At the Crossroads - The World Trading System and the Doha Round

^

SpringerWienNewYork

European Community Studies Association of Austria (ECSA Austria) Publication Series Volume 8 Herausgegeben von der Osterreichischen Gesellschaft fiir Europaforschung (ECSA Austria)

Springer WienNew York

Stefan Griller (ed.) At the Crossroads: The World Trading System and the Doha Round

SpringerWienNewYork

Univ.-Prof. Dr. Stefan Griller Europainstitut, Wirtschaftsuniversitat Wien

Financial support was given by Bundesministerium Wissenschaft undForschung, Wien

fur

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by photocoping machines or similar means, and storage in data banks. © 2008 Springer-Verlag Wien Printed in Austria SpringerWienNewYork is a part of Springer Science + Business Media springer.com Product Liability: The publisher can give no guarantee for all the information contained in this book. The use of registered names, trademarks. etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Typesetting: Camera ready by editor Printing: Ferdinand Berger & Sohne Gesellschaft m.b.H., 3580 Horn, Austria Printed on acid-free and chlorine-free bleached paper SPIN: 11299127 CIP data applied for

ISSN 1610-384X

ISBN 978-3-211-22403-8 SpringerWienNewYork

Preface Since its foundation in 1995 the legitimacy of the WTO – and herewith the notion of organising global trade on the basis of concepts like trade-liberalisation, treaty-based rule of law and enforceability – is challenged for very different reasons: Some – mostly in the industrialised parts of the world – feel that the system as such or particular aspects like, e.g., the regulations for trade in services might be a threat to democracy, others – mostly from developing countries or newly industrializing countries – deem the system fundamentally imbalanced, allowing for the application of double standards by the developed countries. Besides, issues like the protection of environmental or social standards or measures directed at public health under the conditions of trade-liberalisation rank high on the respective agendas. Although it is evident that many of those arguments are contradictory in themselves and irreconcilable as a whole, the WTO tried to react on this widespread criticism: The November 2001 declaration of the WTO’s Fourth Ministerial Conference in Doha, Qatar under the declamatory heading of the ‘Doha Development Agenda’ provided the mandate for negotiations on a range of subjects. The explicit justification was to place the developing countries’ needs and interests ‘at the heart’ of the organisation’s work-programme, a strong commitment to the observance of internationally recognized core labour standards, and to ensure internal transparency and participation of all Members, to mention just a few items of the ambitious agenda. Although this high level of ambition made immediate success rather improbable, the failure of the WTO’s Ministerial Conference in the Mexican sea-resort Cancún in September 2003 came as a shock to many observers. More so, since it were obviously the rifts between the developed and the developing countries’ positions that contributed the main share to this failure: Instead of reconciling those differences, the course of negotiations in Cancún made them look more insurmountable than ever. Summing up, Cancún’s failure

VI

Preface

left the WTO and its efforts to create a new basis for global trade in a state of paralysis, thus adding considerably to its already existing legitimacy-problems that stood at the outset of the entire endeavour. This situation was aggravated considerably by the fact that the negative outcome of Cancún was followed by a rather similar result of the ensuing Ministerial in Hong Kong in December 2005. Shortly after the Cancún-breakdown, in November 2003, ECSA Austria organised a conference in Vienna on ‘The WTO after the Failure of Cancún’. This volume not only documents the proceedings of this conference but, against the backdrop of the ongoing efforts to revive the Doha Round wants to be understood as some sort of stocktaking: What did really go wrong in Cancún? Where does the WTO stand today? Has there been any progress made since? And, finally: Is there any chance left to reach the ambitions goals postulated by the Doha Development Agenda? These and other, related questions are being discussed in the volume at hand. Most of the contributions presented here reflect the state of negotiations as it stood in 2004, and unfortunately, due to technical and organisational problems, the volume can appear only now, three years later. However, progress in the Doha Round negotiations since then are almost negligible. Therefore we decided to publish this volume despite this delay, expecting that readers will share our view that the issues discussed therein are as topical as they have been before. The realisation of the conference ‘The WTO after the Failure of Cancún’ as well as this book was only possible with the financial support provided by the European Commission, DG Education and Culture and the Austrian Ministry for Science and Research. On behalf of ECSA Austria, I would like to thank these institutions for their help. Vienna, September 2007 Stefan Griller

Table of Contents

I.

WTO-Law: Evolution and Context Joost Pauwelyn How to Win a WTO Dispute Based on Non-WTO Law? Questions of Jurisdiction and Merits

1

Gabrielle Marceau / Anastasios Tomazos Comments on Joost Pauwelyn’s Paper: ‘How to Win a WTO Dispute Based on Non-WTO Law?’

54

Erich Vranes Comments on Joost Pauwelyn’s Paper: ‘How to Win a WTO Dispute Based on Non-WTO Law?’

83

William J. Davey The Quest for Consistency: Principles Governing the Interrelation of the WTO Agreements

101

Lorand Bartels Treaty Conflicts in WTO Law – A Comment on William J. Davey’s Paper ‘The Quest for Consistency’ 129 Peter-Christian Müller-Graff Protectionism or Reasonable National Regulation? The Protection of Non-Economic Interests as Barriers to the Free Movement of Goods: A Comparison of EC Law and 147 WTO Law Piet Eeckhout Trade and Non-Economic Policies in the EU and in the WTO. A Comment on Peter-Christian Müller-Graff’s Paper ‘Protectionism or Reasonable National Regulation?’

169

Thomas Cottier / Satoko Takenoshita Decision-making and the Balance of Powers in WTO Negotiations: Towards Supplementary Weighted Voting

181

VIII

Table of Contents

II. Items on the Agenda Fritz Breuss Does the ‘Development Round’ Foster Development?

231

Alan Matthews Agriculture after Cancún

315

Markus F. Hofreither Cancún and beyond – A European Perspective of Agricultural Issues

339

Wolfgang Weiß GATS and Domestic Regulation – A Threat to Democracy?

369

Jacques H. J. Bourgeois Post Cancún WTO TRIPs – A Bumpy Road

385

Abbreviations

401

The Authors of this Volume

407

Joost Pauwelyn*

How to Win a WTO Dispute Based on Non-WTO Law? Questions of Jurisdiction and Merits** I. II.

Introduction Why and How Can WTO Panels Apply Non-WTO Law? III. Because of Non-WTO Law, the WTO Panel Has No Jurisdiction A. A Bilateral Agreement not to Appeal a WTO Panel B. A Bilateral Agreement not to Invoke WTO Dispute Settlement C. A Treaty Conferring Exclusive Jurisdiction to Another Tribunal D. A Treaty Providing for Choice of Forum but Making any Choice Exclusive E. A Treaty Providing for Compulsory (but not Exclusive) Jurisdiction to Another Tribunal F. Res judicata Effect of Rulings by Other Courts or Tribunals IV. Because of Non-WTO Law, a WTO Violation is Justified A. Non-WTO Defenses Explicitly Incorporated into the WTO Legal System B. WTO Violations Permitted (or Even Imposed) Pursuant to the Dispute Settlement Provisions of Another Treaty *

**

2 6 14 16 17 18 24 26 32 35 36

39

Associate Professor, Duke University School of Law, formerly with the Legal Affairs Division and the Appellate Body Secretariat of the WTO. This paper originally was published in Journal of World Trade 37 (2003), No. 6 and is reproduced here by courtesy of Kluwer Law International of WoltersKluwer.

2

Joost Pauwelyn

C. WTO Violations Permitted (or Even Imposed) Pursuant to the Provisions of Another Treaty D. WTO Violations Permitted under Another Treaty on Condition that the WTO Panel Finds that this Other Treaty is Respected/Violated V. Conclusion References

41

44 47 50

I. Introduction It has become standard practice for WTO panels and the Appellate Body to use non-WTO law when interpreting the meaning of terms in the WTO agreement. Such interpretation can, for example, lead to broader GATT exceptions, as in US – Shrimp, where the Appellate Body interpreted the words ‘exhaustible natural resources’ in GATT Article XX(g) with reference to certain environmental treaties.1 It may also narrow the scope of GATT rules or exceptions, as shown in the ICJ Case Concerning Oil Platforms, where a treaty provision similar to GATT Article XXI(b)(iii) on essential security interests was interpreted restrictively with reference to rules of general international law prohibiting the use of force.2 In addition to treaty interpretation, non-WTO law is commonly referred to also to fill largely procedural gaps in the WTO agreement: the WTO agreement is silent on questions such as burden of proof, standing, representation before panels, the retroactive application of treaties or error in treaty formation; as a result, reference has been made to rules of general international law addressing those questions, essentially custom or general principles of law binding on all states.3 1

2

3

Appellate Body Report, United States – Import Prohibition of Certain Shrimp and Shrimp Products (‘US – Shrimp’), WT/DS58/AB/R, adopted 6 November 1998, DSR 1998:VII, 2755, paras. 128-132. On the process of interpreting the WTO treaty with reference to nonWTO law: Marceau (1999) and Pauwelyn (2003a), 244-274 (Treaty interpretation as a conflict-avoidance tool). Case Concerning Oil Platforms (Islamic Republic of Iran v United States of America), Judgment of 6 November 2003, at http:// www.icj-cij.org/icjwww/idocket/iop/iopjudgment/ioptocjudgment(s). htm, discussed infra footnote 9. For an overview of relevant case law: Cameron / Gray (2001) and Pauwelyn (2001), 563. See also Hilf (2001). For an increasingly isolated critique see McGinnis (2003), 36 (‘I would not interpret the

How to Win a WTO Dispute Based on Non-WTO Law?

3

As important as those two processes of reference to non-WTO law may be, it is unlikely that they determine the substantive outcome of a WTO dispute: the process of treaty interpretation stops there where the words in the relevant WTO provision are unambiguously clear (interpretation contra legem is prohibited); applying the largely procedural rules of general international law may eventually decide a case, but hardly influences its substantive merits. This essay takes the relevance of non-WTO law before WTO panels a step further. It examines instances where non-WTO law constitutes an independent defense against claims of violation of WTO law. Elsewhere,4 I have set out a conceptual framework that permits WTO panels to take account of such independent defenses under non-WTO law (summarized in Section I below). In this essay, the main objective is rather to offer real-life examples where non-WTO law can play this role based, first, on specific disputes that have most recently come before WTO panels or other tribunals and, second, on newly negotiated treaties with a trade component (ranging from the Kimberley process on conflict diamonds5 to the WHO Framework Convention on Tobacco Control6). In this exercise, it is useful to distinguish two types of cases: First, cases where the invocation of non-WTO law may lead a panel to find that it has no jurisdiction (Section II); Second, cases where non-WTO law may effectively justify what would otherwise be a

4 5 6

WTO to permit it to be … supplemented by emerging customary international law … The WTO represents a comprehensive code for the overriding purpose of expanding trade. Therefore, it is not amenable to supplementation by other rules with other objectives that will detract from this purpose’): It is difficult to see, however, how WTO panels could operate outside general international law and have to invent from scratch, for example, rules on burden of proof, good faith, state responsibility etc.; nor is it easy to square Professor McGinnis’ view with the Vienna Convention rules on treaty interpretation, explicitly confirmed in Article 3.2 of the DSU and numerous Appellate Body decisions. Pauwelyn (2003a), 440-478 and Pauwelyn (2001), 559-565. The text of the Kimberley Process Certification Scheme can be found at http://www.kimberleyprocess.com. Adopted unanimously by the 56th World Health Assembly on 21 May 2003. The final text is contained in World Health Assembly Resolution 56.1, available at http://www5.who.int/tobacco/ page.cfm?sid=96.

4

Joost Pauwelyn

breach of WTO rules (Section III). By way of conclusion, the essay summarizes the different alternatives available to WTO panels faced with defenses under non-WTO law, as well as some of the practical consequences and policy concerns that they entail. To date WTO panels and the Appellate Body have been able to avoid the question of whether defendants can win a WTO dispute based solely on non-WTO law. In the not so distant future, they will no longer be able to hold off this boat: Firstly, because of the ever increasing interaction between WTO law and other branches of international law, be it regional trade agreements or other treaties or regimes with a trade component (such as the Cartagena Biosafety Protocol7 or recommendations by the International Labor Organization (ILO) in respect of Myanmar8); Secondly, because of the growing willingness of WTO members to explicitly rely on these other sources of law even before a WTO panel, inspired largely by the compulsory nature of WTO dispute settlement: Any trade-related policy of all WTO members can now be challenged at the WTO without the possibility for defendants to block the process. Defendants are, therefore, more likely to invoke all possible defenses, including those to be found under non-WTO law. If and when the occasion arises, it is crucial that the Appellate Body realizes the centrality of this question both for the WTO and the wider system of international law (indeed, questions of overlapping rules of international law arise not only at the WTO, but increasingly also before the International Court of Justice9, the World 7 8 9

Cartagena Protocol on Biosafety to the Convention on Biological Diversity, Jan. 29, 2000, 39 ILM 1027. See infra footnote 99. See the recent Case Concerning Oil Platforms (Islamic Republic of Iran v United States of America), Judgment of 6 November 2003, at http: //www.icj-cij.org/icjwww/idocket/iop/iopjudgment/ioptocjudg ment(s).htm. In that case, the interaction was between Article XX:1(d) of the Iran-US Treaty of Amity of 1955 (permitting ‘measures necessary to protect essential security interests of a party’) and general international law rules prohibiting the use of force, in particular those on self-defense. The question was whether US attacks on Iranian oil platforms could be justified as ‘measures necessary to protect essential security interests’ of the United States, in line with Article XX:1(d), and whether in the examination of this question general international law rules on the use of force played a role. The ICJ decided to interpret Article XX:1(d) with reference to rules on

How to Win a WTO Dispute Based on Non-WTO Law?

5

Bank and the IMF,10 UNCLOS and investment-related arbitrations11 as well as regional institutions such as the courts of the European Union12 and the European Court of Human Rights13). After all, the question is crucial not only when formally raised before a WTO panel, it is decisive also when national parliaments and administrations are considering the adoption of new policies – ranging from trade embargoes to the adoption of a new treaty with a commercial impact – and raise the question of whether a proposal, based on non-WTO rules of international law, would pass muster before a WTO panel. In this sense, offering legal predictability to national decision-makers is as important as resolving the systemic questions concerning the interaction between WTO law and other rules of international law.

10

11

12

13

the use of force (invoking Article 31.3(c) of the Vienna Convention on the Law of Treaties in support); found that the US attacks were not justified under those rules as acts of self-defense and, on that basis, concluded that the US attacks could not be seen either as ‘necessary to protect essential security interests’ of the United States. In the end, however, this lack of justification under Article XX:1(d) did not play a role since the ICJ later decided that the US attacks did not breach the 1955 Treaty of Amity in the first place (hence there was no need to justify US conduct under Article XX:1(d) to begin with). In particular, an increasing number of overlaps arise between the rules of the Bank and the IMF, on the one hand, and UN related rules and decisions, on the other. This overlap is complicated by the stated non-political nature of the Bank and the IMF (set out in Arts. IV, section 10 and III, section 5(b) of the Bank’s Articles of Agreement and Arts. I (v) and V, section 3 of the IMF’s Articles of Agreement). See also the role of, for example, international human rights standards before the World Bank’s Inspection Panel in cases such as Chad: Petroleum Development Project, discussed in World Bank (2003), 96. In respect of UNCLOS, see infra footnotes 19, 64, 68 and 75. For a glimpse of potential overlaps in investment disputes, see Verhoosel (2003) and the ICSID case Compania de Aguas del Aconquija S.A. & Vivendi Universal v Argentine Republic 40 ILM 426 (2001) and 41 ILM 1135 (2002), discussed infra in footnotes 38, 51, 55, 59, 69 and 72. See, for example, Opel Austria v Council, Case C-115/94 [1997] ECR II-39 and Racke GmbH v Hauptzollamt Mainz, Case C-162/96 [1998] ECR I-3655. See, for example, Caflisch (2002), 1-12.

6

Joost Pauwelyn

II. Why and How Can WTO Panels Apply Non-WTO Law? It is undisputed that WTO panels have jurisdiction only to decide on claims under WTO covered agreements. Under Article 1.1, the DSU14 applies only to ‘disputes brought pursuant to the consultation and dispute settlement provisions of the agreements listed in Appendix 1 to the [DSU]’, i.e., the so-called WTO covered agreements. The question here is, rather, whether in the examination of such WTO claims, non-WTO law can offer an independent defense or justification that precludes a panel from finding a breach of WTO law. This question relates to the scope of the applicable law before a WTO panel, a matter to be distinguished clearly from that of the jurisdiction of WTO panels.15 The difference between jurisdiction and applicable law is well known and accepted in other international courts and tribunals,16 though often neglected at the WTO. Making an analogy with the limited jurisdiction of the International Court of Justice (ICJ), the Lockerbie cases decided by the ICJ perfectly illustrate the difference between jurisdiction and applicable law: There, the ICJ had jurisdiction to consider Libyan 14 15 16

Understanding on Rules and Procedures Governing the Settlement of Disputes, Annex 2 to the Marrakesh Agreement (DSU). Bartels (2001) and Pauwelyn (2003a), 443-472. As most recently noted by an Arbitral Tribunal under Annex VII of UNCLOS: ‘The Tribunal agrees ... that there is a cardinal distinction between the scope of its jurisdiction ..., on the one hand, and the law to be applied by the Tribunal ..., on the other hand’ (Mox Plant case (Ireland v United Kingdom), Order No. 3 of 24 June 2003, at http://www.pcacpa.org/PDF/MOX%20Order%20no3.pdf, p. 6, para. 19). In the same vein, see the investor-state dispute settlement mechanism under Chapter 11 of NAFTA: Arts. 1116-7 of NAFTA entitled ‘Claim by an Investor of a Party…’ limits the jurisdiction of NAFTA arbitration tribunals to claims of violation of Section A of NAFTA Chapter 11 and NAFTA Arts. 1503(2) and 1502(3)(a); in contrast, Art. 1131 of NAFTA entitled ‘Governing Law’ sets out the broader scope of the applicable law to be considered in examining the validity of those enumerated NAFTA claims (‘A Tribunal established under this Section shall decide the issues in dispute in accordance with this Agreement [that is, all NAFTA provisions] and applicable rules of international law’, emphasis added).

How to Win a WTO Dispute Based on Non-WTO Law?

7

claims only under the Montreal Convention. However, this did not stop it from also examining other international law, in particular UN Security Council Resolution 748 invoked in defense by the United Kingdom and the United States, as part of the applicable law.17 The question of applicable law before WTO panels is addressed in DSU Article 7. It instructs panels to examine the matter referred to them ‘in the light of the relevant provisions’ of the covered agreement(s) cited by the parties to the dispute and to ‘address the relevant provisions in any covered agreement or agreements cited by the parties to the dispute’. This provision imposes an obligation on panels to address and possibly apply certain WTO rules. At this juncture, two possibilities arise: One either considers this reference to WTO rules as an exhaustive list of all rules that WTO panels can possibly apply;18 Or one holds the view that confirming the relevance of some rules – in casu those that WTO panels will most commonly be asked to apply, i.e. WTO rules – does not preclude that WTO panels may apply also other, non-WTO rules in particular circumstances. In my opinion, the latter view must prevail, for several reasons. First, as a practical matter, WTO case law shows that WTO panels and the Appellate Body have not limited themselves to the four corners of WTO covered agreements: they have referred to general principles of law, customary international law and even other, non-WTO treaties.19 Second, DSU Article 3.2 explicitly con17

18

19

Questions of Interpretation and Application of the 1971 Montreal Convention Arising from the Aerial Incident at Lockerbie (Libya v U.S.), Provisional Measures, 1992 ICJ REP. 114, para. 42 (Apr. 14). See, for example, Trachtman (1999), 342 (stating that the explicit language in the DSU ‘would be absurd if rights and obligations arising from other international law could be applied by the DSB’ and that ‘[w]ith so much specific reference to the covered agreements as the law applicable in WTO dispute resolution, it would be odd if the members intended non-WTO law to be applicable’). See the case law summarized in the references supra footnotes 1, 3 and 15. As a matter of fact, panels had little choice but to apply certain rules of general international law. Or should they have re-invented from scratch rules on, for example, burden of proof, good faith, pacta sunt servanda, etc.? As stated in a recent award under the 1992 Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR Convention):

8

Joost Pauwelyn

firms that WTO covered agreements must be clarified ‘in accordance with customary rules of interpretation of public international law’. Article 31(3) of the Vienna Convention on the Law of Treaties, part of the rules of interpretation thus referred to, directs that in interpreting a treaty, account must be taken not only of the treaty itself (in casu, the WTO treaty), but also of ‘any subsequent agreement between the parties regarding the interpretation of the treaty or the application of its provisions’, as well as ‘any relevant rules of international law applicable in the relations between the parties’. The WTO treaty thereby explicitly frames itself in the wider context of public international law, including other non-WTO treaties. Third, and most importantly, the WTO agreement is a treaty part of public international law. By definition, and even without the explicit confirmation in DSU Article 3.2, the WTO agreement cannot, therefore, be applied in isolation from other rules of international law. Just as private contracts are automatically born into a system of domestic law, so treaties are automatically born into the system of international law. Much the way private contracts do not need to list all the relevant legislative and administrative provisions of domestic law for them to be applicable to the contract, so treaties need not explicitly set out rules of general international law for them to be applicable to the treaty (for example, the text of the Vienna Convention does not have to be attached to the new treaty for general international law rules on the law of treaties to be applicable to it).20

20

‘It should go without saying that the first duty of the Tribunal is to apply the OSPAR Convention. An international tribunal, such as this Tribunal, will also apply customary international law and general principles unless and to the extent that the Parties have created a lex specialis. Even then, it must defer to a relevant jus cogens with which the Parties’ lex specialis may be inconsistent’ (Permanent Court of Arbitration, Dispute Concerning Access to Information under Article 9 of the OSPAR Convention (Ireland v United Kingdom), Final Award, 2 July 2003, para. 84 at http://www.pca-cpa.org/PDF/ OSPAR%20Award.pdf). For example, the obligation to interpret treaty provisions in the light of rules of general international law, as called for in Article 31.3(c) of the Vienna Convention, was confirmed in the Case Concerning Oil Platforms (Islamic Republic of Iran v United States of America), Judgment of 6 November 2003, supra footnote 9. This rule of treaty interpretation was found to apply even though the Treaty of Amity in

How to Win a WTO Dispute Based on Non-WTO Law?

