of five of the larger Indian exporters, andapplied a weighted average anti-dumping duty to exports from otherexporters as is allowed by Art 9.4 of the Anti-dumping Agreement.The most important change between the original investigationand the second was to eliminate the practice of ‘‘zeroing’’ in thecomputation of weighted average dumping margins When zeroing waseliminated, two of the five exporters subject to individual investigation,accounting for 53% of the imports from the five individually investi-gated importers, were found not to be dumping at all The issue beforethe compliance Panel was how this new finding should affect injuryanalysis by the EC

Article 3.5 requires that the importing nation establish a causal linkbetween the dumped imports and injury In purporting to establish thislink when promulgating Regulation No 1644/2001, the EC assumedthat all imports from Indian exporters not individually investigated hadbeen dumped, even though 53% of the imports from the exportersindividually investigated had not been dumped India argued that the

EC thereby violated Art 3.1, which requires that the determination ofinjury be based on ‘‘positive evidence,’’ including an ‘‘objective exami-nation’’ of the ‘‘volume of dumped imports.’’ India argued that the ECshould presume that dumping was occurring by exporters not indi-vidually investigated in the same proportion as imports from exporterswho were individually investigated (47%) This would suggest a smallervolume of dumped imports than the EC had presumed were present,and might reverse the conclusion that dumped imports were causingmaterial injury

The EC argued that the presumption of dumping by exporters notinvestigated individually is permissible under the Anti-dumpingAgreement Its principal argument was based on Art 9.4 of the Anti-dumping Agreement, which permits an anti-dumping duty to beimposed on exporters not individually investigated as long as it doesnot exceed ‘‘the weighted average margin of dumping established withrespect to the selected exporters.’’ The EC contended that because it wasallowed to impose an anti-dumping duty on those exports, it had to also

be allowed to consider them ‘‘dumped’’ for purposes of injury analysis.Next, it argued that the group of exporters that it had chosen toinvestigate individually were not selected to be a statistically validsample, but rather represented the ‘‘largest percentage of the volume

of the exports which can reasonably be investigated,’’ one of theoptions under Art 6.10 Thus, the percentage of exports found to bedumped by the individually investigated exporters could not beassumed to reflect the amount of dumping by exporters not individ-ually investigated.

The Panel agreed with the EC ‘‘We can find no textual obligation inthe AD Agreement to separate out the unexamined producers’ importsinto dumped and not-dumped for purposes of the injury analysis ’’6

It also found India’s position to be logically flawed, given the fact that allnon-investigated imports could be subjected to a positive anti-dumpingduty under Art 9.4: ‘‘Under India’s approach, only a portion of importsfrom producers subject to that anti-dumping duty could be considered

as ‘dumped’ for injury purposes This effectively treats the importsfrom the same producers as dumped for purposes of duty assessment,and not dumped for purposes of injury analysis In our view, this is

an unacceptable outcome, suggesting that the analysis which leads to

it is untenable.’’

The Appellate Body reversed It emphasized the requirement for an

‘‘objective examination’’ of the volume of dumped imports, and notedthat imports not sold at dumped prices were specifically enumerated inArt 3.5 as one of the ‘‘other factors’’ which may cause injury and shouldnot be attributed to dumped imports It was also unpersuaded thatimports from exporters not individually investigated could be presumed

to be dumped simply because Art 9.4 permits them to be subjected to

an anti-dumping duty
‘‘[w]e do not see why the volume of importsthat have been found to be dumped by non-examined producers,for purposes of determining injury under paragraphs 1 and 2 ofArticle 3, must be congruent with the volume of imports from thosenon-examined producers that is subject to the imposition of anti-dumping duties under Article 9.4.’’7The Appellate Body stopped short

of endorsing India’s proposed method for calculating the volume ofdumped imports from exporters not individually investigated, however,allowing for the possibility that ‘‘positive evidence’’ of that volumemight be based on something other than the percentage of exportsdumped by the individually investigated exporters.8

Along the way, the Appellate Body was mindful of the standard ofreview under the Anti-dumping Agreement Article 17.6(ii) of the