9

The same applies as regards other, non-WTO treaties: the WTO treaty was created next to other pre-existing treaties and, in turn, continues its existence in a framework where other, new treaties are created next to it. True, unlike Article 293 of the UN Law of the Sea Convention (UNCLOS), Article 1131 of NAFTA and Article 38 of the ICJ Statute – which explicitly include other rules of international law as part of the applicable law – the DSU does not explicitly confirm its creation and existence in international law. However, given the nature of the WTO agreement as a treaty under public international law, there was no need for the DSU to do so. On the contrary, the principle is that all other international law continues to exist next to the WTO treaty unless the WTO treaty explicitly deviates or contracts out of this other law.21 In other words,

21

dispute did not explicitly incorporate or refer to the Vienna Convention rules on treaty interpretation. Confirmed in the WTO by the Panel Report on Korea – Measures Affecting Government Procurement, WT/DS163/R, adopted 19 June 2000, para. 7.96 (‘Customary international law applies generally to the economic relations between the WTO Members. Such international law applies to the extent that the WTO treaty agreements do not ‘contract out’ from it. To put it another way, to the extent there is no conflict or inconsistency, or an expression in a covered WTO agreement that implies differently, we are of the view that the customary rules of international law apply to the WTO treaties and to the process of treaty formation under the WTO’) and the Appellate Body Report on United States – Anti-Dumping Measures on Certain Hot-Rolled Steel Products from Japan, WT/DS184/AB/R, adopted 23 August 2001, para. 60 and footnote 40 (‘We observe that the rules of treaty interpretation in Articles 31 and 32 of the Vienna Convention apply to any treaty, in any field of public international law, and not just to the WTO agreements’; ‘It might be possible for the parties to a treaty expressly to agree that the rules of treaty interpretation in Articles 31 and 32 of the Vienna Convention do not apply, either in whole or in part, to the interpretation of a particular treaty. Likewise, the parties to a particular treaty might agree upon rules of interpretation for that treaty which differ from those rules of interpretation in Articles 31 and 32 of the Vienna Convention. But this is not the case here’). For a confirmation by other courts and tribunals: Georges Pinson case, Franco-Mexican Commission (Verzijl, President), A.D. 1927-8, No. 292, para. 50 (‘Every international convention must be deemed tacitly to refer to general principles of international law for all questions which it does not itself resolve in ex-

10

Joost Pauwelyn

there was no need for Article 7 of the DSU to explicitly include also other rules of international law as part of the applicable law before WTO panels; to the extent that those other rules were not deviated from in the WTO treaty, this is automatically the case. Finally, an oft-cited provision in support of the alleged obligation on WTO panels to contain themselves to WTO covered agreements, is DSU Article 3.2 which provides: ‘Recommendations and rulings of the DSB cannot add to or diminish the rights and obligations provided in the covered agreements’.22 Should this provision be read as saying that WTO panels, the Appellate Body, and the DSB cannot ever add to or diminish the rights and obligations explicitly set out in WTO covered agreements?23 For the reasons set

22 23

press terms and in a different way’); Chorzow Factory (Merits), PCIJ, Ser. A, no. 17, 29 (1928) (‘Reparation is the indispensable complement of a failure to apply a convention, and there is no necessity for this to be stated in the convention itself’); Legal Consequences for States of the Continued Presence of South Africa in Namibia (South West Africa), Advisory Opinion, ICJ Reports 1971, 16, para. 96 (‘it would be necessary to show that the mandates system … excluded the application of the general principle of law that a right of termination on account of breach must be presumed to exist in respect of all treaties’); Elettronnica Sicula S.p.A (ELSI) case, ICJ Reports 1989, 42, para. 50 (in respect of the obligation to exhaust local remedies, ‘the Chamber finds itself unable to accept that an important principle of customary international law should be held to have been tacitly dispensed with, in the absence of any words making clear an intention to do so’); and Permanent Court of Arbitration, Dispute Concerning Access to Information under Article 9 of the OSPAR Convention (Ireland v United Kingdom), Final Award quoted supra in footnote 19. A provision repeated in Article 19.2 of the DSU. In support of this position, see, for example, McGinnis (2003). For Professor McGinnis, WTO rules can only be affected by amendment, interpretation and waiver procedures in the WTO treaty itself; not by other rules of international, created outside the WTO, even if those rules have been agreed to and are binding on the disputing WTO members. He thus posits the WTO treaty as a completely self-contained regime, de-linked from other treaties and custom, and obscures the fact that treaty relations between states cannot only be affected pursuant to, for example, the amendment procedures in the particular treaty, but also by the conclusion of other treaties which (though not amending the prior treaty) may modify it as between the parties to the

How to Win a WTO Dispute Based on Non-WTO Law?

11

out above, this author thinks not. But what is then the purpose of DSU Article 3.2? In my view, it does not limit the applicable law before WTO panels, nor does it deal with the relationship between WTO covered agreements and all past and future law. Rather, it confirms the rather obvious limits a WTO panel must observe in interpreting WTO covered agreements. The phrase directly follows the one quoted earlier, incorporating rules of interpretation of public international law. In exercising this judicial function of interpretation, WTO panels may clarify the meaning of WTO covered agreements, but they may not ‘add to or diminish the rights and obligations provided in the covered agreements.’ To put it differently, as judicial organs, WTO panels may not create new rights and obligations; they must apply those that WTO members agreed to. However, stating what the judiciary can do with the law differs greatly from stating what the legislature (i.e., WTO members) has done, or can do, with the law. Article 3.2 specifies that the WTO judiciary, like any other judiciary, cannot ‘change’ the WTO treaty. However, that does not limit the extent to which WTO members may conclude or have concluded other treaties that can influence their mutual WTO rights and obligations. As important as the distinction is between panel jurisdiction (WTO claims only) and applicable law (potentially all international law), so too is the distinction between new treaty (Vienna Convention on the Law of Treaties, Articles 41 and 58, see footnotes 25 and 26 below). He does so based on arguments that the WTO treaty has more legitimacy and ‘greater broad consensus’ (p. 42). Though this may be true when comparing the WTO treaty to custom (see footnote 110 below), it does not apply to the inter-action between the WTO treaty and other, non-WTO treaties. Rather than preserve the sovereign will of WTO members, not to give affect to such other treaties would undermine state sovereignty and the democratic legitimacy of WTO rules. It would transform the WTO into a trade-only safe heaven, unaffected by other, equally valid expressions of state will enshrined in, for example, international human rights or environmental treaties (making available an exit option not only amongst branches of international law, but also from domestic law). However, this may well be the very objective of Professor McGinnis’ position when he states (McGinnis (2003), 36) that ‘[t]he WTO represents a comprehensive code for the overriding purpose of expanding trade. Therefore, it is not amenable to supplementation by other rules with other objectives that will detract from this purpose’.

12

Joost Pauwelyn

interpreting WTO rules (and the prohibition to add or detract from those rules in the process) and examining WTO claims in the context of other applicable international law (where the expression of state consent and conflict rules of international law must decide the outcome). The only remaining question relates then to the conditions that non-WTO rules of international law must meet to be applicable before a WTO panel in the examination of WTO claims. First, and most obviously, these other rules must be binding on both disputing parties (and be invoked by either of them). If either of the two parties is not so bound, these other rules cannot be held against it. In technical terms, this means that one either applies other rules binding on the disputing parties as part of the applicable law on the ground that the WTO agreement, as a treaty under public international law, must be applied in the context of such other treaties; or that one interprets the relevant WTO rules in the context of such other treaties based on, for example, Article 31.3(c) of the Vienna Convention referring to ‘any relevant rules of international law applicable in the relations between the parties’. The advantage of the latter approach is that DSU Article 3.2 offers an explicit link to Article 31.3(c) of the Vienna Convention. The disadvantage of this approach is, however, that one risks giving too broad a meaning to the term interpretation. Indeed, what one is effectively doing when dis-applying a WTO norm to the advantage of another, nonWTO norm agreed upon only by the disputing parties (or even when applying a rule of general international law to solve a question on which the WTO treaty itself remains silent) is not so much interpreting WTO terms in the light of other norms agreed upon by WTO members. Rather, one is then applying WTO norms together with such other norms as they are binding (only) in the relationship between the disputing parties. The discussion amongst commentators on whether the reference to other rules in Article 31.3(c) of the Vienna Convention includes only other rules expressing the common intentions of all WTO members or also rules that are binding on just the disputing parties, expresses this dilemma of which technique to use to refer to non-WTO rules. This author considers it more appropriate to draw a line between interpretation with reference to other norms and application of other norms. Others end up with the same result – that is, a panel can refer to non-WTO rules binding only on the disputing parties – based solely on a wider no-

How to Win a WTO Dispute Based on Non-WTO Law?

13

tion of treaty interpretation under Articles 31 and 32 of the Vienna Convention.24 Second, for WTO panels to apply non-WTO rules in their decision on WTO claims, such other rules must be both valid and legal. Most importantly, their very conclusion may not be prohibited in the WTO treaty (an example would be an agreement on voluntary export restraints explicitly prohibited in Article 11 of the WTO Safeguards Agreement25). Moreover, these other rules may not affect the rights or obligations of third parties (an example would be a bilateral agreement in which a trade concession is explicitly reserved to the other party to the agreement, in breach of the MFN rights of other WTO members).26 In addition, a treaty altering WTO rights or obligations as between its parties only, may not be concluded by coercion, fraud or corruption nor be based on error; if not, it is invalid.27 In this respect, the risk of powerful WTO members ‘imposing’ bilateral treaties on weaker states is real. It could even be an argument against WTO panels taking account of nonWTO treaties altogether. Nonetheless, this risk is a reality inherent 24 25

26

27

Marceau (1999); Marceau (2001); Marceau (2002); Palmeter / Mavroïdis (1998); Bartels (2002), 366, fn. 723 (on file with the author). A distinction must be made, however, between, on the one hand, an agreement whose conclusion is explicitly prohibited in the WTO treaty (such as voluntary export restraints under Article 11 of the Safeguards Agreement) and, on the other hand, non-WTO rules that simply contradict rules in the WTO treaty (say, an agreement in which the right of appeal is waived, contrary to Article 17 of the DSU or an agreement permitting trade restrictions otherwise not permitted under GATT Article XX). The former agreement is ‘illegal’ (Article 41.1(b) of the Vienna Convention does not permit the inter se modification of a multilateral treaty if such modification is ‘prohibited by the treaty’) and cannot, therefore, be applied in any event; the latter rules are ‘legal’ but conflict with WTO rules and the question is then which of the two rules – the WTO norm or the other norm – prevails in the specific circumstances of the case. Article 41.1(b)(i) of the Vienna Convention on the Law of Treaties prohibits the inter se modification of a treaty in case it ‘affect[s] the enjoyment by the other parties of their rights under the treaty or the performance of their obligations’. The general principle of pacta tertiis nec nocent nec prosunt is stated in Article 34 of the Vienna Convention. Articles 48-52 of the Vienna Convention.

14

Joost Pauwelyn

in the process of international law making. It is present even in the conclusion of WTO agreements (where the risk of powerful states pressing their views on weaker states is as serious as it is in other multilateral fora). Though panels ought to remain sensitive to this reality, it should not become a scapegoat for neglecting non-WTO rules altogether. Third, for non-WTO rules to justify an otherwise WTO inconsistent measure, the other rule must prevail over the contradictory WTO rule pursuant to conflict rules of international law. This may be so because it is stated explicitly in the WTO treaty itself or in the other, non-WTO treaty, or because the non-WTO rule is later in time (lex posterior) or more specific to the circumstances (lex specialis) as compared to the WTO rule.28 We next examine specific instances where WTO panels may thus be called upon to apply non-WTO rules in a way that can lead the defendant to win a WTO dispute: first, for lack of jurisdiction (Section II); second, on the merits (Section III). III. Because of Non-WTO Law, the WTO Panel Has No Jurisdiction In April 1994, WTO members agreed to submit their disputes under WTO covered agreements to the compulsory jurisdiction of WTO panels and the Appellate Body. Does this mean that since then all disputes between WTO members that have the slightest trade component must necessarily be decided at the WTO? This would be hard to imagine. A variety of reasons exist why WTO members may, by common agreement, decide not to go to the WTO: they may prefer to settle a dispute amicably without resort to third-party adjudication; they may consider that regional trade disputes are best settled before regional tribunals29; they may agree that certain disputes are better decided by non-trade tribunals, etc. In the event that two WTO members have thus agreed not to settle a particular dispute at the WTO, can a WTO panel simply ignore this agreement? Or should it take cognizance of it and, as the case may be, find that it has no jurisdiction to decide the case? In my view, the latter ap28 29

On those conflict rules of international law: Pauwelyn (2003a). On the interaction between WTO dispute settlement and dispute settlement under regional trade agreements, see Kwak / Marceau, (2002), 8 and Pauwelyn (2004).

How to Win a WTO Dispute Based on Non-WTO Law?

15

proach is the correct one. As a result, a defendant can effectively win a WTO dispute with reference to non-WTO law, in casu other agreements or treaties that take away the jurisdiction of WTO panels or the Appellate Body for certain cases. Crucially, in this respect, a WTO panel, like all international tribunals, has the competence – even the obligation – to decide the question of its own jurisdiction.30 This is part of the so-called incidental jurisdiction of all international tribunals and need not be conferred explicitly to panels in the DSU. As a result, a panel is not only entitled, it must examine whether its own jurisdiction remains intact or has been undermined by some other agreement between the disputing parties. If the latter, the panel must decline jurisdiction. This incidental jurisdiction of the competence de la competence provides a useful legal basis on which to justify the examination of other agreements. Yet, the underlying principle remains the same as that applicable when a panel applies non-WTO law on the merits (Section III): in the examination of its own jurisdiction / of the merits of the WTO claims before it, the applicable law before a WTO panel ought not be limited to WTO covered agreements, it should include also other relevant international law. The only difference between the effect of non-WTO law on panel jurisdiction and on the merits of a WTO dispute, is that when it comes to panel jurisdiction, a WTO panel may have to check at its own initiative whether non-WTO law undermines its jurisdiction (even if none of the parties themselves refer to this non-WTO agreement); when it 30

That WTO panels as well as the Appellate Body have the jurisdiction to decide on their own jurisdiction is firmly established. The Appellate Body referred to the ‘widely accepted rule that an international tribunal is entitled to consider the issue of its own jurisdiction on its own initiative, and to satisfy itself that it has jurisdiction in any case that comes before it’ (Appellate Body Report, United States – AntiDumping Act of 1916, WT/DS136/AB/R, WT/DS162/AB/R, adopted 26 September 2000, footnote 30). In the Appellate Body Report on Mexico – Anti-Dumping Investigation of High Fructose Corn Syrup (HFCS) from the United States – Recourse to Article 21.5 of the DSU by the United States (WT/DS132/AB/RW, adopted 21 November 2001, at para. 37), it was stated that panels must check the question of their own jurisdiction at their own initiative (‘panels cannot simply ignore issues which go to the root of their jurisdiction … Rather, panels must deal with such issues – if necessary, on their own motion – in order to satisfy themselves that they have authority to proceed’).

16

Joost Pauwelyn

comes to the merits of the dispute, in contrast, the panel is subject to the principle of non ultra petita, that is, it can only examine defences that have been raised explicitly by the defendant itself (the panel cannot itself add defences). Let us illustrate the potential effect of non-WTO law on panel jurisdiction with reference to a number of examples that have arisen in practice either at the WTO itself or before other international tribunals. This section deals, in turn, with (1) a bilateral agreement not to appeal a WTO panel; (2) a bilateral agreement not to invoke WTO dispute settlement; (3) a treaty conferring exclusive jurisdiction to another tribunal; (4) a treaty providing for choice of forum but making any choice exclusive; (5) a treaty providing for compulsory (but not exclusive) jurisdiction to another tribunal; and (6) the related question of res judicata effect of rulings by other courts or tribunals. A. A Bilateral Agreement not to Appeal a WTO Panel Subsequent to the adoption of the original panel report on Australia – Leather, the United States and Australia agreed on how to proceed with the dispute under Articles 21 and 22 of the DSU. Point 4 of this bilateral agreement read as follows: ‘Both Australia and the United States will unconditionally accept the review panel report [pursuant to Article 21.5 of the DSU] and there will be no appeal of that report (emphasis added)’.31 The review panel report then found in favor of the United States. Australia did not appeal the report, but the following question arises: Had Australia nonetheless decided to appeal the review panel report, could the United States have relied on the bilateral agreement in which both parties gave up their right to appeal? In other words, should the Appellate Body have declined to decide the appeal on the ground of this bilateral agreement, an agreement that is not a WTO covered agreement? In my view, the answer must be in the affirmative: both the United States and Australia agreed not to appeal; they did not thereby affect third party rights; hence, the Appellate Body must respect this agreement and decline jurisdiction. By thus applying the bilateral agreement, the Appellate Body 31

Australia – Subsidies Provided to Producers and Exporters of Automotive Leather, Recourse by the United States to Article 21.5 of the DSU, WT/DS126/8, 4 October 1999.

How to Win a WTO Dispute Based on Non-WTO Law?

17

would not expand its jurisdiction beyond WTO claims; rather, it would expand the applicable law before it and on that basis decline to exercise its limited jurisdiction. B. A Bilateral Agreement not to Invoke WTO Dispute Settlement In the case of India – Quantitative Restrictions, complaint by the EC, the parties reached a mutually agreed solution and notified this solution to the DSB pursuant to Article 3.6 of the DSU.32 In this bilateral agreement between India and the EC, dated 12 November 1997, the following was agreed: the European Communities will refrain from action under GATT Article XXII or Article XXIII as regards those restrictions [maintained by India on import of industrial, agricultural and textile products] during the phasing-out period as defined below, as long as India complies with its obligations under this exchange of letters (emphasis added). Subsequent to this agreement, but before the end of the relevant phasing-out period, the EC initiated the dispute on India – Autos. Before the panel, India argued that this ‘new’ India – Autos dispute is covered by the bilateral settlement in India – Quantitative Restrictions and that, therefore, the EC is precluded from invoking WTO dispute settlement procedures on this matter.33 In other words, in that case the question arose whether a bilateral settlement can take away the jurisdiction of a WTO panel. The EC response was that the India – Autos dispute was not covered by the earlier settlement and that, in any event, the bilateral settlement was not a WTO covered agreement so that it could not be relied on by India before a WTO panel.34 The panel was able to avoid the systemic question of whether India could rely on the bilateral settlement by finding that, as a question of fact, the settlement did not cover the matter in the India – Autos case. Hence, even if India could rely on the bilateral settlement, the India – Autos case was in any event not 32

33

34

India - Quantitative Restrictions on Imports of Agricultural, Textiles and Industrial Products, Notification of Mutually Agreed Solution, WT/DS96/8, 6 May 1998. Panel Report, India – Measures Affecting the Automotive Sector (‘India – Autos’), WT/DS146/R and Corr.1, WT/DS175/R and Corr.1, adopted 5 April 2002, at para. 4.30. Ibidem, para. 4.32 and footnote 71.

18

Joost Pauwelyn

covered by the EC’s promise in this settlement not to invoke WTO dispute settlement. The panel did note the following, however: At the very least, the Panel sees merit in India’s argument that the issue in this respect is not solely whether the mutually agreed solution is a covered agreement, but rather, what effects it may have on the exercise of procedural rights under the DSU in subsequent proceedings.35 Indeed, had the new India – Autos dispute been covered by the EC promise in the bilateral settlement not to invoke WTO dispute settlement procedures, in my view, any WTO panel would have been under an obligation to respect this agreement and to declare that by agreement of the parties it does not have jurisdiction to examine the case. Other instances where particular WTO members may agree not to rely on WTO dispute settlement can be imagined. Taiwan and, especially, China, for example, have sent out signals that they do not intend to resort to WTO dispute settlement to resolve trade disputes between mainland China and Taiwan.36 If a bilateral agreement or even a binding unilateral declaration to this effect can be detected, then a WTO panel ought to respect it and apply it as against the member who made such commitment. C. A Treaty Conferring Exclusive Jurisdiction to Another Tribunal Article 292 of the EC Treaty confers exclusive jurisdiction to the European Court of Justice and other EC bodies as follows: ‘Member States undertake not to submit a dispute concerning the interpretation or application of this Treaty to any method of settlement other than those provided therein’. Similar provisions occur in other contexts, such as the Andean Community37 and investment disputes.38 Does Article 292 prevent 35 36 37

Ibidem, para. 7.116. See Kong (2002), 755. Article 42.1 of the Cartagena Agreement establishing the Andean Community – which is made up of Bolivia, Colombia, Ecuador, Peru and Venezuela – states the following: ‘Member countries shall not submit any dispute that may arise from the application of provisions comprising the legal system of the Andean Community to any court, arbitration system or proceeding whatsoever except for those stipulated in this Treaty.’

How to Win a WTO Dispute Based on Non-WTO Law?

19

an EU member from challenging other EU members before an international tribunal other than the European Court of Justice?39 Obviously not if the dispute does not concern also ‘the interpretation or application of the [EC] Treaty’. But what if the dispute raises questions under both the EC Treaty and another treaty, say the WTO agreement or the United Nations Convention on the Law of the Sea (UNCLOS)? This tension between EC courts and other international tribunals materialised most recently in the Mox Plant case (Ireland v United Kingdom). In that dispute, Ireland submitted claims of violation under UNCLOS concerning discharges into the Irish sea of radioactive waste by a new processing plant (the so-called MOX plant) set up by the United Kingdom close to the Irish border. In an Order on Provisional Measures dated 3 December 2001, the ITLOS found that there was prima facie jurisdiction under Article 288.1 of UNCLOS.40 The Arbitral Tribunal constituted subsequently under Annex VII of UNCLOS (to decide on the merits of the case) decided, in contrast, to suspend its proceedings by Order of 24 June 2003. It did so in response mainly to arguments by the United Kingdom that the dispute falls within the exclusive jurisdiction of EC courts pursuant to Article 292 of the EC Treaty. The Arbitral Tribunal was of the view that the question of whether and what aspects of the UNCLOS dispute fall under the exclusive jurisdiction and competence of the European Communities is a question ‘to be decided within the institutions of the European Communities, and

38

39 40

See, in particular, the exclusive jurisdiction clause often included in investment or concessions contracts in favor of the domestic courts of the host state and how such clauses may play out against the compulsory jurisdiction of international arbitration tribunals granted under an investment treaty, discussed, inter alia, in the ICSID case Compania de Aguas del Aconquija S.A. & Vivendi Universal v Argentine Republic 40 ILM 426 (2001) and 41 ILM 1135 (2002), at para. 98: ‘In a case where the essential basis of a claim brought before an international tribunal is a breach of contract, the [international] tribunal will give effect to any valid choice of forum clause in the contract’. On the general question of overlapping jurisdictions, see Shany (2003); Lowe (1999); and the references supra, footnote 29. Order of 3 December 2001, at www.itlos.org, Case No. 10.

20

Joost Pauwelyn

particularly by the European Court of Justice’.41 Hence, the Arbitral Tribunal considered it inappropriate to continue its proceedings ‘in the absence of a resolution of the problems referred to’ within the context of the EC.42 Interestingly, the Order did so ‘bearing in mind considerations of mutual respect and comity which should prevail between judicial institutions both of which may be called upon to determine rights and obligations as between two States’ and noted that ‘a procedure that might result in two conflicting decisions on the same issue would not be helpful to the resolution of the dispute between the Parties’.43 Unless otherwise agreed or decided, the Arbitral Tribunal will resume its proceedings not later than 1 December 2003, in the hope that by then the dispute will be dealt with by the European Court of Justice.44 This Order to suspend UNCLOS proceedings based on provisions in another agreement (here, the EC Treaty) is in line with the approach that this author would suggest for WTO panels. WTO panels, as well, ought to take cognizance of other agreements in which the disputing parties may have taken away the jurisdiction of a WTO panel to deal with particular cases. When it comes to disputes between EU member states that raise questions under both EU law and WTO law, two problems must be distinguished. First, pursuant to Article 133 of the EC Treaty, the EC’s common commercial policy falls within the exclusive competence of the 41

42 43 44

Order No. 3 of 24 June 2003, at http://www.pca-cpa.org/PDF/ MOX%20Order%20no3.pdf, p. 8, para. 26. The European Commission actually initiated infringement procedures under the EC Treaty against Ireland claiming that Ireland’s initiation of the Mox Plant case under UNCLOS (as well as the OSPAR Convention) violates Ireland’s obligations under the EC Treaty (Ireland Threatened over Sellafield Row, The Independent, 29 June 2003). Order No. 3 of 24 June 2003, p. 9, para. 28. Ibidem. Ibidem, p. 9, para. 30. Note, in contrast, the Award under the OSPAR Convention (quoted supra in footnote 19) where jurisdiction was found, notwithstanding overlaps with EC treaties and directives. However, in that case the United Kingdom did not press its defense under EU law as hard as it did in the UNCLOS Mox Plant dispute. Yet, in my view, the OSPAR Tribunal, as well, should, like the UNCLOS Tribunal, have suspended its proceedings until further clarification was offered by EC institutions on the matter of overlap with EU law and potential exclusive competences of the EC.

How to Win a WTO Dispute Based on Non-WTO Law?

21

EC. This means that individual EU member states no longer have the competence to act in this field.45 As a result, individual EU members not even have the legal capacity to invoke WTO dispute settlement procedures46, at least not in respect of WTO matters falling within the exclusive powers of the EC (for certain GATS and TRIPS matters the competences remain shared between EU member states and the EC47). The resulting lack of capacity to bring a WTO complaint applies not only for disputes between EU members, but also for procedures initiated by an EU member against WTO members that are not members of the EU. Consequently, in case the complaining EU member thus lacks the legal capacity to submit a WTO complaint, any WTO panel ought to recognize this and decline to exercise jurisdiction, even if this lack of capacity results from EU law, not WTO law. As between EU members, the ECJ could then decide the dispute; a WTO complaint by an EU member against a non-EU member could then be re-initiated at the WTO by the EC itself (the EC being a WTO member in its own right). Second, a WTO dispute between EU members may also activate Article 229 of the EC Treaty. On the premise that the WTO provisions relied on by the complainant are a matter of exclusive EC competence, any rights that the complainant thus asserts against another EU member exist – and are a matter of – EC law.48 Conse45

46

47

48

See Case 22/70, Commission v Council (‘ERTA’): ‘... each time the Community, with a view to implementing a common policy envisaged by the Treaty, adopts provisions laying down common rules, whatever form these may take, the Member States no longer have the right, acting individually or even collectively, to undertake obligations with third countries which affect those rules’ ([1971] ECR 263, paragraph 17). Ibidem, at paragraph 18: ‘As and when such common rules come into being, the Community alone is in a position to assume and carry out contractual obligations toward third countries affecting the whole sphere of application of the Community legal system’. See Opinion 1/94, Competence of the Community to conclude international agreements concerning services and the protection of intellectual property, [1994] ECR Page I-05267. Every international agreement entered into by the EC becomes, from its entry into force, an integral part of EC law (Case 181/73, Haegeman v Belgium, [1974] ECR 449 at paragraph 5 and Opinion 1/91, [1991] ECR I-6079 at paragraph 37.

22

Joost Pauwelyn

quently, even if the complainant could frame its complaint exclusively in terms of a violation of WTO rules, the WTO dispute – in so far as it raises a matter within the exclusive competence of the EC – is necessarily also ‘a dispute concerning the interpretation or application of [the EC] Treaty’ which only the ECJ can resolve.49 The subject matter of the WTO dispute is then covered fully also under EC law; as a matter of fact, under EC law as it applies between EU members, it is then an internal question only of EU law, not one of WTO law. In that event, a conflict would arise between Article 292 of the EC Treaty, reserving exclusive jurisdiction to EC bodies to resolve the dispute, and Article 23 of the DSU, stating with equal force that ‘[w]hen [WTO] Members seek the redress of a violation of obligations ... under the covered agreements ..., they shall have recourse to ... the rules and procedures of this Understanding’ (emphasis added). The question is then which of those two norms prevails? In my view, if such genuine conflict does arise, it ought to be Article 292 of the EC Treaty that prevails as the more specific norm or lex specialis (both in terms of membership and subject matter) and arguably even as the later norm in time or lex posterior (the EC treaty was most recently re-concluded in Nice in 2001). Consequently, a WTO panel ought, in those circumstances, to decline jurisdiction. It would then be for the parties, or the European Commission, to re-initiate the dispute before the European Court of Justice. The ECJ could then examine the dispute in terms of EU law as such, or even find violations of WTO obligations given that the WTO treaty is an integral part of the EU legal system and can be relied on directly before the ECJ by the Commission or an EU member in a dispute against another EU member.50 49

50

Crucially in this respect, WTO panels, like all other international tribunals, have the incidental jurisdiction ‘to interpret the submissions of the parties’ in order to ‘isolate the real issue in the case and to identify the object of the claim’ (Nuclear Test cases, ICJ Reports 1974, 262, para. 29 and 466, para. 30 and Fisheries Jurisdiction case (Spain v Canada), ICJ Reports 1998, 437). Based on these powers, the WTO panel may find that the dispute is not only one under WTO covered agreements, but also one ‘concerning the interpretation or application of [the EC] Treaty’. This issue of the enforcement of WTO obligations by the ECJ as between two EU members (or at the request of the Commission against an EU member) – confirmed in, for example, Commission v Germany, C-61/94, Jur., 1996, I-3989, r.o. 52 – must be distinguished

How to Win a WTO Dispute Based on Non-WTO Law?

23

In most WTO cases, the existence of exclusive EC/ECJ competences will be clear so that a WTO panel can apply relevant EC law directly without much interpretation, based on unambiguous EC treaty provisions and ECJ case law. Note, in this respect, that reserving exclusive jurisdiction to EC courts for certain disputes under EC law (as Article 292 does) should not prevent other courts or tribunals – here a WTO panel – from applying relevant rules of EC law in the examination of their own jurisdiction.51 In cases where the situation under EU law is less clear (say, in respect of certain GATS and TRIPS matters for which EU members and the EC share competences), it may be wise for the WTO panel to suspend its proceedings – much like the UNCLOS Arbitral Tribunal did in the Mox Plant case – in the hope that the parties themselves sort out the EC questions before the competent EC bodies. The UNCLOS Tribunal did so based on Article 8 of its Rules of Procedure, providing that ‘[s]ubject to these Rules, the Arbitral Tribunal may conduct the arbitration in such a manner as it considers appropriate, provided that the Parties are treated with equality and that at any stage of the proceedings each Party is given a full opportunity to be heard and to present its case’. Arguably, a WTO panel, as well, is given sufficient flexibility to suspend its own proceedings, even without the agreement of the complainant52, pursuant to Article 12.1 of the DSU, stating that ‘[p]anels shall follow the Working Procedures in Appendix 3 unless the panel decides otherwise after consulting the parties to the dispute’.

51

52

from the general lack of direct effect of WTO law as it can be invoked before the ECJ by private parties or against the EC (Portugal v Council, C-149/96, 1999, ECR I-8395). In support: Decision on Annulment in Compania de Aguas del Aconquija S.A. & Vivendi Universal v Argentine Republic 41 ILM 1135 (2002), at para. 105: ‘it is one thing to exercise contractual jurisdiction (arguably exclusively vested in the administrative tribunals of Tucuman by virtue of the Concession Contract) and another to take into account the terms of a contract in determining whether there has been a breach of a distinct standard of international law, such as that reflected in Article 3 of the BIT’. The suspension of WTO panel proceedings at the request of the complainant is dealt with in Article 12.12 of the DSU.

24

Joost Pauwelyn

D. A Treaty Providing for Choice of Forum but Making any Choice Exclusive Whereas Article 292 of the EC Treaty reserves exclusive jurisdiction to EC courts, Article 1.2 of the 2002 Olivos Protocol – the most recent dispute settlement mechanism set up within MERCOSUR, replacing the earlier Brasilia Protocol – provides a choice of forum in respect of disputes that can be referred to both the WTO and MERCOSUR: Disputes falling within the scope of application of this Protocol that may also be referred to the dispute settlement system of the World Trade Organisation or other preferential trade systems that the MERCOSUR State Parties may have entered into, may be referred to one forum or the other, as decided by the requesting party. Provided, however, that the parties to the dispute may jointly agree on a forum. (emphasis added) Article 1.2 of the Olivos Protocol continues, however, as follows: ‘Once a dispute settlement procedure pursuant to the preceding paragraph has begun, none of the parties may request the use of the mechanisms established in the other fora …’ Chapter 20 of NAFTA sets out a similar regime. Where a dispute regards a matter arising under both NAFTA and the WTO, in principle, the choice of forum is left to the discretion of the complaining party53 (although for certain types of disputes, such as those related to environmental or health protection, the defendant can insist that the dispute be decided under NAFTA54). However, once a forum is chosen, it must be used to the exclusion of all others.55 53

54 55

NAFTA Article 2005, paragraph 1 (entitled ‘GATT Dispute Settlement’) reads: ‘Subject to paragraphs 2, 3 and 4, disputes regarding any matter arising under both this Agreement and the General Agreement on Tariffs and Trade, any agreement negotiated thereunder, or any successor agreement (GATT), may be settled in either forum at the discretion of the complaining Party’ (emphasis added). NAFTA Article 2005, paragraphs 3-5. NAFTA Article 2005, paragraph 6: ‘Once dispute settlement procedures have been initiated under Article 2007 or dispute settlement proceedings have been initiated under the GATT, the forum selected shall be used to the exclusion of the other unless a Party makes a re-

How to Win a WTO Dispute Based on Non-WTO Law?

25

For present purposes, the question arises what a WTO panel should do in case a WTO member first pursues its complaint under MERCOSUR or NAFTA and thereafter re-submits it to the WTO, in violation of the MERCOSUR/NAFTA exclusion provision referred to above. This type of situation arose recently before the WTO panel on Argentina – Definitive Anti-Dumping Duties on Poultry from Brazil. In that dispute, Brazil invoked WTO dispute settlement procedures, after it had unsuccessfully relied on MERCOSUR arbitration (a MERCOSUR arbitration panel had rejected Brazil’s claims of violation in respect of the very same anti-dumping measure imposed by Argentina). In that case, however, the old Brasilia Protocol was still applicable. The panel noted that this earlier Protocol ‘imposes no restrictions on Brazil’s right to bring subsequent WTO dispute settlement proceedings in respect of the same measure’. 56 However, the panel also went on to note the following: We note that Brazil signed the Protocol of Olivos in February 2002. Article 1 of the Protocol of Olivos provides that once a party decides to bring a case under either the MERCOSUR or WTO dispute settlement forums, that party may not bring a subsequent case regarding the same subject-matter in the other forum. The Protocol of Olivos, however, does not change our assessment, since that Protocol has not yet entered into force, and in any event it does not apply in respect of disputes already decided in accordance with the MERCOSUR Protocol of Brasilia. Indeed, the fact that parties to MERCOSUR saw the need to introduce the Protocol of Olivos suggests to us that they recognised that (in the absence of such Protocol) a MERCOSUR dispute settlement proceeding could be fol-

56

quest pursuant to paragraph 3 or 4’ (emphasis added). See also the so-called ‘fork in the road’ provision in many bilateral investment treaties, offering a choice to investors to either submit disputes to the domestic courts of the host state or international arbitration, but stating explicitly that once an avenue is chosen, it is to the exclusion of the other (see, for example, Compania de Aguas del Aconquija S.A. & Vivendi Universal v Argentine Republic 40 ILM 426 (2001) and 41 ILM 1135 (2002), in particular, at paras. 55, 60 and 113). Panel report on Argentina – Definitive Anti-Dumping Duties on Poultry from Brazil, WT/DS241/R, adopted on 19 May 2003 (not appealed), para. 7.38.

26

Joost Pauwelyn

lowed by a WTO dispute settlement proceeding in respect of the same measure.57 These considerations indicate a willingness on behalf of the WTO panel to apply exclusion clauses in other, non-WTO treaties. Indeed, if such non-WTO rules could not ever play a role before a WTO panel then surely the panel would not have bothered explaining and assessing their impact.

If the MERCOSUR dispute had been dealt with under the Olivos Protocol and the MERCOSUR exclusion clause would thus have been triggered, a WTO panel should, in my view, give effect to this exclusion clause. A WTO panel must examine its own jurisdiction. MERCOSUR parties agreed to take away this jurisdiction when two conditions are fulfilled: (i) the dispute is one ‘falling within the scope of application of [the Olivos] Protocol that may also be referred to the dispute settlement system of the [WTO]’; and (ii) the dispute is, or has been examined already by a MERCOSUR panel.58 Consequently, if both of these conditions are met, any WTO panel must come to the conclusion that – by agreement of the disputing parties – it does not have jurisdiction to re-examine the dispute.59 57

58

59

Ibidem. On Argentina’s claim that Brazil was estopped from bringing the same matter to the WTO after bringing it to MERCOSUR, see infra text at footnote 83. Interestingly, in this respect, the Japan-Singapore Economic Partnership Agreement (JSEPA) adds a third condition to be fulfilled before the invocation of one procedure excludes the other (not present under either NAFTA or MERCOSUR), namely the exclusion does not apply ‘if substantially separate and distinct rights or obligations under different international agreements are in dispute’ (Chapter 21, Article 139.3). This may, indeed, be a wise addition in order to avoid exclusion in case the substantive claims under both procedures are markedly different. At the same time, it adds a complexity in that it will then be for the second panel or tribunal to decide whether the procedure before it raises ‘substantially separate and distinct rights or obligations’. Along the same lines, see Compania de Aguas del Aconquija S.A. & Vivendi Universal v Argentine Republic 40 ILM 426 (2001) and 41 ILM 1135 (2002), at para. 113, referred to supra in footnote 55.

How to Win a WTO Dispute Based on Non-WTO Law?

27

E. A Treaty Providing for Compulsory (but not Exclusive) Jurisdiction to Another Tribunal Instead of reserving exclusive jurisdiction to, for example, the ECJ or offering a choice of forum between NAFTA/MERCOSUR and the WTO which once decided upon becomes exclusive, other, nonWTO treaties may also confer compulsory (though not exclusive) jurisdiction to another tribunal to resolve certain disputes with a WTO component. A dispute may, for example, raise questions under both the WTO treaty and a regional trade arrangement with compulsory (though not exclusive nor exclusionary) jurisdiction.60 It may also concern a question of maritime delimitation for which jurisdiction has been conferred to the ICJ and raise trade questions under the WTO agreement for which a WTO panel can be established (witness the Case concerning Maritime Delimitation between Nicaragua and Honduras in the Caribbean Sea (Nicaragua v Honduras),61 the trade aspect of which was brought also before a WTO panel62). A dispute can also raise questions under both UNCLOS and the WTO agreement, witness the WTO case on Chile—Measures Affecting Transit and Importation of Swordfish,63 brought also before the International Tribunal for the Law of the Sea (ITLOS).64 Should the fact that a dispute, or part of a dispute, can be brought also to compulsory dispute settlement procedures under another agreement, prevent a WTO panel from examining the WTO claims before it? In my view, not necessarily so. Rather, in most cases it will be possible to split up the WTO component of the dis60

61 62

63 64

See, for example, the dispute settlement mechanism under the Protocol on Trade of the Southern African Development Community (SADC), discussed in Pauwelyn (2004). See the records of this ongoing case at http://www.icj-cij.org/ icjwww/idocket.htm. Nicaragua – Measures Affecting Imports from Honduras and Colombia, WTO documents WT/DS188/2 and WT/DS202/1 (although a panel was established on this matter, its proceedings have been suspended by mutual agreement). WTO document WT/DS193 (suspended by mutual agreement on March 23, 2001). Conservation and Sustainable Exploitation of Swordfish Stocks in the South-Eastern Pacific Ocean (Chile v Eur. Com.) (Mar. 15, 2001) at http://www.un.org/Depts/los/ITLOS/Order1_2001Eng.pdf (currently suspended on the basis of a provisional arrangement).

28

Joost Pauwelyn

pute from the ICJ/UNCLOS component of the dispute so that the former be decided by a WTO panel and the latter by the ICJ/ITLOS. Examples of cases where the court or tribunal separated the different aspects of a dispute, finding that although it had no jurisdiction to look at one aspect, it continued to have jurisdiction to examine another, are: (i) the Nicaragua case where the ICJ declared that it did have jurisdiction over certain claims under customary international law even though the United States had not accepted ICJ jurisdiction in respect of ‘disputes arising under a multilateral treaty, unless … all parties to the treaty affected by the decision are also parties to the case before the Court’ and the relevant multilateral treaty rules largely overlapped with the customary law,65 (ii) the ICSID Award in Compania de Aguas del Aconquija S.A. & Vivendi Universal v Argentine Republic where the Tribunal accepted jurisdiction over certain claims under a bilateral investment treaty between France and Argentina, even though these claims overlapped with claims under a concession contract for which exclusive jurisdiction had been reserved to domestic Argentine courts,66 and (iii) Order No. 3 of the Arbitral Tribunal constituted under Annex VII of UNCLOS in the Mox Plant case (Ireland v United Kingdom), finding that a concurrent proceeding before an OSPAR67 tribunal limited to claims under the OSPAR convention in relation to the same MOX plant, did not prevent the UNCLOS tribunal from having jurisdiction over the UNCLOS claims in dispute.68 65 66 67 68

Jurisdiction and Admissibility, ICJ Reports 1984, para. 73 and Merits, ICJ Reports 1986, para. 175. 40 ILM 426 (2001), at paras. 53-4. OSPAR refers to the 1992 Convention for the Protection of the Marine Environment of the North-East Atlantic (OSPAR Convention). The Tribunal explained its approach as follows: ‘[T]he Tribunal does not consider that this [relevance of the OSPAR Convention] alters the character of the dispute as one essentially involving the interpretation and application of the [UNCLOS] Convention. Furthermore, the Tribunal is not persuaded that the OSPAR Convention substantially covers the field of the present dispute’ (Order No. 3 of 24 June 2003, at http://www.pca-cpa.org/PDF/MOX%20Order%20no3.pdf,

How to Win a WTO Dispute Based on Non-WTO Law?

29

Such splitting-up or ‘salami-slicing’ of the dispute, though possible in most cases, may be unwarranted in others. First, in exceptional circumstances, the WTO dispute may be so similar to the ICJ/UNCLOS dispute, both in terms of subject matter and substance and scope of the applicable rules, that the two disputes are, in effect, but one and the same.69 A conflict may then arise between, on the one hand, the rule conferring jurisdiction over the dispute to the ICJ/UNCLOS and, on the other hand, the DSU conferring jurisdiction over substantially the same dispute to the WTO. In the absence of explicit conflict clauses in either treaty70, the resolution of such conflict should then depend on normal conflict rules, in particular the principles of lex posterior and lex specialis. If, but only if, based on those rules, the DSU must give way to the other, non-WTO provision, then the WTO panel should find that it has no jurisdiction to examine the dispute. A second reason not to split the dispute into a WTO and an ICJ/UNCLOS component may be that the dispute not genuinely concerns WTO claims (even though such claims can technically be made) but, rather, other rules of international law that the WTO

69

70

6, para. 18). Subsequently, the OSPAR Tribunal issued its own Final Award, notwithstanding the concurrent UNCLOS proceeding, see supra footnote 19. This situation was acknowledged, for example, in the Decision on Annulment in Compania de Aguas del Aconquija S.A. & Vivendi Universal v Argentine Republic. Faced with a situation that gave rise to claims under both a bilateral investment treaty (for which ICSID had jurisdiction) and a domestic concession contract (for which domestic courts had exclusive jurisdiction), the Committee found as follows, at paras. 98 and 101: ‘In a case where the essential basis of a claim brought before an international tribunal is a breach of contract, the [international] tribunal will give effect to any valid choice of forum clause in the contract. … On the other hand, where ‘the fundamental basis of the claim’ is a treaty laying down an independent standard by which the conduct of the parties is to be judged, the existence of an exclusive jurisdiction clause in a contract between the claimant and the respondent state … cannot operate as a bar to the application of the treaty standard’. UNCLOS Art. 282 does, for example, provide for a conflict rule, making UNCLOS jurisdiction subject to certain dispute settlement procedures under other treaties.

30

Joost Pauwelyn

claims are inextricably linked to and that these WTO claims are dependent on to be decided.71 In such extreme cases it could then be submitted that the history, prior procedures, and substantive content of the dispute indicate that the real issue of the case (i.e., the genuine object of the claim) is related to non-WTO claims as to which a WTO panel does not have jurisdiction.72 On these grounds, the WTO panel could then decide that it does not have substantive ju71

72

Imagine that the complainant makes a non-violation complaint under GATT Art. XXIII:1(b) arguing that the defendant has nullified its tariff concessions on, for example, the import of footballs, by suddenly no longer complying with the ILO prohibition on child labor in its domestic production of footballs (making it harder for the complainant to compete). Can the complainant rely on these non-WTO rules even if this would imply that the WTO panel would first have to find a violation of ILO obligations before it could accept the complainant’s non-violation case under WTO rules? Here, the WTO complaint could be said to no longer concern WTO claims but rather ILO claims so that the panel could find that it has no jurisdiction to hear the case. See supra footnote 49 on the power of international tribunals to ‘redefine’ a dispute. It was on these grounds that the ICSID Tribunal in Compania de Aguas del Aconquija S.A. & Vivendi Universal v Argentine Republic (40 ILM 426 (2001), at p. 3 of the Award), though having accepted jurisdiction over Vivendi’s claims, refused to make findings on the merits: ‘the nature of the facts supporting most of the claims presented in this case make it impossible for the Tribunal to distinguish or separate violations of the BIT from breaches of the Concession Contract without first interpreting and applying the detailed provisions of that agreement. By Article 16.4, the parties to the Concession Contract assigned that task expressly and exclusively to the contentious administrative courts of Tucumán. Accordingly, and because the claims in this case arise almost exclusively from alleged acts of the Province of Tucumán that relate directly to its performance under the Concession Contract, the Tribunal holds that the Claimants had a duty to pursue their rights with respect to such claims against Tucumán in the contentious administrative courts of Tucumán as required by Article 16.4 of their Concession Contract’. This Tribunal finding was, however, subsequently (and, based on the facts, in my view correctly) overruled by the Annulment Committee (41 ILM 1135 (2002), at paras. 115 and 105, quoted supra in footnote 51).

How to Win a WTO Dispute Based on Non-WTO Law?