6 Panel Rep ô6.139 7 AB Rep ô126 8 Id ô146.

Agreement provides that ‘‘[w]here the panel finds that a relevantprovision of the Agreement admits of more than one permissibleinterpretation, the panel shall find the authorities’ measure to be inconformity with the Agreement if it rests on one of those permissibleinterpretations.’’ The EC argued that its interpretation of the injuryprovisions was at least a ‘‘permissible’’ interpretation that was entitled

to deference, but the Appellate Body disagreed: ‘‘[W]hatever logy investigating authorities choose for calculating the volume of

methodo-‘dumped imports,’ that calculation and, ultimately, the determination ofinjury under Article 3, clearly must be made on the basis of ‘positiveevidence’ and an ‘objective examination.’ These requirements arenot ambiguous, and they do not ‘admit of more than one permis-sible interpretation’ within the meaning of the second sentence ofArticle 17.6(ii).’’9

2.3 Consideration of all ‘‘relevant factors’’ bearing on injuryArticle 3.4 of the Anti-dumping Agreement requires importing nations

to base their injury analysis on an examination of ‘‘all relevant economicfactors having a bearing on the state of the (domestic) industry.’’

It then provides a non-exhaustive list of such factors India assertedthat the EC failed to gather data on and to evaluate two ‘‘relevantfactors’’
stocks and capacity utilization
when it promulgatedRegulation No 1644/2001 The EC asserted that such data had beenpresented before the investigative authorities, and had been properlyconsidered The Panel ruled for the EC on this point, and India arguedthat the Panel abused its discretion in doing so by, in effect, acceptingthe EC’s unsupported assertions on the matter rather than conducting

a more thorough investigation

The Appellate Body upheld the Panel, which had ‘‘concluded that

it was clear that the European Communities had ‘in its record’ mation on stocks and capacity utilization
the two factors India hadfocused on
and that ‘unlike the original determination, the EC’sconsideration of these factors [was] clearly set out on the face of theredetermination.’’ ’10 While India wished to characterize the Panel’sconclusion as an abuse of its discretion, the Appellate Body saw it as

infor-a finfor-actuinfor-al conclusion by the Pinfor-anel thinfor-at winfor-as within its proper discretionand should not be disturbed on appeal

9 AB Rep ô118 10 AB Rep ô154.

3 Critical analysis3.1 Non-attribution and the procedural issue

One can quibble with the willingness of the Appellate Body to permitthe EC to rely on its original ‘‘other factors’’ analysis In light of itsresolution of the second issue in the case, discussed below, the EC isrequired to restate its assessment of the quantity of dumped imports,revising the estimate downward The quantity of fairly traded imports,

an ‘‘other factor’’ that might cause injury, must be revised upward.One might thus argue that the EC should redo both its analysis ofharm attributable to ‘‘dumped imports’’ and its analysis of harm due

to ‘‘other factors.’’ The Appellate Body does not reach this conclusion,however, perhaps because India’s arguments focused on EC input andoutput prices as the ‘‘other factors’’ to be considered

The much more interesting aspect of the ruling on this issue,however, is its procedural implications The WTO treaty text does notspecifically address res judicata and related issues, leaving to Panels and

to the Appellate Body the task of evolving sensible principles in the area

In this case, the Appellate Body insisted that ‘‘India did not raise a newclaim before the Article 21.5 panel; rather, India reasserted in theArticle 21.5 proceedings the same claim.’’ At some level, it is difficult toquarrel with the proposition that parties to WTO disputes should not

be permitted to relitigate the same claim over and over again What theAppellate Body masks with this language, however, is that the concept

of ‘‘sameness’’ can be interpreted broadly or narrowly

Recall the facts: India had raised the ‘‘non-attribution’’ issue in itsoriginal complaint, but did not advance factual arguments in relation

to that issue sufficient to make out a prima facie case Then, before thecompliance Panel, it sought to make those arguments seriously for thefirst time, pointing to ‘‘other factors’’ such as high EC input prices andlow EC output prices Here, to say that India had lost the same claimearlier is to imply that all arguments relating to a particular legal issueare part of the ‘‘same’’ claim, and are waived if they are omitted fromthe first round of litigation in which that issue appears The Panelopinion hinted at an even broader principle when it stated that it wouldnot afford India ‘‘an opportunity to obtain a ruling in an Article 21.5proceeding that they could have sought and obtained in the originaldispute.’’ This language suggests that all legal issues that could have beenraised in an earlier proceeding, but were not, are waived For termi-nological simplicity, we refer to these principles as rules of waiver,

although the reader should be aware that civil procedure treatises oftenattach the labels res judicata, issue preclusion, or claim preclusion tothese types of rules.