31

risdiction over the dispute based on the compulsory jurisdiction conferred to another tribunal.73 The following are examples of cases where it was found that a dispute cannot be split in two and, as a result, the tribunal declined to exercise jurisdiction: (i) the ICJ Fisheries Jurisdiction case (Spain v Canada), where the ICJ ‘re-defined’ Spain’s complaint relating to Canada’s ‘lack of entitlement to exercise jurisdiction on the high seas’ into a dispute ‘arising out of or concerning conservation and management measures’ for which Canada had made a reservation to its grant of jurisdiction to the ICJ. On that basis, the Court found that it did not have jurisdiction to hear the case;74 and (ii) the UNCLOS Arbitration Award on Southern Bluefin Tuna. In the latter case, the tribunal found that the dispute ‘while centered in the 1993 [trilateral Convention for the Conservation of Southern Bluefin Tuna], also arises under [UNCLOS]’. It continued, nonetheless, by saying that ‘[t]o find that, in this case, there is a dispute actually arising under UNCLOS which is distinct from the dispute that arose under the [1993 Convention] would be artificial’.75 Since the tribunal later declared not to have jurisdiction over the 1993 Convention part of the dispute, it automatically declined jurisdiction also over the UNCLOS part (notwithstanding the compulsory jurisdiction in Part XV of UNCLOS) on the ground of its ‘single dispute’ theory.

73

74 75

Advocating that WTO panels decline jurisdiction in certain cases where non-WTO rules are at stake, see Marceau (2001). Note the difference between this second reason to reject jurisdiction (no jurisdiction to begin with, since no WTO claims at issue) and the earlier, first reason to reject jurisdiction (a conflict between two rules conferring jurisdiction to different tribunals decided, as the case may be, in favor of the non-WTO tribunal). ICJ Reports 1998, 437. Southern Bluefin Tuna case (Australia and New Zealand v Japan, Jurisdiction and Admissibility), Arbitral Tribunal constituted under Annex VIII of UNCLOS, posted on the internet at www.worldbank. org/icsid/bluefintuna/main.htm, paras. 52 and 54.

32

Joost Pauwelyn

F. Res judicata Effect of Rulings by Other Courts or Tribunals Finally, a WTO panel may have to decline jurisdiction based on an earlier ruling by another court or tribunal on the same matter (say, based on a panel under the Southern African Development Community (SADC) which does not have an exclusion clause similar to the one in NAFTA or the Olivos Protocol discussed earlier76). As a result, the defendant can win a WTO dispute, not so much with reference to another, non-WTO treaty, but based on the so-called res judicata effect of a judgment or ruling by another court or tribunal. There are, however, three conditions for the principle of res judicata to apply. They are: (i) identity of parties; (ii) identity of object or subject matter (it must be the very same issue that is in question); and (iii) identity of the legal cause of action.77 It is undisputed that WTO panel and Appellate Body reports, once adopted by the DSB, have binding legal effect as between the parties to the particular dispute. The Appellate Body recently confirmed the above three conditions in respect of WTO panel reports when it stated as follows: ‘… in our view, an unappealed finding included in a panel report that is adopted by the DSB must be treated as a final resolution to a dispute between the parties in respect of the particular claim and the specific component of a measure that is the subject of that claim’.78 The Appellate Body also confirmed the principle of res judicata in respect of its own reports.79

76 77 78

79

See Pauwelyn (2004). See Lowe (1996), 40. Appellate Body report on EC – Anti-Dumping Duties on imports of Cotton-Type Bed Linen from India, Recourse to Article 21.5 of the DSU by India, WT/DS141/AB/RW, 8 April 2003, para. 93. For a discussion of the res judicata effect of WTO panel reports, see also the Panel Report on India – Autos, supra footnote 33. Appellate Body report on United States – Import Prohibition of Certain Shrimp and Shrimp Products, Recourse to Article 21.5 of the DSU by Malaysia WT/DS58/AB/RW, 22 October 2001, paras. 9296.

How to Win a WTO Dispute Based on Non-WTO Law?

33

Obviously, for a WTO panel to give res judicata effect to another WTO panel or Appellate Body report is one thing; to give the same effect to a ruling or report by another court or tribunal (say, a SADC panel), is quite another. Two hurdles must be passed in this respect. First, WTO panels must recognize that res judicata is a principle of general international law that WTO panels must apply irrespective of whether the earlier ruling in question comes from within or outside the WTO. In my view, this hurdle is easy to pass: the applicable law before WTO panels includes general principles of law and it is widely accepted that res judicata is such a principle.80 Second, the ruling or report by the other court or tribunal must meet the three conditions referred to earlier. In other words, the parties, subject matter and legal cause of action before the other court or tribunal must be the same as those before the WTO panel. Even if another tribunal may have dealt with the same subject matter as between the same parties (say, a SADC panel may have decided a safeguards dispute between South Africa and Mozambique), it is unlikely that it will have examined the matter under the same cause of action, that is, as a question of violation of WTO rules: the SADC panel will have examined claims of violation of SADC safeguard rules; the WTO panel will be asked to examine claims of violation of the WTO safeguards agreement. Does this mean that WTO panels will never have to give res judicata effect to other courts or tribunals? Not necessarily so. In case both the SADC and the WTO provision under which the respective claims are made, are in substance the same (say, both raise a violation of the MFN principle in respect of safeguard measures), the argument could be made that the doctrine of ‘issue estoppel’ or ‘collateral estoppel’ applies. The English law requirements for issue estoppel have been explained as follows: ‘(1) that the same question has been decided; (2) that the judicial decision which is said to create the estoppel is final; and, (3) that the parties to the judicial decision or their privies were the same persons as the parties to the proceedings in which the estoppel is raised or their privies’.81

80 81

See Lowe, Res Judicata, supra footnote 77. Carl Zeiss Stiftung v Rayner & Keeler Ltd. (No. 2) [1967] A.C. 853 at 935 (per Lord Guest).

34

Joost Pauwelyn

In other words, in English law, the requirements for issue estoppel are the same as those for the traditional res judicata principle to apply, minus the requirement of identity of legal cause of action. In US law, a similar doctrine is known as collateral estoppel: ‘Collateral estoppel extends the res judicata effect of a judgment to encompass the same issues arising in a different action (‘issue preclusion’) and even to different parties where the issue has been determined in prior litigation with adequate opportunity to be heard for the party to be precluded’.82 As a result, if the principle of issue estoppel were applied also before a WTO panel, then a WTO panel could preclude a SADC member from bringing, for example, a safeguards claim at the WTO in case substantively the same claim was previously decided upon by a SADC panel, as between the same parties. A WTO panel could also apply the principle of issue estoppel to the determination of specific facts or the legal characterization of facts by the previous SADC panel (or vice versa). The US doctrine of collateral estoppel could even go further and give res judicata effect also to a previous SADC panel finding on the same issue even if that panel was constituted at the request of another SADC member, different from the one now challenging the same measure at the WTO, in the event the former SADC member had ‘adequate opportunity to be heard’ before the original SADC panel. Note, however, that in a recent panel report, the panel refused to apply the basic principle of estoppel in respect of a claim by the defendant Argentina that the same measure had earlier been decided in MERCOSUR arbitration. In Argentina – Definitive Anti-Dumping Duties on Poultry from Brazil, Argentina argued, indeed, that the fact that Brazil had challenged the same measure previously before MERCOSUR estopped Brazil from bringing the case again before the WTO.83 Argentina explicitly refused, however, to invoke the principle of res judicata.84 Although the panel refused to decide on whether or not the principle of estoppel can apply before a WTO 82 83

84

Scoles et al. (2000), 1141. Panel report on Argentina – Definitive Anti-Dumping Duties on Poultry from Brazil, WT/DS241/R, adopted on 13 May 2003 (not appealed), paras. 7.37 et seqq. Ibidem, fn. 53.

How to Win a WTO Dispute Based on Non-WTO Law?

35

panel85, it used the following three conditions for estoppel to be activated: ‘(i) a statement of fact which is clear and unambiguous, and which (ii) is voluntary, unconditional, and authorized, is (iii) relied on in good faith’. When applying the first condition, the panel did not consider ‘that Brazil has made a clear and unambiguous statement to the effect that, having brought a case under the MERCOSUR dispute settlement framework, it would not subsequently resort to WTO dispute settlement proceedings. … This is especially because the Protocol of Brasilia [unlike the Olivos Protocol discussed earlier], under which previous MERCOSUR cases had been brought by Brazil, imposes no restrictions on Brazil’s right to bring subsequent WTO dispute settlement proceedings in respect of the same measure’.86 On these grounds, the panel refused to apply the principle of estoppel and continued to examine Brazil’s claims. It is worth noting, however, that rather than applying WTOlike anti-dumping rules, the earlier MERCOSUR arbitration panel found that no specific anti-dumping rules applied to the contested measure.87 As a result, it was, indeed, hard to say that the earlier MERCOSUR ruling dealt with substantively similar claims as those raised before the subsequent WTO panel. Hence, even under the wider notions of ‘issue estoppel’ or ‘collateral estoppel’, Brazil should not have been precluded from submitting the WTO case. IV. Because of Non-WTO Law, a WTO Violation is Justified In Section II, we examined instances where non-WTO law may lead a panel to decline jurisdiction. In this Section, we assess how non-WTO law may, on the merits, prevent a finding of violation of WTO rules. As mentioned in the introduction, we are concerned not with cases where non-WTO law may simply influence the interpretation of WTO terms, be it by expanding or limiting the scope of

85 86 87

Ibidem, fn. 58. Ibidem, para. 7.38. Award by the Ad Hoc MERCOSUR Tribunal (Brazil v Argentina) concerning the Application of anti-dumping Measures on the Export of Poultry from Brazil, 21 May 2001, at http://www.mercosur.org.uy/ pagina1esp.htm.

36

Joost Pauwelyn

WTO rules or exceptions.88 Rather, we focus on the exceptional case where a harmonious interpretation of WTO rules and other rules is not possible and the non-WTO rule offers an independent defense for what would otherwise be a violation of the WTO treaty. When it comes to the substantive evaluation of WTO claims, two DSU provisions imply the power of WTO panels to apply and make findings under other rules of international law (in addition to the other reasons stated in Section I in support of a broad notion of applicable law). First, the obligation in Article 11 of the DSU for panels to make an ‘objective assessment of ... the applicability of ... relevant covered agreements’ may require a panel to refer to and apply other rules of international law; these other rules may show that the relevant WTO rules do not apply and have therefore not been violated; in contrast, failure to look at these other rules would preclude an ‘objective assessment of ... the applicability of . . . the relevant covered agreements’. Second, the panel’s mandate in Article 7.2 of the DSU to ‘make such findings as will assist the DSB in making the recommendations or in giving the rulings provided for in that/those agreement(s)’, further acknowledges that WTO panels may need to resort to and apply rules of international law beyond WTO covered agreements as long as it assists the DSB in resolving the WTO claims before it. It may be useful to distinguish four types of situations where a non-WTO defense can be raised on the merits of a WTO claim, of increasing order of complexity (in that each time it becomes more difficult to justify why a WTO panel should accept the defense): (1) defenses under non-WTO law explicitly incorporated into the WTO legal system; (2) measures allegedly violating the WTO treaty but specifically permitted (or even imposed) pursuant to the dispute settlement provisions of another treaty; (3) measures that a WTO member must enact (or is explicitly permitted to enact) pursuant to the provisions of another treaty; (4) measures normally in breach of WTO rules but permitted under another treaty on condition that the WTO panel finds that this other treaty is respected/violated. A. Non-WTO Defenses Explicitly Incorporated into the WTO Legal System Several avenues exist where WTO law itself explicitly refers to or incorporates defenses under non-WTO rules. First, WTO treaty 88

See supra footnotes 1 and 2.

How to Win a WTO Dispute Based on Non-WTO Law?

37

provisions sometimes incorporate non-WTO defenses: (i) the SPS and TBT agreements refer to certain international standards adopted outside the WTO (such as Codex Alimentarius standards on hormones89); if the measure challenged conforms to such a standard, the measure will be presumed to be consistent also with WTO rules; (ii) the TRIPS agreement not only incorporates obligations under other, WIPO conventions; it also incorporates certain exceptions to be found in these conventions;90 (iii) under the Subsidies agreement an export credit practice is not considered a prohibited export subsidy if it is in conformity with the interest rate provisions of the OECD Arrangement on Guidelines for Officially Supported Export Credits;91 (iv) Article XXI of GATT and Article XIVbis of GATS explicitly permit ‘any action in pursuance of [a WTO member’s] obligations under the United Nations Charter for the maintenance of international peace and security’; if, therefore, the UN Security Council, acting under Chapter VII of the UN Charter, imposes economic sanctions on a WTO member, the relevant Security Council resolution – an instrument not part of WTO covered agreements – can be used in defense of a WTO complaint that the target of the sanctions would submit before a WTO panel. Second, waivers adopted by the WTO Ministerial Conference (or General Council) pursuant to Article IX:3 of the Marrakesh Agreement may authorize a WTO member to adopt otherwise WTO inconsistent measures with reference to non-WTO rules. The waiver granted to the EC in respect of its import regime for bananas, for example, refers to the Lome Convention and permits the EC ‘to provide preferential treatment for products originating in ACP States as required by the relevant provisions of the Fourth Lome Convention’.92 The more recent waiver on conflict diamonds per89

90 91

92

See, for example, Appellate Body Report, EC Measures Concerning Meat and Meat Products (Hormones), WT/DS26/AB/R, WT/DS48/AB/R, adopted 13 February 1998, DSR 1998:I, 135. See, for example, Panel Report, United States – Section 110(5) of the US Copyright Act, WT/DS160/R, adopted 27 July 2000 (not appealed). See, for example, Panel Report, Canada – Export Credits and Loan Guarantees for Regional Aircraft, WT/DS222/R and Corr.1, adopted 19 February 2002 (not appealed). The Fourth ACP-EEC Convention of Lome, Decision of the CONTRACTING PARTIES of 9 December 1994, L/7604, 19 December 1994.

38

Joost Pauwelyn

mits measures ‘necessary to prohibit the export [and import] of rough diamonds to [and from] non-Participants in the Kimberley Process Certification Scheme consistent with the Kimberley Process Certification Scheme’.93 Neither the Lome Convention nor the Kimberley scheme (or the WTO waiver, for that matter) are WTO covered agreements. Still, they can, and have been,94 referred to, interpreted and applied by a WTO panel in defense of an alleged violation of WTO rules. When non-WTO rules are explicitly incorporated into the WTO legal system, WTO panels and the Appellate Body will feel rather comfortable to apply those rules (even if, in the case of waivers, the link to those rules – that is, the waiver decision itself – is, strictly speaking, not part of a WTO covered agreement).95 Yet, the basis for referring to those rules is the same as that for referring to other rules not explicitly incorporated: the consent of the WTO members involved to accept what could otherwise be violations of the WTO treaty. Whether this consent was given within or outside the WTO should not matter. To the contrary, quite often the consent given inside the WTO will be less explicit and direct than that given outside the WTO: for example, the international standards referred to in the SPS and TBT agreements include those objected to by WTO members in the relevant standard-making bodies;96 the reference to the OECD arrangement is an ongoing one which permits OECD members to create safe-havens never accepted by WTO members not party to the OECD;97 the WTO waiver on conflict diamonds exempts trade restrictions consistent with the Kimberley scheme that are imposed on non-participants to this scheme, that is,

93

94

95 96 97

Waiver decision, WTO document G/C/W/432/Rev. 1, dated 24 February 2003, referred to in the Decision by the WTO General Council at its meeting on May 15-16, 2003 (WTO document WT/GC/W/498, dated 13 May 2003, Item VI), adopted by consensus. See also Pauwelyn (2003b). See the Appellate Body Report, European Communities – Regime for the Importation, Sale and Distribution of Bananas (‘EC – Bananas III’), WT/DS27/AB/R, adopted 25 September 1997, DSR 1997:II, 591. Ibidem. See Appellate Body Report on EC – Hormones, supra footnote 89. See supra footnote 91.

How to Win a WTO Dispute Based on Non-WTO Law?

39

WTO members not having any influence on how this scheme develops.98 B. WTO Violations Permitted (or Even Imposed) Pursuant to the Dispute Settlement Provisions of Another Treaty As explained in Section I, when WTO panels examine the validity of WTO claims, they should not limit themselves to WTO covered agreements, nor to non-WTO rules explicitly referred to in the WTO treaty or WTO waiver decisions. The easiest situation for panels to thus apply non-WTO rules would be where a measure is inconsistent with WTO rules, but specifically imposed or permitted by a decision under the dispute settlement mechanism of another treaty. A case in point is the Resolution of the International Labor Conference recommending action against Myanmar for grave breaches of the ILO’s Forced Labor Convention.99 There, the conclusion that ILO rules had been breached was made by an independent Commission of Inquiry set up under the ILO Constitution. The subsequent ILO recommendation was the very first case where sanctions were called for under Article 33 of the ILO Constitution since the ILO’s creation in 1919. A number of WTO members have imposed trade embargoes on Myanmar since that ILO resolution.100 If Myanmar – a member of the WTO – were now to challenge those embargoes, could the WTO member imposing the embargo not rely on the ILO recommendation so as to justify any potential WTO violations? In my view, this should be possible: Myanmar, as all ILO members, agreed to the ILO dispute settlement mechanism; See, supra footnote 93. The ILO recommended that ILO members ‘review, in the light of the conclusions of the Commission of Inquiry [which had found the serious violations of the Forced Labor Convention], the relations that they may have with the member State concerned [Myanmar] and take appropriate measures to ensure that the said Member cannot take advantage of such relations to perpetuate or extend the system of forced or compulsory labor referred to by the Commission of Inquiry, and to contribute as far as possible to the implementation of its recommendations made’, Resolution of the International Labour Conference (88th session, 2000) at http://www.ilo.org/public/english/standards/ relm/ilc/ilc88/resolutions.htm#II. 100 In the United States, see the Burmese Freedom and Democracy Act 2003, H.R. 2330, 108th Congress (2003).

98 99

40

Joost Pauwelyn

this mechanism found that Myanmar had breached ILO rules and recommended action against it; although it is hard to say that other ILO members where thereby obliged to impose a trade embargo on Myanmar, in case those other members are ‘good ILO citizens’ and do what the ILO is explicitly calling for, then who is the WTO to question the validity of ILO decisions? In that event, a conflict may arise between WTO rules (prohibiting the embargo101) and the ILO recommendation (calling for an embargo). As the later and more specific norm, the ILO recommendation should then prevail over the WTO prohibition. A similar situation could arise when two WTO members agree in an economic cooperation agreement that in case either violates fundamental human rights, the other may impose trade sanctions.102 If such violation of human rights were now impartially determined either by a human rights court or a tribunal or commission set up under the cooperation agreement and the other member imposes trade sanctions, can the perpetrator of the human rights violation go to the WTO and obtain a ruling that the trade sanction is illegal? In my view, not. The WTO panel should then apply not just WTO covered agreements, but also take cognizance of the ruling where breach of human rights was found and of the provision explicitly permitting trade sanctions in such event. Turning the tables around, for a WTO panel to thus recognize the outcome of dispute settlement procedures under other treaties would be very similar to a situation where the World Intellectual Property Organization (WIPO) refuses to find a violation of the Berne Convention when such violation was explicitly authorized by the WTO as a countermeasure under Article 22.6 of the DSU. A case in point is the DSB authorization obtained by Ecuador to suspend the protection of EC copyrights under WIPO conventions in response to prior WTO violations by the EC’s import regime for bananas. In contrast, for a WTO panel not to take account of ILO recommendations would be the same as WIPO finding a violation under its conventions notwithstanding an explicit WTO authorization to engage in those violations.

101 Even if the case can be made also that the embargo is justified already under GATT itself (e.g. under GATT Arts. XX and/or XXI). 102 The EC, for example, often includes such provisions in its agreements with third states, see Bartels (2002).

How to Win a WTO Dispute Based on Non-WTO Law?

41

C. WTO Violations Permitted (or Even Imposed) Pursuant to the Provisions of Another Treaty When another court, tribunal or commission has previously found that a trade measure is justified under another treaty, it is relatively easy for a WTO panel to give effect to such ruling: the justification under non-WTO law is then decided upon not by the WTO panel, but by the actors in charge of the other treaty. The WTO panel then only gives effect to a ruling very specific to the parties and subject matter in dispute, which is simply the result of a procedure explicitly agreed upon by both parties in another international organization or under another treaty. The slightly more difficult situation may arise also where a measure is prohibited under WTO rules, but prescribed or explicitly permitted under the self-standing provisions of another treaty (or even customary international law103). In that event, it would be for the WTO panel itself to interpret the non-WTO treaty provision and to decide for itself whether this provision justifies the measure in question. Examples of such situation are trade measures that violate WTO rules but which a WTO member must take (or has an explicit right to take) under a multilateral environmental or health agreement that is binding on both disputing parties – such as the Cartagena Protocol on Biosafety or the recent WHO Framework Convention on Tobacco Control104 – or even under a bilateral agreement in which two WTO members agreed on the imposition as between them of otherwise WTO-inconsistent measures.105 Although in most cases measures under these other conventions will be justified also under the WTO exceptions related to health and the environment, in exceptional situations a genuine conflict may 103 See, however, the caveat in the text below at footnote 108. 104 See supra footnotes 6 and 7. 105 Imagine that the United States and Malaysia had settled the Shrimp/turtle dispute and agreed, among other things, that the United States can continue the imposition, as against Malaysia, of otherwise WTO inconsistent measures. Should a WTO panel, subsequently constituted at the request of Malaysia to strike down these very U.S. measures, not take account also of this bilateral settlement and on that basis decline to find a WTO violation? In my view, it should. This position implies the qualification of WTO obligations as essentially bilateral obligations, see Pauwelyn (2003c).

42

Joost Pauwelyn

arise.106 A WTO panel should examine whether this is, indeed, the case and may thereby have to interpret the provisions of the other treaty. If it finds a conflict, it should then decide also which of the two norms prevails under conflict rules of international law. If it is the WTO norm, then it should find a violation of WTO covered agreements. If it is the other norm, then the panel should accept the justification under non-WTO law and not find a WTO violation. Already now, WTO panels regularly interpret non-WTO rules. Although in most (though not all) of those cases there is an explicit reference to these rules in WTO provisions themselves, panels and the Appellate Body have shown that they have the qualifications and impartiality to properly interpret ‘foreign’ rules.107 Interpreting and applying non-WTO rules not explicitly referred to in the WTO treaty would, indeed, take it a step further but not inherently change a function that WTO panels already exercise at the present day. One particular reference to non-WTO law may, however, cause additional concerns, namely the application by WTO panels of rules of general international customary law. While other, nonWTO treaties agreed upon by the disputing parties are the product of deliberative and explicit state consent, international custom derives from ‘a general practice accepted as law’.108 As a result, when it comes to defining and interpreting custom, WTO panels have less explicit guidance and may feel inhibited, in particular, to decide on whether WTO treaty provisions have been altered by an allegedly supervening custom. In EC – Hormones, for example, the Appellate Body was extremely hesitant when addressing the EC claim that the precautionary principle as a rule of customary law ought to supple-

106 On the definition of ‘conflict’, see Pauwelyn (2003a). 107 See the panel and/or Appellate Body interpretations of (i) the Lome Convention in EC – Bananas III, supra footnote 95; (ii) the Berne Convention in US – Copyright, supra footnote 90; (iii) international Codex Alimentarius standards in EC – Hormones, supra footnote 89 and especially in European Communities – Trade Description of Sardines, WT/DS231/R and AB/R, adopted on 23 October 2002; (v) the OECD arrangement on export credits, in Canada – Export Credits, supra footnote 91, (v) environmental conventions and declarations in US – Shrimp, supra footnote 1; and (vi) the precautionary principle in EC – Hormones, supra footnote 89. 108 Article 38.1(b) of the Statute of the International Court of Justice.

How to Win a WTO Dispute Based on Non-WTO Law?

43

ment the provisions of the SPS Agreement.109 This trepidation is justified and WTO panels ought, indeed, be extremely careful and on solid grounds before concluding that a new rule of custom has emerged. At the same time, the risk of new custom overruling prior WTO provisions is extremely limited. Although it is generally accepted that no inherent hierarchy exists between treaties and custom110, in practice, it is rare for custom to, first of all, emerge notwithstanding the continuing existence of a contradictory treaty norm: it is not as if custom can be established over night; custom requires a general and consistent practice of states, including the (at least) tacit consent of those who concluded the pre-existing treaty (persistent objectors cannot be bound by it). Moreover, even if new custom does emerge in the face of a treaty dealing with the same subject matter, given the often vague and general nature of custom, a genuine conflict between custom and treaty is exceptional: in most cases it will be possible to interpret the treaty in line with the new custom. Finally, in those cases where a genuine conflict does arise, the treaty is most likely to prevail as lex specialis based on its 109 Appellate Body Report, EC Measures Concerning Meat and Meat Products (Hormones) (‘EC – Hormones’), WT/DS26/AB/R, WT/DS48/AB/R, adopted 13 February 1998, DSR 1998:I, 135, at para. 123: ‘The precautionary principle is regarded by some as having crystallized into a general principle of customary international environmental law. Whether it has been widely accepted by Members as a principle of general or customary international law appears less than clear. We consider, however, that it is unnecessary, and probably imprudent, for the Appellate Body in this appeal to take a position on this important, but abstract, question’. For a critique, see Pauwelyn (2001), 569-570, and Pauwelyn (2003a), 481 et seq. 110 See, for example, Kontou (1995) and the references in Pauwelyn, (2003), 94-97. Contra: McGinnis (2003), arguing that ‘[g]lobal multilateral agreements should dominate customary international law because they rest on a more certain consensus and have fewer agency costs’ (42) and even ‘skeptical that [the] substantive aspects [of custom that is part of jus cogens] should have priority over global multilateral treaties’ (note 137). While this author has previously argued (see footnotes 109 and 111) and hence agrees with Professor McGinnis – that, in practice, and in the particular context of the WTO, it will be extremely rare for subsequent custom to overrule WTO treaty provisions, this should not, however, mean that the starting principle of absence of hierarchy between treaty and custom in the wider field of public international law is no longer valid.

44

Joost Pauwelyn

often more specific and explicit expression of state will.111 In sum, the fear expressed by some authors112 that for WTO panels to apply custom risks high-jacking the contractual, consent-based nature of the WTO is unwarranted: WTO members can only be held to custom if the strict rules for its emergence are met (states who explicitly and consistently objected to the custom cannot be bound by it); moreover, even if custom was explicitly or tacitly consented to, it is unlikely to prevail over the WTO treaty. D. WTO Violations Permitted under Another Treaty on Condition that the WTO Panel Finds that this Other Treaty is Respected/Violated Undoubtedly the most controversial use of non-WTO law would be for a panel to accept a defense under another treaty even if that defense – say, an explicit right to impose a trade restriction – is conditional on prior breach of the other treaty by the complaining party (or conditional on simultaneous compliance of the other treaty by the defendant itself). Recall the example of an economic cooperation agreement where two Members agree that in case either of them violates fundamental human rights, the other may impose trade sanctions.113 Now, if such prior breach of human rights has not been independently determined by a court, tribunal or other competent commission, can the WTO member who is of the view that its partner has, indeed, committed such breach, unilaterally decide that this breach occurred and, consequently, impose trade sanctions? If so, what happens if the victim of the sanctions challenges them before a WTO panel? Can the member who enacted the sanctions justify its conduct based on the cooperation agreement? A similar situation would arise in case a participant in the Kimberley scheme against conflict diamonds is of the view that another participant to that scheme does not abide by the restrictions imposed on conflict diamonds. As a result, that member could impose import restrictions on diamonds coming from this other participant (the Kimberley scheme only permits trade in certified conflict-free diamonds and the member concerned is convinced that diamonds from this other 111 For a full discussion, see Pauwelyn (2003a), 131-143. 112 See, in particular, McGinnis (2003), discussed also supra fn. 23 and 110. 113 See supra footnote 102.

How to Win a WTO Dispute Based on Non-WTO Law?

45

participant are, in fact, conflict diamonds). Now, if the target of those trade restrictions were to challenge them before a WTO panel, could the WTO member imposing the restrictions rely on the Kimberley scheme in defense of what it is doing?114 A slightly different yet similar situation would arise also when two WTO members conclude a bilateral settlement of a WTO dispute in which they agree that the defendant can maintain a WTO inconsistent measure for a certain period of time for as long as it conforms to the other provisions in the settlement.115 If the complainant were, thereafter, to challenge this WTO inconsistent measure, notwithstanding the agreement that it can be maintained, should the defendant be allowed to rely on the bilateral settlement in its defense, even if the validity of this defense depends on whether the defendant itself has implemented the other provisions of the settlement? Although, in all three examples, the applicable law should, in my view, include also defenses under the cooperation agreement, the Kimberley scheme or the bilateral settlement (all of which were, after all, accepted by the complainant), any WTO panel would then be faced with an additional hurdle: for the defense to be valid, either the parties ought to agree that human rights / Kimberley certification requirements had, indeed, been breached or the WTO panel itself must decide that such breach has, indeed, occurred (in respect of the bilateral settlement, parties would have to agree, or the panel itself would have to find, that the defendant had otherwise complied with the settlement). Put differently, the WTO panel would then not simply apply other rules ‘as it finds them’ in other treaties, with minimal interpretation by the panel, but actually have to decide first whether or not these other rules have been breached/complied with. As noted several times, however, WTO panels have jurisdiction only to examine claims of violation under WTO covered agreements. In the examination of those WTO claims – and even accepting the argument made here that the applicable law to conduct this examination can include also other rules of international law – 114 Recall, in this respect, that the WTO waiver on conflict diamonds can be invoked only for trade restrictions on non-participants; it does not operate in the WTO relationship as between two participants to the scheme. See supra text at footnote 98. 115 See, for example, the bilateral settlement in India – Quantitative Restrictions, supra footnote 32.

46

Joost Pauwelyn

are WTO panels permitted to go as far as making prior findings of violation of (or compliance with) non-WTO rules? In my view, this power for panels to make prior findings of violation under non-WTO rules should not be easily accepted and may depend on the circumstances of each case. At that juncture, the dividing line between, on the one hand, substantive jurisdiction to decide on certain claims of violation (WTO claims only) and, on the other hand, the applicable law to decide on the validity of those claims (potentially all international law), becomes rather blurred. If the required finding of violation or compliance with obligations under non-WTO rules is both legally and factually straightforward, a panel may decide to move forward. Especially if no compulsory dispute settlement mechanism exists under the other treaty, a panel may be more inclined to decide the issue itself. After all, even if it finds a violation under the other treaty, it would only permit the consequential trade sanction and reject the WTO complaint. It would, in other words, simply confirm the status quo. If, in contrast, there is a special dispute settlement procedure available to determine violation of those non-WTO obligations – and the resolution of this matter is intricately linked to, and decisive for the outcome of, the WTO claims – then a panel may be wise to suspend its proceedings so as to give the parties the chance to first obtain a ruling under the other treaty.116 Yet, in other cases the additional task thus put on the panel may take it outside of its limited substantive jurisdiction. If so, the question is then how the panel should decide the case? It could simply disregard the non-WTO defense at issue and find a violation of the WTO treaty (even if it could later turn out that the violation is actually justified under the other treaty); Or the panel could decide that the dispute not genuinely concerns WTO claims (even though such claims can technically be made) but, rather, claims under another treaty to which the WTO claims are inextricably linked and for which the panel does not have substantive jurisdiction (in casu, the cooperation agreement, the Kimberley scheme or the bilateral settlement). The panel could, on that basis, decide that it does not have jurisdiction over the dispute and, thereby, for all practical purposes, reject the WTO complaint. If the defense under non-WTO rules is serious enough, in my view, the latter approach (decline jurisdiction) should be the preferred one, albeit (i) to avoid 116 See supra text at footnote 52.

How to Win a WTO Dispute Based on Non-WTO Law?

47

conflicting rulings and a fragmentation of international regimes and (ii) to provide an incentive to the parties to resolve the non-WTO questions amicably, after which they can always come back to the WTO for a resolution of the remaining dispute. V. Conclusion This essay analyses the different situations where non-WTO law can constitute an independent defense against claims of violation of WTO law. These situations are summarized in the table below: Other Treaties / Rulings Un- Other Treaties / Rulings Undermine the Merits of the dermine the Jurisdiction of WTO Complaint the WTO Panel 1. Bilateral agreement not to 1. Other treaties/standards exappeal; not to invoke the DSU plicitly referred to in the WTO treaty 2. Rulings/recommendations 2. Another treaty with exclusive jurisdiction under another treaty specifically imposing/permitting the measure at issue 3. Another treaty impos3. Another treaty with choice ing/permitting the measure at of forum but making any issue choice exclusive 4. Another treaty with compul- 4. Another treaty impossory (though not exclusive nor ing/permitting the measure at exclusionary) jurisdiction issue on condition that this 5. Res judicata effect of other other treaty is violated/complied with rulings In conclusion, and taking some distance from the specific solutions advocated here, it may be useful to sum up the alternatives available to panels when faced with non-WTO defenses, as well as some of the practical consequences and policy concerns that they entail. The panel completely disregards the defense under non-WTO law Be it in response to a defense that may undermine the jurisdiction of the WTO panel or one that may change the substantive outcome of the WTO complaint, the WTO panel could simply disregard the applicability of other rules of international law: x Not to apply other international rules consented to by the parties amounts, in the first place, to a disregard for the sovereign will of the disputing parties.

48

x

x

Joost Pauwelyn

To limit the resolution of the dispute to WTO rules, there where other rules are also applicable, does not resolve the dispute. At best, it resolves the dispute only partially; at worst, it leads to conflicting rulings and WTO countermeasures in response to something that is perfectly legal under international law. To resolve WTO complaints within the four corners of WTO covered agreements portrays the WTO as a self-contained regime and the field of public international law as a fragmented system with sealed-off compartments. It puts fuel on the argument that the WTO is only concerned about economic welfare; not about other values that may be expressed with equal force in other treaties. The panel declines jurisdiction on the ground of non-WTO law

Based on an agreement not to submit a certain dispute to a WTO panel or to submit it rather elsewhere, a WTO panel may decide that, by agreement of the parties, it does not have jurisdiction to decide the case. The panel may also decline jurisdiction on the ground that the dispute is in fact not genuinely a dispute under WTO rules, but raises rather claims under non-WTO rules for which it has no jurisdiction: x In all of these cases, the status quo is confirmed in that the WTO does not condemn nor condone the measure that was challenged. x The dispute can subsequently be brought before another court or tribunal, either because the parties had agreed earlier to resolve it there or because in substance the dispute is one under another treaty. In the latter case, any remaining WTO aspect of the dispute can then be referred back to a WTO panel, which should, in turn, take account of any rulings made under the other treaty. x Declining jurisdiction because the dispute is too intricately linked to non-WTO claims (and requires, for example, the making of findings of violation under another treaty) in a situation where no compulsory dispute settlement mechanism is available under the other treaty, amounts in practice to the continuation of the allegedly WTO inconsistent measure (the WTO complainant cannot go anywhere, for example, to confirm that it is not violating human rights so that it should not be subjected to trade sanctions). This result – though regrettable

How to Win a WTO Dispute Based on Non-WTO Law?

49

in case the WTO complainant is in the right – ought to provide an incentive for states to ameliorate the enforcement mechanism under other treaties. The panel takes account of a substantive defense under non-WTO law For a panel to take account of a defense under non-WTO law is not the same as accepting that defense. The WTO panel may find that the non-WTO rule does not require nor permit the challenged measure. Moreover, even if the opposite is true and a conflict arises between a WTO prohibition and an obligation or explicit right under another treaty, the conflict may still be decided in favor of the WTO provision. If, but only if, the other treaty norm prevails, the only consequence is that no finding of violation under the WTO treaty can be made (in other words, status quo). x Irrespective of the legal arguments made earlier, are members of WTO panels and the Appellate Body capable of interpreting non-WTO treaties and applying them to the facts? What is the WTO’s legitimacy when deciding questions under, for example, environmental or human rights treaties? Does it risk tainting those treaties with an inherent trade-bias? From the perspective of the drafters, enforcement agencies and other stakeholders under other, non-WTO treaties, the situation may boil down to choosing between two evils: either (1) the WTO completely ignores those treaties – thereby disregarding the sovereign will of the disputing parties as expressed in those other treaties – and prohibits something called for under another treaty; or (2) the WTO takes account of these other treaties with the risk that it misinterprets or waters down the other treaty. In my view, the latter solution constitutes, by far, the lesser evil (an evil that the WTO is, after all, engaging in already now when it regularly interprets, with competence and impartiality, non-WTO rules117). Moreover, this potential evil can be minimized or prevented relatively easily: First, as is already happening to date, members of the Appellate Body can be selected based not only on their trade expertise, but also on their broader knowledge of public international law or even domestic law generally speaking. Of the seven individuals currently on the Appellate Body at least four are not specialized in 117 See supra footnote 107 and recall, in this respect, the position expressed in the text supra at footnote 51.

50

Joost Pauwelyn

trade.118 When it comes to panels as well, room can be (and has been) made to have generalists and specialists in non-trade fields serving as panel members. Second, when applying non-WTO rules, WTO panels and the Appellate Body can seek expert advice not only from the disputing parties, third parties and the staff of the WTO secretariat, but also from other international organizations or even other international tribunals.119 To this effect, WTO panels could even suspend their proceedings.120 In this sense, they could adopt what is called in EU law the doctrine of acte claire121: if the meaning and application of the non-WTO rule offers no complication, a WTO panel could apply it without much need for expert advice or input from other sources. If, in contrast, the non-WTO rule offers ambiguity, then advice or even decisions from other bodies could be sought, not only to help the WTO in reaching the correct and most legitimate decision, but also to preserve the uniform interpretation of the other treaty.

References Lorand Bartels (2001), Applicable Law in WTO Dispute Settlement Proceedings, in: Journal of World Trade 35 (2001), 499519. Lorand Bartels (2002), The Human Rights Clause in the International Agreements of the European Community, doctoral thesis (European University Institute), April 2002. 118 These four are: Abi-Saab (expert in public international law, especially international human rights and international criminal law); Tanaguchi (expert in civil procedure); Lockhart (former Judge in Australia); and Sacerdoti (public international law, focusing on international investment). 119 See Pauwelyn (2002). 120 As done by the UNCLOS Arbitration Tribunal in the Mox Plant case, supra footnote 15, discussed supra in text at footnotes 41 and 52. 121 See Article 234 of the EC Treaty, setting out the procedure under which national courts can request preliminary rulings from the ECJ on matters of EU law. A national court must thus refer to the ECJ ‘if it considers that a decision on the question [under EU law] is necessary to enable it to give judgment’.

How to Win a WTO Dispute Based on Non-WTO Law?

51

Lucius Caflisch (2002), The Rome Statute and the European Convention on Human Rights, in: Human Rights Law Journal 23 (2002), No. 1-4, 1-12. James Cameron / Kevin R. Gray (2001), Principles of International Law in the WTO Dispute Settlement Body, in: International and Comparative Law Quarterly 50 (2001), 248-298. Meinhard Hilf (2001), Power, Rules and Principles – Which Orientation for WTO/GATT Law? in: Journal of International Economic Law 4 (2001), No. 1, 111-130. Qingjiang Kong (2002), Can the WTO Dispute Settlement Mechanism Resolve Trade Disputes Between China and Taiwan? in: Journal of International Economic Law 5 (2002), No. 3, 747758. Nancy Kontou (1995), The Termination and Revision of Treaties in the Light of New Customary International Law, Oxford (Oxford University Press) 1995. Kyung Kwak / Gabrielle Marceau (2002), Overlaps and Conflicts of Jurisdiction between the WTO and RTAs, Conference on Regional Trade Agreements, WTO, 26 April 2002, available at http://www.wto.org/english/tratop_e/ region_e / sem_april02_e /sem_april02_reading_e.htm. Vaughan Lowe (1996), Res Judicata and the Rule of Law in International Arbitration, African Journal of International Law 8 (1996), 38-50. Vaughan Lowe (1999), Overlapping jurisdiction in international tribunals, in: Australian Year Book of International Law 20 (1999), 191-204. Gabrielle Marceau (1999), A Call for Coherence in International Law – Praises for the Prohibition Against ‘Clinical Isolation’ in WTO Dispute Settlement’, in: Journal of World Trade 33 (1999), 87-152. Gabrielle Marceau (2001), Conflicts of Norms and Conflicts of Jurisdictions, The Relationship between the WTO Agreement and MEAs and other Treaties, in: Journal of World Trade 35 (2001), 1081-1131. Gabrielle Marceau (2002), WTO Dispute Settlement and Human Rights, in: European Journal of International Law 13 (2002), No. 4, 753-814.

52

Joost Pauwelyn

John McGinnis (2003), The Appropriate Hierarchy of Global Multilateralism and Customary International Law: The Example of the WTO, Northwestern Law and Economics Research Paper No. 03-09, available at http://papers.ssrn.com/sol3/papers. cfm?abstract_id=421661. David Palmeter / Petros Mavroïdis (1998), The WTO Legal System: Sources of Law, in: American Journal of International Law 92 (1998), 398-413. Joost Pauwelyn (2001), The Role of Public International Law in the WTO: How Far Can We Go?, in: American Journal of International Law 95 (2001), 535-578. Joost Pauwelyn (2002), The Use of Experts in WTO Dispute Settlement, in: International and Comparative Law Quarterly 51 (2002), 325-364. Joost Pauwelyn (2003a), Conflict of Norms in Public International Law, How WTO Law Relates to Other Rules of International Law, Cambridge (Cambridge University Press) 2003. Joost Pauwelyn (2003b), WTO Compassion or Superiority Complex? What to Make of the WTO Waiver for ‘Conflict Diamonds’, in: Michigan Journal of International Law 24 (2003), 1177-1207. Joost Pauwelyn (2003c), A Typology of Multilateral Treaty Obligations: Are WTO Obligations Bilateral or Collective in Nature?, in: European Journal of International Law 14 (2003), 907-951. Joost Pauwelyn (2004), Going Global or Regional or Both? Dispute settlement in the Southern African Development Community (SADC) and Overlaps with other Jurisdictions, in: Minnesota Journal of Global Trade 13 (2004), 231-304. Eugene F. Scoles et al. (2000), Conflict of Laws, Saint Paul / MN (Westgroup) 32000. Yuval Shany (2003), The Competing Jurisdictions of International Courts and Tribunals, Oxford (Oxford University Press) 2003. Joel Trachtman (1999), The Domain of WTO Dispute Resolution, in: Harvard International Law Journal 40 (1999), No. 2, 333377. Gaetan Verhoosel (2003), The Use of Investor-State Arbitration under Bilateral Investment Treaties to Seek Relief for Breaches

How to Win a WTO Dispute Based on Non-WTO Law?

53

of WTO Law, in: Journal of International Economic Law 6 (2003), No. 2, 493-506. World Bank (2003), The Inspection Panel 10 Years on, Washington / DC 2003.

Gabrielle Marceau and Anastasios Tomazos*

Comments on Joost Pauwelyn’s Paper: ‘How to Win a WTO Dispute Based on Non-WTO Law?’ I. Introduction II. Where Non-WTO Law Leads a Panel to Decline Jurisdiction III. Where Non-WTO Law Justifies What Would Otherwise Be a Violation of WTO Law IV. Conclusion References

55 57 66 78 80

I. Introduction We have been asked to provide general comments on our colleague’s paper entitled, ‘How to win a WTO dispute based on nonWTO Law’. As the title suggests, the primary focus of Joost Pauwelyn’s paper is to examine how and in what circumstances non-WTO can be applied by World Trade Organization (WTO) panels in disputes. In examining this issue, Pauwelyn distinguishes the following two types of situations, which in our view are nevertheless related: (1) where non-WTO law leads a panel to decline jurisdiction and (2) where non-WTO law justifies what would otherwise be a violation of WTO law. In the first situation, Pauwelyn argues that panels and the Appellate Body should decline jurisdiction when WTO Members have bilaterally agreed to set aside the jurisdiction of the WTO dispute settlement over WTO-related disputes and each time another inter-

*

Gabrielle Marceau is a Counsellor in the Cabinet of the DirectorGeneral of the WTO and Anastasios Tomazos at the time of writing this comment was a Legal Consultant for the Legal Affairs Division of the WTO Secretariat. The views expressed in this paper are strictly those of the authors and do not bind the WTO Secretariat or the WTO Members.

56

Gabrielle Marceau / Anastasios Tomazos

national tribunal has been set up to resolve a dispute or has already ruled on a similar matter pursuant to another treaty provision. In our view, in arguing that WTO panels should decline jurisdiction when another treaty also grants jurisdiction, which disregards the compulsory and automatic nature of the WTO dispute settlement system and, in creating unnecessary conflict of norms that facilitates pushing disputes outside the ambit of WTO law, Pauwelyn misconstrues fundamental aspects of the WTO dispute settlement system and overlooks the successful results of the Uruguay Round negotiations, where States (i.e., WTO Members) agreed to construct a more rule-based international trading system, primarily through the present dispute settlement system. In allowing WTO panels and the Appellate Body to recognize such treaties – which we suggest may be considered contrary to the automatic nature1 of the WTO dispute settlement system – one of the implications of Pauwelyn’s thesis is that it would leave weaker trading partners in the hands of more powerful States, capable of imposing other jurisdictional mechanisms that are less impartial, comprehensive and rule-orientated than that of the WTO dispute settlement system. In the second situation, Pauwelyn argues that in some instances, WTO adjudicating bodies should find, on the merits of a dispute, that there is no WTO violation where non-WTO law prevails over WTO law. In reaching this conclusion, Pauwelyn suggests a very broad definition of conflict of norms and provides several examples of possible conflicts between WTO provisions and other international obligations. Pauwelyn broadens the notion of conflict of norms by including conflicts between rights and obligations or exceptions under Article XX of GATT 1994. He also contends that WTO relations are bilateral by nature and therefore, two Members can agree to derogate from provisions of the WTO. Then, Pauwelyn suggests that WTO panels and Appellate Body are required to resolve any and all conflicts that may exist between WTO provisions and other international obligations and rights of States when they are involved in a WTO dispute. Ostensibly, for Pauwelyn, identifying conflicts appears to ensure that panels and the Ap1

This term refers to the so-called negative consensus rule set out, inter alia, in Article 6.1 of the DSU, whereby all Members present at a DSB meeting, including the complainant requesting the establishment of the panel, decide by consensus not to establish a panel.