Rules of waiver have the obvious consequence of encouraginglitigants to raise issues sooner rather than later, and can hasten the finalresolution of a dispute Many legal systems have them.11 It is possiblethat such rules are economically desirable when all the costs and benefits

of the legal system are taken into account, but that is not obvious

We have found no treatment of the issue in the existing law and nomics literature on procedure, perhaps because a complete accounting

eco-of all the relevant considerations in any particular context is exceedinglydifficult to provide An exhaustive treatment is beyond the scope of thiscomment as well, but we will sketch some of the pertinent consid-erations that bear on the design of optimal waiver principles Beforeaddressing waiver, however, we set forth our understanding of thejustification for res judicata in its narrower sense

3.1.1 Res judicataCompliance with the law generally has a social value, and the promptresolution of legal proceedings can hasten valuable compliance Thisobservation seemingly applies as much to the WTO as to other legalcontexts But legal decision makers are imperfect, and may make errors

in their findings of law or fact When litigants are required to complywith erroneous decisions, error costs arise, often of the same nature

as the gains from compliance with correct decisions A desire to avoiderrors motivates principles of ‘‘due process’’ in many legal systems.Process itself is costly, however, and so it is unrealistic for most legalsystems to avoid error altogether The task of designing an optimalprocedure thus balances competing considerations: the value of resolv-ing legal issues sooner and of reducing process costs on the one hand,against the costs of errors on the other

Because of concerns about error, it is not uncommon for litigants

to be permitted to raise issues more than once The usual setting forrevisiting issues is the ‘‘appeal,’’ a common feature in many legalsystems, including now the WTO But there will generally bediminishing returns to reopening issues that have been previously

11

On the American rules in this area, see generally, Friedenthal, Kane and Miller ( 1999 ); James, Hazard and Leubsdorf ( 2001 ).

decided
at some point, the likelihood of error becomes sufficientlysmall that the benefits of ending the dispute and the associated processcosts predominate over any concerns about error Thus, rights of appealare always limited (and some matters may not be appealable at all).Res judicata in its narrowest sense simply precludes a litigant raising

an identical claim in a new proceeding when the claim was previouslyadjudicated It can be understood as a presumption that the legal system

in question already provides an appropriate error-correction mechanismthrough its appellate process Once a litigant has raised an argument,lost, and exhausted all available appeals, no further delays and litigationcosts are likely to be justified

To be sure, scenarios may arise in which concern for error isparticularly acute, and the limits on the process available in typical casesmay appear too stringent The usual solution to such problems, how-ever, is for the legal system to add a more extensive appellate process forparticular categories of cases rather than to permit tribunals to retreatfrom res judicata in its narrow form Criminal defendants in the UnitedStates, for example, receive additional layers of appellate review notmade available to civil litigants and capital defendants
becausethe costs of error are great and irreversible
are afforded proceduralprotections not made available to other criminal defendants

3.1.2 WaiverThe rationale for rules of waiver must be somewhat different By defi-nition, waiver applies to arguments and issues that were not adjudicatedpreviously but that could have been There can be no presumption thattheir prior disposition was correct if there was no prior disposition.But rules of waiver might be based on a related presumption
if

a litigant did not bother to raise an argument previously, perhaps thelitigant has revealed it to be weak, so that the likely error cost of ignoring

it is small Rules of waiver encourage litigants to bring all potentiallymeritorious arguments before the court at once so that the dispute can

be resolved with dispatch and the gains from compliance with the lawcan be realized more quickly; any claims ‘‘waived’’ are presumed to be

so weak that they need not be addressed

This simple intuition may have much to do with the justificationfor doctrines of waiver, but it is incomplete for two reasons First, to thedegree that complainants internalize the costs of delay in bringingother parties’ behavior into conformity with the law, the legal system

seemingly has no interest in encouraging complainants to pursue pliance at a faster clip Second, litigation becomes more expensive asmore claims are brought Each claim must be researched, briefed, andargued Factual support must be amassed Even if the adjudicative bodycan exercise ‘‘judicial economy’’ to avoid issues that need not be reached

com-to resolve the case, the parties com-to the proceeding must still bear tional costs as the number of issues and arguments grows Hence,

addi-if a complainant prefers to start with what it believes to be its strongestclaims, and to leave others in abeyance should the initial claims fail,some of the costs of litigation (including some that are externalized)will be avoided if the initial claims succeed and resolve the dispute.This consideration, too, seems to argue for allowing the complainant