Comments on Joost Pauwelyn’s Paper

57

pellate Body are able to adjudicate the issue objectively as well as respecting international law. In our view, Pauwelyn’s perspective overlooks the nature and the state of public international law at present. States are bound at all times by all their international obligations and must always respect the rights of other States. Moreover, States must also negotiate and apply their international obligations in good faith at all times. Although Pauwelyn does acknowledge in his paper that it would be exceptional if a harmonious interpretation of WTO law and non-WTO law was not possible and the latter provided an ‘independent defence’ to a violation of the WTO treaty, we are of the view that WTO law and other provisions of international law can generally be applied harmoniously and effectively through good faith interpretation. Even though WTO panels and the Appellate Body cannot interpret and enforce non-WTO law, other than to the extent necessary to interpret and apply WTO provisions, one should not underestimate the potential coherence that exists between WTO law and the other systems of international law. We believe Pauwelyn overemphasizes the role of conflict of norms in resolving WTO disputes. Pauwelyn ignores the ‘chaotic’ nature of international law and seeks to compartmentalize each system of international law in such a manner that almost assures that conflicts will be created as issues often overlap between different sub-systems of international law. If Pauwelyn’s argument is accepted, it would grant a specialized tribunal, such as a WTO panel, powers for which it has not been conferred or possess the capacity to address. More importantly, instead of favoring an evolving coherence and mutual respect between international legal systems, Pauwelyn’s viewpoint would require the WTO dispute settlement system and other international tribunals to compartmentalize, separate and dissociate functions and duties contrary to their overlapping mandates. This cannot be correct. Although no one would argue that the WTO is not an important international organization, it is not a world government and its dispute settlement system even less so. II. Where Non-WTO Law Leads a Panel to Decline Jurisdiction WTO panels are not courts of general jurisdiction; they have a delegated and limited jurisdiction. Therefore, panels must comply with the mandate received from WTO Members through the Dis-

Gabrielle Marceau / Anastasios Tomazos

58

pute Settlement Body (DSB). Pursuant to Articles XXII and XXIII of GATT 1994 and the Dispute Settlement Understanding (DSU) Members have delegated to panels and the Appellate Body the power to adjudicate their disputes and make recommendations to the DSB. The parameters of this delegation and the jurisdiction of the WTO adjudicating bodies are determined by the DSU and the Marrakesh Agreement Establishing the WTO. This may explain why GATT/WTO jurisprudence has established that panels and the Appellate Body may still rule on the WTO compatibility of measures that have been removed during proceedings. For instance, in Argentina – Textiles and Apparel, Argentina had changed the level of its statistical tax during the proceedings. The Appellate Body nevertheless examined the level of the tax as it was at the time the Panel was established: ‘At the time the Panel proceeding commenced, there was in effect in Argentina an ad valorem tax of 3 per cent on imports, without a minimum or a maximum charge, which was called a ‘statistical tax’ and was described as designed to cover the cost of providing a statistical service intended to provide a reliable data base for foreign trade operators.i i

According to Argentina’s statement at the oral hearing on 23 February 1998, this ad valorem statistical tax was modified to 0.5 per cent in December 1997.’2

Likewise, in US – Certain EC Products, the Appellate Body concluded that although the measure at issue was no longer in existence, it nonetheless found that the measure was inconsistent with Article II:1 of GATT 1994. However, since the measure was no longer in existence, no recommendation could be made to the DSB.3 Other panels and the Appellate Body have analyzed and reached conclusions on measures removed or modified at the time of their rulings.4 This is further evidence, as elaborated below, that WTO adjudicating bodies have a defined and limited jurisdiction. 2 3 4

Appellate Body Report on Argentina – Textiles and Apparel, WT/DS56/AB/R, at para. 64 Appellate Body Report on US – Certain EC Products, WT/DS 165/AB/R, at paras, 81, 82 and 129. See for instance, the US - Gasoline Panel Report, WT/DS2/R, para. 6.19; Panel Report on US - Shirts and Blouses, WT/DS33/R, at para. 6.2; Panel Report on Indonesia – Autos, WT/DS54/R, WT/DS55/R,

Comments on Joost Pauwelyn’s Paper

59

To examine the issue of the jurisdiction of WTO adjudicating bodies discussed by Pauwelyn, we must first turn our attention to Article 23 of the DSU.5 This provision is perhaps one of the most fundamental provisions of the DSU. As a means of prohibiting unilateral measures, the purpose of Article 23 of the DSU is to ‘multilateralize’ the dispute settlement system by obliging Members to seek redress only through the DSU, as well as, renouncing the use of unilateral trade countermeasures other than in conformity with WTO law. In light of Article 23 of the DSU, which provides that a violation of the WTO Agreement can be addressed only according to the WTO/DSU rules, would the invocation of a Regional Trade

5

WT/DS59/R, WT/DS64/R;. Panel Report on EEC - Measures on Animal Feed Proteins, adopted on 14 March 1978, BISD 25S/49; and Panel Report on US - Tuna and Tuna Products from Canada, adopted on 22 February 1982, BISD 29S/91. Article 23 of the DSU reads as follows: ‘Strengthening of the Multilateral System 1. When Members seek the redress of a violation of obligations or other nullification or impairment of benefits under the covered agreements or an impediment to the attainment of any objective of the covered agreements, they shall have recourse to, and abide by, the rules and procedures of this Understanding. 2. In such cases, Members shall: (a) not make a determination to the effect that a violation has occurred, that benefits have been nullified or impaired or that the attainment of any objective of the covered agreements has been impeded, except through recourse to dispute settlement in accordance with the rules and procedures of this Understanding, and shall make any such determination consistent with the findings contained in the panel or Appellate Body report adopted by the DSB or an arbitration award rendered under this Understanding; (b) follow the procedures set forth in Article 21 to determine the reasonable period of time for the Member concerned to implement the recommendations and rulings; and (c) follow the procedures set forth in Article 22 to determine the level of suspension of concessions or other obligations and obtain DSB authorization in accordance with those procedures before suspending concessions or other obligations under the covered agreements in response to the failure of the Member concerned to implement the recommendations and rulings within that reasonable period of time.’

60

Gabrielle Marceau / Anastasios Tomazos

Agreement (RTA) provision be sufficient to stop the automatic nature of the WTO dispute settlement system? How can Article 23 and the automatic nature of the DSU be reconciled with the preference and, in some circumstances, the exclusivity given to the RTA dispute settlement process concerning obligations which are similar in the RTA and in WTO for the same facts? Pauwelyn argues that if a RTA panel has ruled on a matter between the parties, the WTO panel should decline to rule on the similar WTO claim; and that if the RTA process is ongoing, the WTO process should be halted. Although we would do not in principle disagree with Pauwelyn’s suggestion, it does not reflect the present state of international law. Rather, it is our view that in light of the automatic nature of the mechanism, once a dispute is initiated under the DSU, it is unlikely that a WTO panel would give any consideration to the respondent's request to halt the procedures just because similar or related procedures are being pursued under a RTA. Parties could de facto not bring the matter to the WTO, but we think that it would be difficult for a WTO panel to refuse to hear a WTO Member complaining about a measure claimed to be inconsistent with the WTO Agreement on the ground that the complaining or responding Member is alleged to have a more specific or more appropriate defence or remedy in another forum concerning the same legal facts. In practice, the WTO panel will most likely read and consider the related RTA ruling(s) and may even use it to confirm its own interpretation of the WTO provisions and their application to the matter at issue. However, to decline jurisdiction because a RTA dispute ruling has already been rendered or ongoing is a totally different matter. In initiating a parallel WTO dispute, a RTA party may be found to be violating the obligation under the RTA not to initiate a dispute outside the RTA. In these circumstances, the RTA party opposed to the parallel WTO panel (the ‘opposing RTA party’) would claim that the WTO panel initiated by the other RTA party is impairing some of its benefits under the RTA. The opposing RTA party would arguably win this claim before the RTA panel. Theoretically, that opposing RTA party would then be entitled to some retaliation, the value of which could probably correspond to (part of) the benefits that the other RTA party could gain in initiating its WTO panel. In other words, even if it may not be practical or useful for a RTA party to duplicate a dispute that should be dealt with in RTA, there would be no legal obstacle against such a possibility,

Comments on Joost Pauwelyn’s Paper

61

since, legally speaking, the RTA tribunal and WTO panel would be considering different ‘matters’, and different ‘applicable law’ in mechanisms that offers different remedies, implementation and enforcement rules.6 There could be an overlap or conflict of jurisdiction between the dispute settlement mechanism of the WTO and that of RTAs. The wording of Article 23 of the DSU seems to make it clear that a WTO adjudicating body always has the authority (and even the obligation) to examine claims of violations of WTO obligations. WTO rights and obligations can be challenged only pursuant to the WTO dispute settlement procedures and only before a WTO adjudicating body.7 In addition, WTO jurisprudence has decided that any WTO Member that is a ‘potential exporter’ has the sufficient legal interest to initiate a WTO panel process.8 That is to say, in the context of a dispute between two WTO Members involving issues covered by both an RTA and the WTO Agreement, any WTO Member which considers that any of its WTO benefits have been nullified or impaired has the absolute right to trigger the WTO dispute settlement mechanism and to request the establishment of a panel on this matter.9 Such a WTO Member cannot be asked (and arguably cannot even agree) to take its WTO dispute to another forum, even if that other forum appears to be more relevant or better equipped to deal with the issues involved. Tensions may also arise from the availability of RTA noncompulsory dispute settlement mechanism with no binding effect even in the absence of strict de jure conflicts (but when faced with overlaps of jurisdictions). For instance, trade measures taken pursuant to non-compliance with an RTA adjudication process could be 6

7

8

9

However, it is worth pondering on the potential issue of how a WTO arbitration panel would deal with retaliation already enforced under the North American Free Trade Agreement if the potential WTO retaliation is to be exercised on the same trade flows. Even an arbitration pursuant to Article 25 of the DSU would be a WTO arbitration, and thus covered by the exclusivity provision of Article 23 of the DSU. Appellate Body Report on EC – Bananas III, WT/DS27/AB/R at para. 136; See also the Panel Report on Korea – Dairy, WT/DS98/R, at para. 7.13. Appellate Body Report on US – Shirts and Blouses, WT/DS33/AB/R, at p. 13.

62

Gabrielle Marceau / Anastasios Tomazos

argued to be inconsistent with Article 23 of the DSU and Article XI of GATT. The benefits gained from such RTA countermeasures may be nullified by the consequences of a violation of Article 23 of the DSU and Articles II and XI of GATT 1994. It is therefore for WTO Members to negotiate how they want to allocate jurisdiction between RTAs and the WTO, and how the dispute settlement mechanism of RTAs and those of the WTO will operate. Otherwise, they will evolve in parallel. If so, they should be used and operated in good faith. As of now, they generally appear to be doing so. In recent years, treaties and organs of jurisdiction have dramatically increased in number. An obvious example is that of the multiplicity of treaties, organs and jurisdictions involved in human rights issues.10 It seems accepted practice that States may adhere to different but parallel dispute settlement mechanisms for parallel or even similar obligations. As stated by the Arbitral Tribunal (ICSID/ITLOS) in the recent Southern Bluefin Tuna case: But the Tribunal recognizes as well that there is a commonplace of international law and State practice for more than one treaty to bear upon a particular dispute. There is no reason why a given act of a State may not violate its obligations under more than one treaty. There is frequently a parallelism of treaties, both in their substantive content and in their provisions for settlement of disputes arising thereunder. [...] the conclusion of an implementing convention does not necessarily vacate the obligations imposed by the framework convention upon the parties to the implementing convention.11 (emphasis added) There may be situations where two Members agree not to bring one or several disputes to the WTO. However, if a Member brings a dispute to the WTO, contrary to a bilateral agreement or to an exclusive jurisdiction clause in another treaty, that Member will probably argue that the bilateral agreement or the exclusive jurisdiction clause is not applicable or does not cover the matter at issue before the WTO on the bases that the applicable law, the remedies available and the enforcement mechanisms, etc. are not the same. Thus, a panel would need to examine the substance of the claims of WTO violation to assess whether and how they overlap with the 10 11

See Roucounas (1987), 197. Award on Jurisdiction and Admissibility of 4 August 2000, Southern Bluefin Tuna Case, Australia and New Zealand v Japan, at p. 91.

Comments on Joost Pauwelyn’s Paper

63

provisions of the bilateral agreement or the treaty. After all this work, it is doubtful that in the present state of international law, a panel would simply decline jurisdiction and rule that another judicial or quasi-judicial body is the most appropriate forum to examine the issue, or has already solved the dispute. Moreover, in the absence of an agreement among states, the principle of forum non conveniens seems not to be applicable in the present state of international law.12 The same holds true for an allegation based on the principle of res judicata. We agree with Vaughan Lowe that there is no such general principle yet applicable between international tribunals.13 Also, we doubt that the establishment of the WTO has changed this situation. Furthermore, it would also be very difficult for the principle of res judicata to find application in the context of WTO dispute settlement between an international tribunal and a WTO panel, as the notion of ‘matter’14 pursuant to Article 7.1 of the DSU determines the panel's terms of reference, which in turn defines the scope of a panel's jurisdiction.15 Even between WTO litigants within WTO dispute settlement, it would still be difficult for the respondent to invoke the principle of res judicata, as confirmed below by the Panel in India – Autos: ‘Because the policy underlying res judicata is to bring litigation of a particular nature to an end at an appropriate stage, the key to its application should be to compare what has already been ruled on to what is being brought before the adjudicating body in the subsequent proceedings. Both India and the United States have used a comparison between the ‘matter’ ruled on in the India - Quantitative Restrictions case and the ‘matter’ brought before this Panel in identifying the similarities or dissimilarities between 12 13 14

15

Infra, Lowe (1999), 12. Lowe (1999), 12; Sawaki (1979–80); Fawcett, 5 et seq. and 10. See also Lowe (1996). See the Appellate Body Report in the Guatemala – Cement I, WT/DS60/AB/R, at para. 72, where the Appellate Body stated that the ‘matter’ referred to the DSB consists of two elements: the specific measures at issue and the legal basis of the complaint (or the claims). See the Appellate Body Report on India – Patents (US), WT/DS50/AB/R, at paras. 92-93.

64

Gabrielle Marceau / Anastasios Tomazos

the two disputes for the purposes of assessing whether the issues before this Panel can be considered to be res judicata.16 [footnotes omitted] [...] The Panel therefore considers that for res judicata to have any possible role in WTO dispute settlement, there should, at the very least, be in essence identity between the matter previously ruled on and that submitted to the subsequent panel. This requires identity between both the measures and the claims pertaining to them. There is also, for the purposes of res judicata, a requirement of identity of parties which is clearly met with regard to the United States in this instance.’17[footnotes omitted] Another issue that Pauwelyn addresses is bilateral agreements between WTO Members not to invoke WTO dispute settlement. In this regard, Pauwelyn considers the European Communities’ decision to subsequently initiate a dispute in India – Autos, despite reaching a mutually agreed solution with India in the context of the India – Quantitative Restrictions18 dispute, whereby both parties had agreed to refrain from any action during the phasing-out period on the condition that India complied with its obligations under certain exchange of letters. Although the Panel in India – Autos was able to avoid the issue of whether India could rely on the mutually agreed solution on the basis that the agreement did not cover the same matter within the meaning of Article 7.1 of the DSU, Pauwelyn is of the view that ‘any WTO panel would have been under an obligation to respect this agreement and to declare that by agreement of the parties, it does not have jurisdiction to examine the issue’. We disagree with Pauwelyn that this is an example where a Panel would rule that it had no jurisdiction ‘because of non-WTO law’. In our view, the Panel would have been able to resolve this issue without having to resort to applying substantive ‘non-WTO law’. The Panel seemed to agree with India’s argument that the issue related to the European Communities’ procedural rights under the 16 17 18

Panel Report on India – Autos, WT/DS146/R and Corr.1, WT/DS175/R and Corr. 1, para. 7.64. Ibidem, para. 7.66. Panel Report on India – Quantitative Restrictions, WT/DS90/R.

Comments on Joost Pauwelyn’s Paper

65

DSU, rather than India’s substantive obligations under a covered agreement. In this regard, the Panel was correct to point out that, ‘the status of mutually agreed solutions under the DSU and their impact in subsequent dispute settlement proceedings is not expressly indicated in the DSU’19 and that the issue was ‘what effects it may have on the exercise of procedural rights under the DSU in subsequent proceedings’.20 In other words, the issue was what effect does the agreement have on the ability of a Member to subsequently request the establishment of a panel. We believe that this is one of those instances where basic general principles of procedural law could be employed by a panel in a future dispute to determine whether a mutually agreed solution negatively affects the procedural rights of a complainant in such a way that it bars that Member's ability to ‘re-litigate’ on a matter that has already been settled. In other words, as suggested by the Panel in India – Autos, a panel could apply the principle of estoppel in such circumstances.21 We also hasten to add that this interpretation would be one in which a panel would be authorized to make under Article 11 of the DSU. Since WTO panels are obliged to presume that WTO Members comply with their international law obligations, they should interpret and apply WTO law accordingly. Moreover, since WTO panels and the Appellate Body are quasi-judicial they must, and do comply with basic general principles of procedural law. Both panels and the Appellate Body have interpreted and applied Article 11 of the DSU to encompass general principles of procedural law and due process rules, which have evolved into rules on the burden of proof, on the representation of parties before panels, standing, judicial economy and so forth. Here, we generally agree with Pauwelyn’s assertion that, ‘as important as those processes of references to non-WTO law may be, it is unlikely they determine the substantive outcome of a WTO dispute [...] applying the largely procedural rules of general international law may eventually decide a case, but hardly influences its substantive merits.’ Finally, it is also worth citing the following excerpt from India – Autos, where the Panel implies that it is not necessarily interpreting and applying non-WTO law (i.e., a non-covered agreement) when examining a mutually agreed solution: 19 20 21

Supra, India – Autos, para. 7.113. Ibidem, para. 7.116. Ibidem, footnote 364.

66

Gabrielle Marceau / Anastasios Tomazos

‘[...] These possibilities suggest that the issue cannot necessarily be resolved simply through an acknowledgement that an MAS [mutually agreed solution] is not a covered agreement as was argued by the EC. That argument simply is another way of noting that the DSU does not expressly give a panel a mandate to consider whether a ‘violation’ of such an agreement might exist as a distinct basis for a dispute under the DSU. It does not necessarily prove that a panel may not in some circumstances need to consider the terms of such agreed solutions in order to fulfill its duties under the DSU. Here the Panel notes that disputes concerning the application of the DSU itself can be the object of proceedings under the DSU. This might possibly include disputes concerning mutually agreed solutions, since these are expressly referred to in the DSU.’22 III. Where Non-WTO Law Justifies What Would Otherwise Be a Violation of WTO Law Before discussing the relationship between WTO law and general international law and the alleged capacity of WTO adjudicating bodies to assess Members’ compliance with non-WTO treaty provisions, one must understand the manner in which the WTO dispute settlement process functions. In many ways, the WTO dispute settlement system is an attractive forum to settle disputes for States that are members of the WTO as it can be triggered fairly easily and quickly. Also, another appealing feature is that panels and the Appellate Body are expected to render relatively quick decisions. Virtually any allegation would generally suffice to formally trigger the WTO dispute settlement process through a simple request for consultations in writing, copied to the DSB. All Members are presumed to have an economic and legal interest to initiate a WTO dispute settlement mechanism.23 As noted above, once triggered, the WTO dispute settlement system is automatic in view of the so-called negative consensus rule found in Article 6.1 DSU. 22 23

Ibidem, footnote 364. However, in light of the negative consensus rule, nothing would prevent a Member from initiating a formal WTO dispute even if that Member did not have either an economic or legal interest.

Comments on Joost Pauwelyn’s Paper

67

As previously mentioned, Article 23 of the DSU is arguably one of the most fundamental provisions of the DSU. One of the unique features of the WTO is that it has managed to regulate and limit unauthorized unilateral countermeasures from powerful Members. The WTO system of law contains, inter alia, specific: (1) rights and obligations (2) claims and causes of action (3) violations (4) enforcement mechanisms and (5) remedies in case of violation. The DSU defines the jurisdiction, the capacity and the mandate of panels (and the Appellate Body) with reference to (1) allegations of WTO violations by the complaining party(ies) (2) the specific type of remedies/conclusions that panels and the Appellate Body may recommend and (3) the prohibition not to add to or diminish WTO law. Indeed, paragraph 1 of Article 23 of the DSU has been interpreted as being an ‘exclusive dispute settlement clause’.24 The mandate of panels and the Appellate Body is to determine whether provisions of the WTO ‘covered agreements’ have been violated. Pursuant to Article 1.1, the DSU applies to disputes brought under the covered agreements. The covered agreements listed in Appendix 1 of the DSU are all the WTO multilateral trade agreements, (except the Trade-Related Review Mechanism) and the plurilateral agreements, to the extent that Members have adopted the latter. The covered agreements would also include WTO decisions and secondary legislation. Article 4 provides that consultations can be initiated on allegations of violations of any of the covered agreements. Article 7.1 states that the mandate of the panel is ‘(t)o examine, in the light of the relevant provisions in (name of the covered agreement(s) cited by the parties to the dispute), the matter referred to the DSB by the Member in document [...] and to make such findings as will assist the DSB in making the recommendations or in giving the rulings provided for in that/those agreement(s)’ (emphasis added). Article 7.2 adds that ‘Panels shall address the relevant provisions in any covered agreement or agreements cited by the parties to the dispute’ (emphasis added). Article 11 of the DSU also provides for a limited jurisdiction for panels, which is circumscribed by the covered agreements. It requires a panel to ‘make an objective assessment of the matter before it, including an objective assessment of the facts of the case and the applicability of and conformity with the relevant covered agreements, and make such other 24

Panel Report on US – Section 301 Trade Act, WT/DS 152/R, at para. 7.43.

68

Gabrielle Marceau / Anastasios Tomazos

findings as will assist the DSB in making recommendations or in giving the rulings provided for in the covered agreements’ (emphasis added). Finally, Article 19.1 of the DSU provides that the standard recommendation is that the losing Member ‘brings its measure into conformity with that [covered] agreement(s)’. No other conclusion or remedy is envisaged. At most, WTO adjudicating bodies can make ‘suggestions’25 as to the manner in which the measure should be brought into conformity. However, we hasten to add that such suggestions are not binding and cannot be enforced. Therefore, a WTO adjudicating body could not enforce a suggestion regarding the manner in which a Member could amend its laws, regulations and the like, so as to bring them into conformity with non-WTO norms, unless it could be related to compliance with a provision of the covered agreements. The limited jurisdiction of WTO adjudicating bodies is confirmed by the cautious jurisprudence of the Appellate Body. In EC – Poultry, Brazil claimed that the European Communities had not provided it with the full allocation of a tariff quota on frozen chicken imports, contrary to obligations under the EC schedules and their bilateral ‘Oilseeds Agreement’. Here, the Appellate Body acknowledged that the Oilseeds Agreement was not ‘applicable law’ and thus could not be enforced by WTO dispute settlement mechanism. It stated: ‘In our view, it is not necessary to have recourse to either Article 59.1 or Article 30.3 of the Vienna Convention … As such, it [the Schedule of the EC] forms part of the multilateral obligations under the WTO Agreement. The Oilseeds Agreement, in contrast, is a bilateral agreement negotiated by the European Communities and Brazil under Article XXVIII of the GATT 1947, as part of the resolution of the dispute in EEC – Oilseeds. As such, the Oilseeds Agreement is not a “covered agreement” within the meaning of Articles 1 and 2 of the DSU. Nor is the Oilseeds Agreement part of the multilateral obligations accepted by Brazil and the European Communities pursuant to the WTO Agreement, which came into effect on 1 25

Article 19.1 in fine reads as follows: ‘In additions to its recommendations, the panel or the Appellate Body may suggest ways in which the Member concerned could implement the recommendations.’

Comments on Joost Pauwelyn’s Paper

69

January 1995. The Oilseeds Agreement is not cited in any Annex to the WTO Agreement. Although the provisions of certain legal instruments that entered into force under the GATT 1947 were made part of the GATT 1994 pursuant to the language in Annex 1A incorporating the GATT 1994 into the WTO Agreement, the Oilseeds Agreement is not one of those legal instruments.’26 According to the Appellate Body, Schedule LXXX rather than the Oilseeds Agreement, contained the relevant obligations of the European Communities under the WTO Agreement. Therefore, it was Schedule LXXX, rather than the Oilseeds Agreement, which formed the legal basis for this dispute and which must be interpreted in accordance with ‘customary rules of interpretation of public international law’ under Article 3.2 of the DSU.27 This seems to suggest that even agreements negotiated under the auspices of the WTO Agreement, such as those negotiated under Article XXVIII of GATT, remain useful tools of interpretation, but as such, cannot be enforced by WTO adjudicating bodies (unless the WTO provision explicitly provides otherwise).28

26 27

28

Appellate Body Report on EC – Poultry, WT/DS69/AB/R, at para. 79. Appellate Body Report on EC – Poultry, WT/DS69/AB/R, at para. 81. The Appellate Body continued and stated that the Oilseed Agreement could be used pursuant to Article 32 of the Vienna Convention, as part of the circumstances of the negotiation of the Schedule. See the ruling of the Chairman in United States – Margin of Preferences, 9 August 1949, BISD II/11 to the effect that a bilateral agreement cannot be enforced by a GATT Panel. The issue of the WTO compatibility of a regional trade agreement with WTO provisions, including Article XXIV, is not really different since, should panels have wide jurisdiction to assess the overall compatibility of regional trade agreements, panels would still be examining whether a Member’s specific measure or its regional trade agreement with other Members is compatible with the WTO agreements, taking into account the possible exceptions authorized by Article XXIV. In all cases panels would not be ‘enforcing’ the provisions of the regional trade agreement, something that could be done by the parties to the regional trade agreement only pursuant to the dispute settlement procedures of the regional trade agreement itself. On the relationship

70

Gabrielle Marceau / Anastasios Tomazos

Moreover, in EC – Bananas III, the Appellate Body upheld the panel’s decision that it ‘had no alternative but to examine the provisions’ of a non-WTO agreement ‘in so far as it is necessary’ to interpret WTO rules (the Lomé waiver referred to the Lomé Convention and the panel was examining the scope of the Lomé Waiver).29 The implication is that the role of non-WTO instruments is limited to interpreting the non-WTO rule to extent necessary to decide the matter at hand. Similarly, a panel may be called upon in the future to examine the Kimberley Certification Scheme for rough diamonds – a non-WTO agreement – ‘in so far as it is necessary’ to interpret the scope of the Kimberley Waiver30 which refers to the Kimberley Certification Scheme. Having said that the WTO adjudicating bodies have a limited jurisdiction – to interpret and apply the covered agreements – does not mean that the WTO should be interpreted and applied in isolation from the rest of international law. In its first report, US – Gasoline, the Appellate Body noted that the WTO agreements must not be interpreted in ‘clinical isolation’ from public international law.31 In this regard, the Appellate Body cited Article 3.2 of the DSU which requires panels and the Appellate Body to use ‘customary rules of interpretation’ to interpret the provisions of the WTO agreements. Article 3.2 provides: ‘[...]The dispute settlement system of the WTO] serves to clarify the existing provisions of those [WTO] agreements in accordance with customary rules of interpretation of public international law [...].’ Customary rules of international law include (at least) Articles 31, 32 and 33 of the Vienna Convention. Of particular interest is Article 31.3(c) which provides that when interpreting the ordinary meaning of treaty terms, an interpreter shall also ‘take into account’ any applicable rule of international law. Article 31.3(c) would bring

29 30

31

between the dispute settlement procedures under regional trade agreements and that of the GATT or the WTO see Marceau (1997). Appellate Body Report on EC - Bananas III , WT/DS27/AB/R, at para. 162. Waiver Concerning Kimberly Process Certification Scheme for Rough Diamonds: Communication, G/C/W/432//Rev.1 (24 February, 2003). Appellate Body Report on US – Gasoline, WT/DS2/AB/R at page 17. See also McRae (1997), 99.

Comments on Joost Pauwelyn’s Paper

71

into the interpretative analysis of any WTO provision a series of relevant international rights and obligations which may alter and influence the meaning of WTO provisions. Article 31.3(c) is a tool, an indicator of the norms/rules of international law that must be taken into account in the interpretation of specific treaty provisions, such as those of the WTO. In fact, Article 31.3(c) of the Vienna Convention can be viewed as an obligation on the interpreter to be ‘aware of’ – and take into account – what is otherwise international law between the WTO disputing parties. In sum, Article 31.3(c) of the Vienna Convention aims at promoting some ‘coherence’ in international law, so that the treaty being interpreted and other relevant international law rules are read in a way that mutually supports and avoids conflicts with other treaties. The WTO Agreement, as any other treaty, should be interpreted in a fashion that takes into account other relevant and applicable rules of international law. In doing so in good faith, it should generally be possible to interpret WTO provisions in a way that allow WTO Members to respect all their international law obligations and other States’ rights. We agree with Donald McRae that Article 3.2 of the DSU is a mandate for interpreting the agreement and the real concern is: ‘[...] whether the treaty obligations of the WTO Members are to be supplemented by additional obligations derived from customary international law. On the face of it, Article 3.2 of the DSU says they cannot. Dispute settlement cannot add to the obligations of the WTO Members as set out in the Agreements.’32 ‘Principles of customary international law do not leapfrog into treaty regimes an add to substantive obligations under those regimes. They exist alongside and are relevant to the interpretation and application of the substantive treaty rules.’33 This distinction seems to have been recognized by WTO jurisprudence. Even if WTO applicable law seems to exclude the direct application of rules on State responsibility,34 or some of them, these 32 33 34

McRae (2003), 713. Ibidem, 715 See the Panel Report in US – Certain EC Products at para. 6.133: ‘[…] In short the regime of counter-measures, reprisals or retaliatory measures has been strictly regulated under the WTO Agreement. It is

72

Gabrielle Marceau / Anastasios Tomazos

rules – to the extent they are customary – bind WTO Members and remain a relevant benchmark for the interpretation of WTO law which is presumed to evolve consistently with international law. In US – Cotton Yarn, the Appellate Body, after analyzing the safeguard provisions of the WTO Agreement on Textiles and Clothing, stated: ‘Our view is supported further by the rules of general international law on state responsibility, which require that countermeasures in response to breaches by states of their international obligations be commensurate with the injury suffered.’35 In a nutshell, the mandate of panels and the Appellate Body is defined and limited: to interpret WTO law and decide whether a provision of the covered agreements has been violated. In doing so they apply and enforce the WTO law. In accordance with their respective mandates, the WTO panels and the Appellate Body only have the jurisdiction to interpret and apply WTO law. As such, they cannot interpret, let alone reach any legal conclusions of violation or compliance with other treaties or customs in complete isolation from the covered agreements. If and when they do so, it is only to the extent necessary to interpret provisions of the covered agreements. As mentioned above, WTO panels and the Appellate Body must presume that WTO Members comply with their international law obligations and therefore, they should interpret and apply WTO law accordingly. Moreover, since WTO panels and the Appellate

35

now only in the institutional framework of the WTO/DSB that the United States could obtain a WTO compatible determination that the European Communities violated the WTO Agreement, and it is only in the institutional framework of the WTO/DSB that the United States could obtain the authorization to exercise remedial action.’ [footnote 170: ‘Therefore, in the WTO context, the provision of Article 60 of the Vienna Convention on the Laws of Treaties (1969) on this matter does not apply since the adoption of the more specific provisions of Article 23 of the DSU.’] Appellate Body Report on United States – Cotton Yarn, WT/DS192/AB/R, at para. 120. The same is true for the use of the principle of ‘due diligence’ in paras. 67, 76, 77 and 79 as with the use of ‘good faith’ in para. 81 of the same Report.

Comments on Joost Pauwelyn’s Paper

73

Body are quasi-judicial they must, and do comply with basic general principles of procedural law. Pauwelyn, however, limits the scope for a coherent interpretation between WTO provisions and non-WTO provisions. To do so, Pauwelyn asserts as ‘conflicts’ differences between treaty provisions. For instance, he considers that in situations where a Member invokes compliance with an Multilateral Environmental Agreement (MEA) as justification under Article XX of GATT 1994 for a breach of a WTO rule, a ‘conflict’ may exist between the MEA and Article XI of GATT 1994, which generally disallows trade prohibitions and restrictions other than duties, taxes or other charges, despite the existence and applicability of Article XX of GATT 1994. Since Article XX of GATT 1994 explicitly allows Members to give priority to policies other than trade, including those policies affected by an MEA, it is erroneous to claim that there is conflict between a WTO rule that disallows trade prohibitions and restrictions and the MEA. In international law, for a ‘conflict’ to exist between two treaties, three conditions have to be met. First, the treaties must have some overlap in membership. Second, the treaties must cover the same substantive subject matter. Otherwise, there would be no possibility for conflict. Third, the provisions must conflict, in the sense that the provisions must impose mutually exclusive obligations.36 The general principle of good faith in the interpretation and application of treaties call for a presumption against conflicts. The presumption against conflict is especially reinforced in cases where separate agreements are concluded between the same parties, since it can be presumed that they are meant to be consistent with one another in the absence of any evidence to the contrary.37 For a conflict to exist, it must be established that a provision of the WTO Agreement mandates an action that a provision of another treaty prohibits or vice-versa, when a provision of another treaty mandates an action that the WTO Agreement prohibits. The occurrence of such a situation would be quite rare. In fact, one would be required to demonstrate that compliance with the WTO necessitates violation of that other treaty.

36 37

Jennings / Watts (1992), 1280; Fitzmaurice (1957), 237; Sinclair (1984), 97. See also Jenks (1953), 425 et seq.

74

Gabrielle Marceau / Anastasios Tomazos

In the WTO context,38 the Appellate Body in Guatemala – Cement I elaborated on the issue of the possibility of conflicts between the special or additional rules of procedures in Appendix 2 of the DSU with respect to anti-dumping disputes and the general provisions of the DSU and concluded as follows: ‘[...] A special or additional provision should only be found to prevail over a provision of the DSU in a situation where adherence to the one provision will lead to a violation of the other provision, that is, in the case of a conflict between them. An interpreter must, therefore, identify an inconsistency or a difference between a provision of the DSU and a special or additional provision of a covered agreement before concluding that the latter prevails and that the provision of the DSU does not apply.’39 Conflicts are concerned with situations where two ‘obligations’ cannot be reconciled. If the issue is one in which a treaty contains a right and the other an obligation a conflict cannot exist. ‘Rights’ within a treaty and contained in other treaties, must also be respected and enforced by States at all times. There is, however, no need to expand the concept of conflict as other rules of international law ensure the respect, and in some circumstance the primacy of provisions allowing for rights over other provisions imposing ‘obligations’: ‘[...] [T]echnically speaking, there is a conflict when two (or more) treaty instruments contain obligations which cannot be complied with simultaneously. [...] Not every such divergence constitutes a conflict, however.[...] Incompatibility of contents is an essential condition of conflict.’40 38

39 40

We emphasize here that the Appellate Body has held that the WTO Agreement is a ‘single undertaking’ and a ‘inseparable package of rights and disciplines’; See Brazil – Desiccated Coconut, WT/DS22/AB/R, 12 et seq. and Argentina – Footwear (EC) WT/DS121/AB/R, para. 81, respectively. Appellate Body Report on Guatemala – Cement I, WT/DS60/AB/R, para. 60. Karl (1984), 468; see also Jenks (1953), 425 et seq. For, in such a case, it is possible for a State which is a signatory of both treaties to comply with both treaties at the same time. The presumption against conflict is especially reinforced in cases where separate agreements

Comments on Joost Pauwelyn’s Paper

75

‘[…] a conflict of law-making treaties arises only where simultaneous compliance with the obligations of different instruments is impossible. [...] There is no conflict if the obligations of one instrument are stricter than, but not incompatible with, those of another, or if it is possible to comply with the obligations of one instrument by refraining from exercising a privilege or discretion accorded by another. […] The presumption against conflict is especially reinforced in cases where separate agreements are concluded between the same parties, since it can be presumed that they are meant to be consistent with themselves, failing any evidence to the contrary.’41 In taking into account explicit ‘rights’ provided for (within a treaty or in another treaty), one may use the lex specialis derogat generalis (lex specialis) principle of interpretation which favours the application of a more specific provision over a general one. Therefore, it may emerge from the intention of the parties and in the application of the lex specialis principle that a State exercises an express and more specific right provided for in an earlier or later treaty, albeit appearing inconsistent with a subsequent or earlier treaty obligation drafted in general terms. Contrary to Pauwelyn’s argument, no conflicts exists between Article XX of GATT and the basic market access provisions of the GATT (Articles I, II, III and XI). The Appellate Body has stated that, in assessing the interpretation and application of Article XX, there is a need to maintain a ‘balance’ between these provisions and has referred to this ‘balance’ in terms of weighing the ‘rights and obligations’ of Members under both sets of these provisions.42 In increasing the number of possible conflicts, Pauwelyn gives the impression that the identification of a conflict or several conflicts would facilitate the adjudicatory duties of panels and the Appellate Body and ensure an objective resolution of the matter and the respect of international law. In expanding the jurisdiction of the

41 42

are concluded between the same parties, since it can be presumed that they are meant to be consistent with each other, failing any evidence to the contrary; see also Vierdag (1988), 100; Jennings / Watts (1992), 1280; Fitzmaurice (1957), 237; and Sinclair (1984), 97. Jenks (1953). See the Appellate Body Report on US – Shrimp, WT/DS58/AB/R, para. 142.

76

Gabrielle Marceau / Anastasios Tomazos

WTO adjudicating bodies, and in encouraging them to resolve all potential conflicts, Pauwelyn asks WTO panels and the Appellate Body to resolve issues that Members have not been able to reconcile (or may not want to reconcile). Pauwelyn seems to suggest that in assessing potential conflicts, WTO panels and Appellate Body will be able to resolve the matter at issue more appropriately than if it limits itself to interpreting the WTO provision coherently with other regimes of international law. This cannot be correct. Even in a rare occasion of conflict between two treaty provisions, determining which of the two provisions supersedes the other is not a simple exercise. The WTO adjudicating body would need to interpret the other treaty and decide whether the WTO Member invoking it has complied with the procedural and substantive provision of that other treaty. But, what if, applying Pauwelyn’s treaty conflict rules, a WTO panel reaches the conclusion that the nonWTO treaty should prevail, but that the WTO Member invoking such non-WTO treaty has not complied with procedural requirement(s) of that other treaty. Should the WTO adjudicating body rule that the non-WTO provision prevails and that there is no WTO violation, despite that Member's non-compliance with the procedural requirement(s) of the other treaty? In our view, Pauwelyn’s approach creates more problems than solutions. If the WTO applicable law cannot be interpreted so as to avoid conflict of norms with a provision of another treaty, WTO adjudicating bodies would not be able to enforce non-WTO provisions or give them direct effect in the WTO applicable law, if the outcome was that the superseding provision added to or diminished or even amended the rights and obligations in the covered agreements. Setting aside jus cogens, States do have the right to create judicial and remedy systems that would not be able to enforce all international obligations of the same States. Consequently, these non-WTO norms are binding on the same States (also WTO Members), and if States violate them, they will be held responsible, but in another jurisdiction (and rules on State responsibility may apply). Both systems of State responsibility operate in parallel; the WTO simply being a specific one. Our approach simply reflects the present state of international law which is characterized by an imperfect coherence between international legal regimes and jurisdictions exists.

Comments on Joost Pauwelyn’s Paper

77

Elsewhere,43 Pauwelyn has also argued that Members are presumed to have accepted the binding effect and the direct application of general international law into the WTO law. According to Pauwelyn, general international law can fill in the gaps of the WTO treaty and provisions of non-WTO treaties that are binding on the parties can be invoked in defense of alleged WTO violations. If the relevant customary conflict rules demonstrate that a non-WTO rule must prevail, ‘then the WTO rule cannot be applied (and the defendant wins)’. For Pauwelyn, this would not result in ‘requiring the WTO panel to judicially enforce the other rule of international law’ and ‘the panel would not be ‘diminishing’ the rights of the complainants, as the complainant would have agreed to these conflicting rules in the first place’. In addition to being contrary to the wording of the DSU, Pauwelyn’s suggestion would require WTO adjudicating bodies to interpret the other treaty to decide on its compliance or violation by the concerned WTO Member and to draw legal consequences over and above the provisions of the WTO treaty. As previously stated, the WTO adjudicating bodies are not courts of general jurisdiction and they cannot interpret and apply all treaties involving WTO Members. The covered agreements are explicitly listed and it cannot be presumed that Members wanted to provide the WTO remedial system to enforce obligations and rights other than those listed in the WTO treaty. If they were to allow a non-WTO provision to supersede and set aside a WTO provision and therefore, give it legal effect and enforce a non-WTO provision in superseding a WTO provision, they would be adding or diminishing the WTO covered agreements (or even amending it). There is no evidence whatsoever to even suggest that during the Uruguay Round the drafters of the WTO treaty ever wanted to provide non-WTO norms with direct effect into WTO law and allow Members to benefit from free use of the WTO remedial mechanism to enforce rights and obligations other than those of the WTO treaty. In our view, Pauwelyn’s position disregards the fact that no perfect coherence between international legal regimes exists and there is not yet a perfect organisational coherence of international jurisdictions. When faced with imperfect sharing of international responsibilities, Pauwelyn ironically elevates WTO panels and the Appellate Body into a world dictator of sorts. Moreover, Pauwelyn 43

Pauwelyn (2001).

78

Gabrielle Marceau / Anastasios Tomazos

overlooks that a solution to any conflict between WTO rules and other non-WTO rules should also take into account the legal fact that sovereign States decided to agree to trade rules and an enforcement mechanism different from that of other international obligations. Whether this is good or bad, is a question beyond the scope of this paper. Importantly, genuine situations of conflict of norms will occur very rarely and through good faith interpretation, WTO law and other provisions of international law can generally be applied harmoniously and effectively. Having said that WTO panels and the Appellate Body cannot interpret and enforce non-WTO other than to the extent necessary to interpret and apply WTO provisions, one should not underestimate the coherence that already exists between WTO law and the other systems of international law. It is our suggestion that conflict of norms should continue to be construed narrowly. A broad definition of conflict would lead to a result in which a third party (an adjudication body or a treaty interpreter) is conferred the power to set aside provisions that have been voluntarily negotiated by States. IV. Conclusion It is important to emphasize that States must at all times comply with all their international obligations and respect the rights of other States. The issue of WTO Members' State responsibility for a violation of an international obligation or their obligation to respect other States’ rights under treaties other than the WTO is not a matter for the WTO dispute settlement. It is important to recall that WTO Members remain responsible for the consequence of violations of their international obligations. In situations where a nonWTO treaty provision could potentially supersede a WTO provision, but could not be directly applicable into the WTO legal system or enforced by WTO adjudicating bodies, the State invoking the non-WTO provision would still be able to invoke the application of the general international law rules on State responsibility against other States (also WTO Members) or other relevant systems of law for the violation of the non-WTO provisions.44 44

Paragraph 5 of Article 30 of the Vienna Convention on the Law of Treaties provides that: ‘Paragraph 4 [lex posterior] is without prejudice to article 41, or to any question of the termination or suspension

Comments on Joost Pauwelyn’s Paper

79

At the moment, in international law, there does not seem to be any complete co-ordination of systems of international law and there are several treaties on multiple and overlapping matters. A single measure may violate a treaty and be consistent with another one. It is thus possible to envisage a situation where an international tribunal would reach a conclusion that a measure (that is also [part of] a WTO measure) is inconsistent with a (non-WTO) treaty, while the WTO adjudicating body may reach the conclusion that the same measure is consistent with the WTO treaty. As required under international law, the measure would have to be modified to comply with the non-WTO treaty law while continuing to be compatible with WTO law; and most of the time this should be feasible. The ruling of the WTO adjudicating body would only relate to the WTO aspects of the measure and would not affect or deal with the compatibility of the same measure with the non-WTO treaty. Various suggestions have been offered to try and ‘constitutionalize’ international courts and jurisdictions. We still believe that States’ multiple international rights and obligations can be interpreted and applied harmoniously and coherently. In this regard, we share the following perspective, that seems the most astute in the present state of international law: ‘With the greatest respect to the past two Presidents of the International Court, I do not share their view that the model of Article 234 (the renumbered Article 177) of the Rome Treaty provides an answer. It is simply cumbersome and unrealistic to suppose that other tribunals would wish to refer points of general international law to the International Court of Justice. Indeed, the very reason for their establishment as separate judicial instances militates against a notion of intra-judicial reference. The better way forward, in my view, is for us all to keep ourselves well informed. Thus the European Court of Justice will want to keep abreast of the case law of the international Court, particularly when it deals with treaty law or matters of customary international law; and the International Court will want to make sure it fully understands of the operation of a treaty under article 60 or to any question of responsibility which may arise for a State from the conclusion or application of a treaty, the provisions of which are incompatible with its obligations towards another State under another treaty.’