to bring claims at its own pace, in preference to rules of waiver thatpenalize claimants for failing to bring issues before the dispute process

a complaining nation does not bear all of these fixed costs

The presence of considerable fixed costs to litigation can supply

a positive externality to the consolidation of claims in an initial ceeding Plausibly, a complainant might prefer to proceed more or lessseriatim with its claims to save itself the variable costs of litigatingmatters that may prove unnecessary But if considerable economies ofscale are lost when the complainant proceeds in this fashion and thosecosts are borne by others, the system may gain by foreclosing such

pro-a strpro-ategy

Of course, the mere existence of fixed costs is not sufficient to justifyrules of waiver Their magnitude must be considered in relation to theadded variable costs of litigating more claims at once, claims that mayprove unnecessary to litigate ex post Roughly speaking, the greater thefixed costs of a proceeding in relation to the variable costs per claim,the stronger the case for insisting that more issues be raised at once.These points also suggest the possibility of more refined waiver rules.Some types of claims may have very low marginal litigation costs,

perhaps because they resolve themselves readily with reference to factsalready in evidence Others may require much additional fact-finding

or investigation Similarly, some claims may be legally straightforward,and others may be highly debatable and complex The case for a rule

of waiver is stronger with respect to the relatively cheaper factual andlegal claims

Likewise, the fixed costs of each proceeding may vary with the nature

of the proceeding With particular reference to Art 21.5 compliancePanels, we note that these Panels are comprised of the same members

as the original dispute Panel Each panelist is presumably familiar withthe facts and basic legal issues of the case already Under these circum-stances, it may make sense to have somewhat more lenient rules ofwaiver because the fixed costs of the second proceeding will tend

to be smaller in relation to the variable costs of litigating moreissues initially

We note one further consideration that may have some bearing onrules of waiver The compliance Panel in the Bed Linen case noted itsconcern for ‘‘manipulative and abusive’’ litigation tactics It did notdetail its fears in this regard, but its phrasing hints at concern forvexatious proceedings, brought not because of their potential legalmerits but because of their capacity to harass the respondent There is

a considerable economic literature on the use of frivolous litigation

to extract settlements, and it is possible that complainants in theWTO might hope to extract trade concessions in meritless casesfrom respondents anxious to avoid litigation costs For such a strategy

to justify rules of waiver, of course, it must be the case that litigation

is more expensive seriatim than in a consolidated proceeding,presumably because of the fixed costs noted above
otherwise,vexatious claims brought all at once would be just as effective

‘‘harassment.’’

Although waiver rules may make vexatious litigation less some in the presence of fixed costs, other procedural devices are bettertailored to address the problem of vexatious claims Parties whobring claims that are adjudged to be frivolous can be sanctioned in

trouble-a vtrouble-ariety of wtrouble-ays (‘‘Rule 11 strouble-anctions’’ in the ptrouble-arltrouble-ance of Americtrouble-an civilprocedure) A potentially effective sanction is fee shifting, where thecomplainant must pay the litigation costs of the respondent Suchmeasures target frivolous litigation directly without affecting the timing

of potentially meritorious claims Hence, waiver rules are at most

a ‘‘second-best’’ response to vexatious litigation

3.1.3 Modeling the effect of waiver rules in the WTOFollowing the literature on efficient legal procedure, we would ideallylike to model the problem of designing optimal rules of waiver.One would ask the question whether, in the absence of waiver rules,WTO complainants would bring too few claims at a time from a socialstandpoint If so, one would then inquire whether waiver rules couldcorrect the problem.

Such an analysis would be extremely complex, however, requiringattention to the social gains from litigation, the timing of those gains,and the magnitude and timing of litigation costs The matter becomesall the more complicated in cases with multiple legal claims, in thatthe social returns to the proper adjudication of each claim may vary.Yet another complication is the fact that WTO litigants are govern-ments It is a commonplace in the procedure literature to treat litigants

as expected profit maximizers, but governments cannot be presumed

to behave in this fashion Indeed, in WTO litigation, money rarelychanges hands

We thus limit ourselves to a very simple treatment of one piece of thepuzzle that abstracts from these difficult issues Our focus is on thequestion of how a rule of waiver affects the number of claims broughtbefore the dispute resolution process, and thus the total variable costs

of litigation

Consider a two-period model If the claimant wins at least one claim

in period 1, it wins an award that it values at B1 If it fails to win oneclaim in period 1 but wins at least one claim in period 2, it wins anaward that it values at B2, B2< B1 The difference between B1 and B2reflects the cost to the claimant of delay in receiving the remedy Assumethat there are many potential arguments available to the claimant,and approximate these by a continuum of claims with measure M The(small) claim i has the (small) probability p(i)di of success It is brought

at marginal cost c(i)di It is optimal for the claimant to bring its ‘‘best’’claims first; i.e those with the highest p(i)/c(i) Thus, we order theclaims so that p(i)/c(i) is a non-increasing function Let C(x) be thetotal cost of bringing the set of claims [0, x].12

In a judicial system with waiver, the claimant must raise all arguments

in a single ‘‘case.’’ The case comprises the claims [0, x], where x is

a decision variable for the claimant In a judicial system without waiver,the claimant may return to the adjudicating body with additional claims

12 That is, CðxÞ ¼ R x

cðiÞdi:

in period 2, if it fails to win at least one of its claims in period 1 In thissetting, the claimant brings claims [0, x1] in period 1 and claims[x1, x1þx2] in period 2 (if there has been no success in period 1).Let F(x) be the probability that there is at least one successful claimfrom among those in [0, x]; F(0) ¼ 0, F(M)  1 The density, f (x), is theprobability that the first success comes on claim x The hazard rate,

f (x)/[1
F(x)] is the probability that a success comes on x, given thatthere has been no success on any claims before x Since the claimsare independent, this is just p(x) Bayes Rule tells us that the probability

of at least one success before x1þx2, F(x1þx2), is equal to theprobability of at least one success before x1plus the probability of nosuccesses before x1times the probability of at least one success between

FðxÞB1
CðxÞThe first order condition is

f ðxÞB1 ¼cðxÞ;

that is, the claimant chooses the marginal claim x to equate theprobability that the first success will come on that claim times theaward from winning the case to the marginal cost

3.1.3.2 Judicial system without rule of waiver Here the claimantcan spread claims over the two periods, coming back in period 2 ifthe claims brought in period 1 do not succeed The claimant chooses

f ðx1þx2ÞB2
½1
Fðx1Þcðx1þx2Þ 0, ð2Þwhere expression(2)holds with equality if and only if x24 0 Assumethat it is profitable to bring a second case, which requires a sufficientlylarge B2(if a second case is not profitable, a rule of waiver has no effect).Then f (x1þx2)B2
[1
F(x1)]c(x1þx2) ¼ 0, which we can substituteinto expression(1)to obtain

f ðx1ÞðB1
B2Þ
f ðx1ÞCðx1Þ
Fðx1Þcðx1Þ þf ðx1ÞCðx1þx2Þ ¼0:

ð3Þ

3.1.3.3 An example Consider the case of a constant hazard rate,which arises when the probability of success in each claim is the same.Then F(x) ¼ 1
e
lxfor some l 4 0 Suppose also that every claim hasthe same cost, so that c(x) ¼ c and C(x) ¼ cx All claims are symmetricunder these assumptions, but there are still diminishing returns tofiling more and more claims, and the claimant generally will not fileall of them

3.1.3.4 The example with waiver The first order condition becomes

le
lxB1¼cor

elx¼lB1

c :For convenience, define yi¼lBi/c, so the last expression can bewritten as

elx¼
1 or x ¼ log
1=l:

3.1.3.5 The example with no rule of waiver If x24 0, thenexpression(2)implies

le
lðx1 þx 2 ÞB2 ¼e
lx1cor

elx2 ¼lB2

c ¼
2:Taking logs, x2¼log y2/l

Now, for the choice of x1, expression (3)becomes

le
lx1ðB1
B2Þ
le
lx1cx1
ð1
le
lx1Þc þ le
lx1cðx1þx2Þ ¼0or

celx1¼lB1
lB2þcð1 þ log
2Þ,where this last expression uses lx2¼log y2 from the solution for x2

We can now write

elx1 ¼
1
ð
2
1
log
2Þ:

The term in parentheses is positive for all y24 1 (and unless

y24 1, x2¼0) Thus, comparing the solution for the case with a rule

of waiver, the claimant files more claims in the first round under a rule

of waiver than without waiver, as one might expect

Note further that

elðx1 þx 2 Þ

¼elx1elx2¼
2ð
1

2þ1 þ log
2Þ,

elðx1 þx 2Þ

1 ¼ ð
1

2Þð
2
1Þ þ
2log
2>0,

which implies that the total number of claims filed in the two caseswithout a rule of waiver (if the second case is necessary) exceeds thenumber of claims that will be filed in a single case under a rule of waiver.Finally, consider the expected number of claims brought undereach regime With no rule of waiver, expected claims are

x1þ ½1
Fðx1Þx2 ¼x1þe
lx1x2 Using the results above, this sion becomes

expres-log
2lð
1

2þ1 þ log
2Þþ

logð
1

2þ1 þ log
2Þ

This expression can be greater or smaller than the solution for x(the expected and actual number of claims) under a rule of waiver,log y1/l We have evaluated the difference numerically, and find thatits sign depends on the sizes of the two parameters, y1and y2

For example, setting y1¼3 and allowing y2 to vary over its fullpossible range (from 1 at the lowest, given that x2 positive requires

y24 1, to 3 at the highest, since y14 y2), we find that the differencerises steadily from zero In this case, the expected number of claimswithout a rule of waiver is higher for all y22 (1,3], and the difference

is larger the greater y2is This suggests that a rule of waiver will reducethe variable costs of litigation

But when y1¼5, a somewhat different picture emerges Thedifference between the expected number of claims without and with

a rule of waiver rises above zero as y2 rises initially, but it reaches

a maximum and turns negative as y2approaches y1 And when y1¼10,the difference is negative for all y2above approximately 4.0, so that theexpected number of claims with a rule of waiver can clearly exceedthe number without a rule of waiver for some parameter values

In particular, when y1is ‘‘high’’ and the cost of delay is not too great(y1
y2is not too big), a rule of waiver actually increases the expectednumber of claims filed

The explanation for these findings is rather subtle Fix y1and consider

an increase in B2, which increases y2 Clearly, this has no effect on thenumber of claims brought in a regime with a rule of waiver When there

is no rule of waiver, an increase in y2 reduces the number of claimsbrought in a first case, but it increases both the probability that there

will be a second case and the number of claims brought in such an event.The net effect on the expected number of claims can be positive ornegative We calculate that an increase in y2 actually decreases theexpected number of claims in a regime without a rule of waiver ifand only if 1 þ (y1
y2)(2
y2)
(y2
1
logy2)< 0 If y1
y2 is small(little cost of delay), this inequality is satisfied for y24 2 þ log y2;i.e y2greater than approximately 3.15 Thus, when y1and y2are bothlarge and the difference is small, the expected number of claims in

a system without a rule of waiver is relatively small compared tothe large number of arguments that the claimant brings with a rule

of waiver

Our results are only the beginning of a full treatment of the issues forreasons noted earlier
we have not modeled the social returns tolitigation or any litigation cost externalities But even the modest piece

of the problem that we explore is quite complex, as the reader will nodoubt have noticed We can offer little definitive advice other than tourge caution in the evolution of rules of waiver in the WTO (and moregenerally) It is not at all obvious that social welfare will improve iftribunals insist that claimants bring all claims to an initial proceedinglest they be waived Particularly when the fixed costs of additionalproceedings are modest, it may be better to proceed on the strongestclaims first and then to litigate others later only if necessary.And because the fixed cost aspects of Art 21.5 compliance Panels maytend to be relatively modest, special caution is appropriate there.Compliance Panels should perhaps employ a rather narrow conception

of what constitutes the ‘‘same’’ claim when following the AppellateBody’s directive to deny claimants a second bite of the apple

Returning to the facts of the Bed Linen case, it is not obvious to usthat India’s failure to develop its case fully on the non-attribution issueduring the first proceeding should have precluded it from raising theissue again in the compliance proceeding Its arguments regarding

‘‘other factors’’ such as EC input and output prices had not been vettedearlier, nor is it clear that the EC would have suffered any seriousprejudice if forced to address them before the Art 21.5 Panel.The decision by the Panel and the Appellate Body to foreclose thosearguments may encourage WTO litigants to throw the ‘‘kitchen sink’’into their initial complaints and arguments, so that initial Panelproceedings become even more (and perhaps unduly) cumbersome.This is particularly true if, as we imagine being the case, the rules ofres judicata and waiver that apply before compliance Panels will apply