Gabrielle Marceau / Anastasios Tomazos

80

the circumstances in which these issues arise for its sister court in Luxembourg [...].’45 In conclusion, despite some criticisms offered on our colleague’s paper, it should be borne in mind that we greatly appreciate yet another of his outstanding contributions to the field of public international law and WTO law as well as the relationship between both. In the final analysis, however, Joost’s assertion that the purpose of his paper is to take the relevance of non-WTO law before WTO panels a ‘step further’ is instructive, inasmuch as WTO panels should be reluctant, at present, to take that ‘step further’.

References James J. Fawcett (1995), Declining Jurisdiction in Private International Law, Oxford (Oxford University Press) 1995. Gerald Fitzmaurice (1957), The Law and procedure of the International Court of Justice, in: British Year Book of International Law 34 (1957), 237. Roselyn Higgins (2003), The ICJ, The ECJ, and the Integrity of International Law, International and Comparative Law Quarterly 52 (2003) No. 1, 1-20. Wilfred Jenks (1953), The Conflict of Law-Making Treaties, in: British Year Book of International Law 30 (1953), 401-453. Robert Jennings / Arthur Watts (eds.) (1992), Oppenheim’s International Law, London (Longman) 91992. Wolfram Karl (1984), Conflicts Between Treaties, in: Rudolf Bernhardt (ed.), Encyclopedia of Public International Law Vol. 7, Amsterdam / London / New / Tokyo (North Holland) 1984, 468-473. Vaughan Lowe (1996), Res Judicata and the Rule of Law in International Arbitration, African Journal of International Law 8 (1996), 38-50. Vaughan Lowe (1999), Overlapping jurisdiction in international tribunals, in: Australian Year Book of International Law 20 (1999), 191-204. 45

Higgins (2003), 20.

Comments on Joost Pauwelyn’s Paper

81

Gabrielle Marceau (1997), Dispute Settlement Mechanisms, Regional or Multilateral Agreement: Which One is Better?, in: Journal of World Trade 21 (1997) No. 3, 169-179. Donald M. McRae (1996), The contribution of international trade law to the development of international law, in: Recueil des cours 260 (1996), 99-238. Donald M. McRae (2003), Claus-Dieter Ehlermann’s Presentation on ‘The Role and Record of the Dispute Settlement Panels and the Appellate Body of the WTO’, in: Journal of International Economic Law 6 (2003) No. 3, 709-717. Joost Pauwelyn (2001), The Role of Public International Law in the WTO. How Far Can We Go?, in: American Journal of International Law 95 (2001) No. 3, 535-578. Emmanuel Roucounas (1987), Engagements parallèles et contradictoires, in: Recueil des cours 206 (1987-VI), 9-288. Takao Sawaki (1979-80), Battle of Lawsuits – Lis Pendens in International Relations, in: Japanese Annual International Law 23 (1979–80), 17-29. Ian M. Sinclair (1984), The Vienna Convention on the Law of Treaties, Manchester (Manchester University Press) 21984. E. W. Vierdag (1988), The Time of the ‘Conclusion’ of a Multilateral Treaty: Article 30 of the Vienna Convention on the Law of Treaties and Related Provisions, in: British Year Book of International Law 59 (1988), 100.

Erich Vranes*

Comments on Joost Pauwelyn’s Paper: ‘How to Win a WTO Dispute Based on Non-WTO Law?’ I. Introduction: Areas of Disagreement II. The Definition of Norm Conflict III. Jurisdiction and Applicable Law in WTO Proceedings A. The Positions Juxtaposed B. Legal Starting-Points and Central Question C. Substantive Law: Restricted Capacity of Panels and WTO Members? D. WTO Dispute Settlement: An (Absolutely) Exclusive Forum? IV. Conclusions References

83 84 89 89 92 93 95 97 98

I. Introduction: Areas of Disagreement The following comments concentrate on the areas of disagreement in the positions of Joost Pauwelyn on the one side and Gabrielle Marceau and Anastasios Tomazos on the other. As Joost Pauwelyn and Gabrielle Marceau are two of the main proponents in the debate on the relationship between international law and the WTO system, in dealing with their views one has to address several of the core issues raised in this discussion so far. In order to better understand these issues and the respective positions presented, it appears appropriate to take some of their earlier writings into perspective. Joost Pauwelyn’s position can be summarized quite succinctly. According to him, one has to distinguish jurisdiction and applicable law. Against this background, he argues first that a panel’s jurisdiction can be taken away through international agreements that do not form part of WTO law. Similarly, Pauwelyn submits that substantive rules of international law conflicting with WTO law can be *

Europainstitut, Vienna University of Economics and Business Administration.

84

Erich Vranes

invoked as a defence even within WTO proceedings. Both contingencies depend on the principles of international law, in particular Articles 41 and 30 of the Vienna Convention. Underlying Pauwelyn’s approach is a wide definition of conflict. Although not extensively addressed by Pauwelyn in his present paper, Marceau and Tomazos deal with this fundamental issue in more detail. According to Pauwelyn, explicit permissions to restrict trade in other treaties may prevail over WTO law, depending on the rules of international law. Importantly, it follows that Article XX of the GATT and other WTO exceptions, WTO provisions incorporating non-WTO norms by reference, and waivers are not the unique or primary links to non-WTO international law: a WTO Member which is required or permitted to restrict trade by a nonWTO treaty need not comply with clauses like Article XX (and need not employ the least trade-restrictive means e.g. under Article XX(b)), if WTO law is superseded by a given non-WTO treaty. On all of these issues, there is disagreement between Pauwelyn on the one hand and Marceau and Tomazos on the other. We will address these questions in consecutive order, starting with that of the appropriate definition of norm conflict, which underlies the whole topic. II. The Definition of Norm Conflict In contrast to Pauwelyn, Marceau and Tomazos argue that the notion of conflict of norms should be defined narrowly. In their view, a conflict exists only where ‘a provision of the WTO Agreement mandates an action that a provision of another treaty prohibits or vice versa, when a provision of another treaty mandates an action that the WTO Agreement prohibits.’1 This is in line with what arguably constitutes the predominant position in writings on international law.2 While Marceau and Tomazos cite the Appellate Body 1 2

Cf Marceau / Tomazos in this volume at 77 (italics added). Cf Jenks (1953), 426; Marceau (2001), 1081. Several other authors, among them Czaplinski / Danilenko (1990), 12-13; and most recently Wolfrum / Matz (2003), 4 have also opted for narrow definitions of conflict; yet it is not always clear whether in doing so they are aware of the particular problem of conflicts between permissive and prescriptive or prohibitive norms. Moreover, the proposed definitions are often open to different interpretations. Thus Klein (1962), 555, Wilting (1994), 2-12 and Kelsen (1979), 99-100 have been under-

Comments on Joost Pauwelyn’s Paper

85

report on Guatemala – Cement I in their support,3 it must be emphasized that a different position has been taken by the panel in the Bananas III case.4 It is indispensable briefly to recall some fundamentals of the theory of norms, in order to be in a position to assess the diametrically opposed positions of Marceau and Tomazos on the one hand and Pauwelyn on the other. It has already been pointed out by Bentham in his Of Laws in General that the multitude of legal provisions and complex legal concepts (such as ‘competence’, ‘property’ etc) can be reduced to (sets of) ‘complete norms’, that is norms of conduct, namely prohibition, obligation, positive and negative permission.5 In this regard, deontic logic6 and the imperative theory of law are congruent. A norm of conduct consists of two parts, that is to say a deontic operator which expresses the obligation, prohibition, or permission, and a descriptive proposition which can be any conduct. All of these ‘basic units’ of legal thinking are interdefinable through negations. Thus, if the prohibition of a given conduct is negated, this same conduct is permitted, and vice versa. In other words, the prohibition of a given conduct constitutes the contradictory opposite of the permission of this conduct (‘non-prohibition’, positive permission). The same is true for the permission to forbear from adopting a given conduct and the obligation to adopt this con-

3

4

5 6

stood as giving strict definitions, excluding any incompatibility of permissions and prescriptive norms. In relation to Kelsen, in particular, this interpretation appears problematic, as he adopted a wide definition in his General Theory of Norms, where he expressly stated ‘... that one cannot deny that a permission and a prescription mutually exclude each other’ (Kelsen (1979), 79). A wide definition has recently been adopted by Weiss (2003), 201 et seqq., who concurs with the 1997 Bananas panel report (cf in the following). Cf WTO Appellate Body, Guatemala–Anti-Dumping Investigation Regarding Portland Cement from Mexico, WTO Doc. WT/DS60/AB/R, in particular at paras 65-66 (Nov. 5, 1998). Cf WTO panel report, European Communities-Regime for the Importation, Sale and Distribution of Bananas, WTO Docs. WT/DS27/R/ECU/GUA/HON/MEX/USA, paras 7.158-7.159 and fn 792 (September 27, 1997). Cf Bentham (1970), 93-109 and 156-183. On this cf e.g. Weinberger (1989), 228 et seqq.; Lenk (1974), 198 et seqq.; Alexy (1994), 182-194; Adomeit (1986), 26-29 and 82-86.

86

Erich Vranes

duct: negating this obligation yields a permission of contradictory content (‘non-command’, negative permission), and vice versa. Prohibition and obligation are also interdefinable, if the conduct which is regulated (the descriptive proposition) is negated.7 All of these constellations – permission vs prohibition, permission vs obligation, prohibition vs obligation etc – are recognized as conflicts in legal theory. The first two are normally referred to as contradictory conflicts; the last (prohibition vs obligation) is termed contrary conflict. In short, there is a conflict ‘if a given behaviour appears in abstracto or in concreto as prescribed and not prescribed, or as prohibited and not prohibited, or even as prescribed and prohibited’.8 Importantly, legal definitions, exceptions (such as Article XX of the GATT), norms referring to other norms or setting out legal fictions etc are merely parts of the antecedent (the if-clause) of complete norms.9 Thus, a complete norm is equivalent to the complete expression of the legislator’s (or the contracting parties’) will in respect of a given conduct or class of conduct. It therefore varies, in extent and complexity, from a simple command to a multitude of legal provisions, depending in particular on the degree of a norm’s generality,10 and the number of exceptions, definitions etc which have to be taken into account as preconditions by the decisionmaker called upon to apply the norm, or the conflicting norms. As indicated, Pauwelyn has submitted that the notion of conflict of norms should be defined widely so as to encompass conflicts between prohibitions and obligations (prescriptive norms) on the one side and permissions on the other, a position which in fact is in line with the understanding in legal theory. Marceau and Tomazos in contradistinction only accept those constellations as 7

8

9 10

Let us assume the conduct in question is ‘to stay in Vienna’. A given norm may prohibit a person from ‘staying in Vienna’. The negation of this conduct is ‘not staying in Vienna’. If this person is prohibited from ‘not staying in Vienna’, the person is actually under an obligation ‘to stay in Vienna’. Thus, negating the descriptive part of a prohibitive norm yields the contrary obligation and vice versa. Engisch (1977), 162 (‘Ein Verhalten [erscheint] in abstracto oder in concreto zugleich als geboten und nicht geboten oder als verboten und nicht verboten oder gar als geboten und verboten’). Cf e.g. Hart (1971), 59-66; Hart (1992), 26-48; Thon (1878) 2 et seqq., 325, 338-339 and passim; Röhl (1995) 228. Bentham (1970) 159.

Comments on Joost Pauwelyn’s Paper

87

constituting a conflict of norms that we have just referred to as contrary conflicts.11 Obviously, it would be unconvincing to confirm or reject either of these definitions merely on the ground that it is not in line with deontic logic, WTO jurisprudence or the ‘prevailing opinion’ in international law doctrine. What then are decisive arguments? In order to answer this question, it is crucial to distinguish between two types of definitions, that is analytical definitions and stipulative definitions. An analytical definition analyses and explains the actual way a term is used in a given language.12 It is an assertion concerned with past or present usage and has truthvalue.13 A stipulative definition on the other hand establishes the meaning of a word. It takes the form of a command or proposal on the meaning of a given term.14 This does not imply, however, that the author of a stipulative definition enjoys unrestricted discretion; he is bound in particular by teleological considerations. For purposes of legal doctrine this means above all that a stipulative definition must adequately fit into the legal system within which it is meant to operate.15 Turning back to the problem of how to define ‘conflict of norms’, it is arguably impossible to derive a complete and uniform analytical definition from international legal texts. Hence, the definition to be introduced can only be stipulative and must, therefore, be teleologically adequate. The telos of norms is to regulate behaviour. It follows that if a given conduct is at the same time permitted and prohibited, it is not unequivocally but contradictorily regulated from the viewpoint of the addressee of these norms. The same is true for the other constellations discussed above. In other words, if attaining this telos is impaired by a permissive norm incompatible with a prescriptive norm, one should recognize these norms as conflicting. 11

12 13 14 15

As pointed out supra, in their view there is only a conflict of norms if ‘a provision of the WTO Agreement mandates an action that a provision of another treaty prohibits or vice versa, when a provision of another treaty mandates an action that the WTO Agreement prohibits.’ Klug (1982), 93; Weinberger (1989), 360-361. Dubislav (1981), 131. Cf Robinson (1968), 19, 21, 59-92; Klug (1982), 99-109; Weinberger (1989), 360-361. Klug (1982), 93; Weinberger (1989), 360.

88

Erich Vranes

Thus, although the contrary position prevailing in international legal doctrine which is also held by Marceau and Tomazos cannot be labelled as ‘incorrect’, it is teleologically inadequate. Above all, this approach disregards the fundamental interpretive requirement of investigating the actual consent of the contracting parties, in particular in that it prevents the lex posterior and lex specialis principles from coming into play when a permission collides with a prohibition or obligation.16 This approach therefore passes over relevant objective criteria such as the timing of treaties and the specificity of norms. Ironically, the most pertinent critique has been advanced by the most distinguished advocate of this view: Jenks had pointed out himself that incompatibilities between permissions and prescriptive norms may ‘from a practical point of view be as serious’ as those between incompatible obligations, as such constellations ‘may render inapplicable provisions designed to give one of the divergent instruments a measure of flexibility of operation which was thought necessary to its practicability’.17 This remark has not hindered his definition from finding its way into recent works and, more strikingly, even into WTO jurisprudence.18 Still, it should in itself be sufficient to cast into doubt the adequacy of the narrow definition of conflict, which therefore should be rejected on this ground and those just outlined. A concluding caveat must be issued in this context: international lawyers, in particular, are familiar with the notion that it is necessary to distinguish conflicts according to the configurations of the parties involved. Thus, conflicts of norms may arise within one treaty, between norms contained in two treaties with identical parties (AB/AB, ABCD/ABCD etc), between norms contained in two 16

17 18

It must be conceded that Gabrielle Marceau has held in one of her earlier writings on this subject that ‘[t]o take into account explicit ‘rights’ provided in another treaty, one should refer to the lex specialis principle of interpretation’ (Marceau (2001) 1086). The lex specialis principle, at least as employed by Marceau in her reasoning (ibidem 1086), constitutes a conflict rule; this is at odds with Marceau’s overall position that there is no conflict in such constellations. Jenks (1953), 426-427. Cf Marceau (2001); WTO panel report, Indonesia – Certain Measures Affecting the Automobile Industry, WTO Docs. WT/DS54/R, WT/DS59/R, WT/DS64/R, at footnote 649 (July 23, 1998); WTO panel report, Turkey–Restrictions on Imports of Textile and Clothing Products, WTO Doc. WT/DS34/R, paras 9.92-9.95 (May 31, 1999).

Comments on Joost Pauwelyn’s Paper

89

treaties with decreasing or increasing membership (ABCD/AB, ABC/ABCD etc) or between norms contained in two treaties with overlapping parties (AB/AC, ABCD/ABEF etc). On the one hand, the principles of lex specialis and lex posterior come into play if there is identity of parties (ie if states A and B have concluded two bilateral agreements that give rise to a conflict of norms). On the other hand, and more importantly, these principles may also apply when the contracting parties to two agreements overlap (e.g. if states A and B have concluded a bilateral agreement, while they are also parties to a multilateral agreement concluded by A, B, C and D); in the latter case, whether the more special or later-in-time bilateral treaty may be invoked depends on the pacta tertiis principle: in case there is a conflict, say from the perspective of A, between the provisions of both agreements, A is prohibited from invoking the AB treaty as a lex specialis or lex posterior in dispute settlement proceedings under the ABCD treaty, if the rights of the third parties C and D would thereby be infringed. The same is true if A is party to two bilateral agreements with B and C: if state A is under an obligation vis à vis state B to adopt a given behaviour b under one agreement, and if state A is obligated at the same time towards state C to adopt the irreconcilable conduct c under the other agreement, then there is a conflict from the viewpoint of state A. But state A cannot invoke its treaty with state B as a lex specialis or lex posterior to justify breaching its agreement with state C. The same holds true mutatis mutandis for its relation with state B: in such constellations, the pacta tertiis principle prevents state A from invoking the specificity and timing of the conflicting treaty.19 These considerations are essential for the following issues. III. Jurisdiction and Applicable Law in WTO Proceedings A. The Positions Juxtaposed Pauwelyn has argued essentially that the respondent WTO Member can invoke non-WTO rules of international law even within WTO proceedings, if certain conditions are fulfilled. First, such non-WTO rules must be binding on both parties (these are the ABCD/AB and ABCD/ABEF constellations referred to above). Second, the establishment of such rules by way of an international agreement must not be prohibited by the WTO treaty (cf Art 41(1)(b) VCLT). 19

Cf Art 30(4) lit b VCLT.

90

Erich Vranes

Moreover, these rules may not affect the rights or obligations of third parties (cf Art 41(1) (b)(i) VCLT). Finally, the non-WTO rule of international law must be shown to prevail over WTO law according to the established conflict principles of lex specialis and lex posterior. Pauwelyn argues that these principles also apply to the question whether non-WTO rules of international law have taken away a panel’s jurisdiction, the sole difference to issues of substantive law being that this question may have to be addressed ex officio. While Marceau and Tomazos have not addressed the issue of conflicts in substantive law which Pauwelyn has dealt with in his present paper, Gabrielle Marceau has held in widely noticed earlier publications that panels cannot take account of inter se modifications of WTO law for lack of a ‘constitutional capacity’ to do so. Moreover, Marceau has insinuated that even inter se modifications by WTO Members themselves may be impossible.20 In their present paper, Marceau and Tomazos essentially restate a stance recently taken by Kwak and Marceau,21 according to which ‘WTO adjudicating bodies always have the authority (and even the obligation) to examine claims of violations of WTO obligations’. This means, in the view of Kwak, Marceau and Tomazos, that WTO Members have an ‘absolute right to trigger the WTO dispute settlement mechanism ... [A] WTO Member cannot be asked (and arguably cannot even agree) to take its WTO dispute to another forum, even if that other forum appears to be more relevant or better equipped’.22 Incidentally, there is little guidance to be derived from GATT 1947 and WTO dispute settlement decisions, which so far do not appear consistent in the treatment of conflicting non-WTO rules of international law. While several GATT panel reports took quite restrictive stances,23 there are some indications – as rightly pointed 20

21 22 23

Marceau (2001), 1104-1105 (‘If the WTO obligations are always the same for all Members – and it can be argued they are – such bilateral modification of WTO rights and obligations may simply not be possible without affecting the rights of other third WTO Members’). Kwak / Marceau (2002), 12. Cf Marceau / Tomazos in this book; in the same sense Kwak / Marceau (2002), 12 (italics added). Cf e.g. the 1984 GATT panel report, US – Imports of Sugar from Nicaragua, adopted on 13 March 1984, L/5607, BISD 31S/67, in

Comments on Joost Pauwelyn’s Paper

91

out by Pauwelyn – in recent WTO decisions that non-WTO law might be recognized as prevailing over conflicting WTO rules even within WTO proceedings: the panels on India – Automobiles and Argentina – Poultry indicated obiter that they might take into account non-WTO agreements that take away a panel’s jurisdiction.24 However, these indications are not accompanied by explicit legal reasoning. The same is true of the Argentina – Footwear case, which has even given rise to diametrically opposite readings in the literature.25 Finally, the panel on Korea – Government Procurement stated that customary international law applies between WTO

24

25

which the panel confined itself to an examination of the claims ‘solely in the light of the relevant GATT provisions’ despite the fact that the US invoked non-GATT international law in its defence (para 4.1); see also the 1988 GATT panel report, Canada – Herring and Salmon, adopted on 22 March 1988, L/6268, BISD 35S/98, where the panel tersely noted in the last paragraph ‘Canada referred in its submission to international agreements on fisheries and the Convention on the Law of the Sea. The panel considered that its mandate was limited to the examination of Canada's measures in the light of the relevant provisions of the General Agreement. This report therefore has no bearing on questions of fisheries jurisdiction’; cf also the 1984 FIRA case, in which the Council decided ‘that it be presumed that the Panel would be limited in its activities and findings to within the four corners of GATT’ (cf the panel report, Canada – Administration of the Foreign Investment Review Act (FIRA), adopted on 7 February 1984, L/5504, BISD 30S/140, at para 1.4); these cases are discussed by Pauwelyn (2003) 456-459; Palmeter/Mavroïdis (1998) 411 refer to a ‘tendency of GATT panels to disregard public international law’ in their discussion of the 1994 US – Tuna case. The fact that GATT/WTO jurisprudence is not yet settled is also pointed out by Bartels (2001), 509; on these issues cf also Reinisch (2003), 449 et seqq.; Hilpold (2002), 46 et seqq. Cf the WTO panel report, India – Measures Affecting the Automotive Sector, WT/DS146/R, adopted 5 April 2002, at para 7.116; and the WTO panel report, Argentina – Definitive Anti-Dumping Duties on Poultry from Brazil, WT/DS241/R, adopted on 19 May 2003, para 7.38. Appellate Body report in Argentina – Footwear, WT/DS56/R, adopted on 22 April 1998, paras 65-72; while Pauwelyn cites this report in support of his argument, Trachtman reads this case as implying the opposite, namely that international law can not supersede WTO law in WTO dispute settlement; cf Trachtman (1999), 343.

92

Erich Vranes

Members ‘to the extent there is no conflict’; it did not indicate, however, that conflicting non-WTO international could prevail over WTO law in WTO proceedings.26 B. Legal Starting-Points and Central Question Regarding the issues raised by Pauwelyn, Marceau and Tomazos, there are two evident starting-points which, however, often do not receive sufficient attention in the literature. On the one side, despite its economic importance, the WTO treaty is a normal international agreement. Therefore, the questions delineated in the preceding section have to be assessed according to the rules of general international law. This means in particular that inter se modifications of the WTO treaty are permissible under the established principles of general international law, in particular the pacta tertiis principle as laid down in Article 41 VCLT, unless it can it can be shown that WTO Members have explicitly or implicitly contracted out from these rules. The latter proviso constitutes the core of the problem, as will be shown in the following. On the other side, it is firmly established in the jurisprudence of international tribunals and in WTO dispute settlement practice that adjudicating bodies have the implied jurisdiction to decide all issues inextricably linked to the exercise of the judicial function, including the competence to decide on their own competence (Kompetenzkompetenz).27 This arguably includes the issue of whether a panel’s jurisdiction has been taken away through modifications of the treaty by which the panel is established.28 Therefore, the central question is whether these two starting points have been modified expressly or implicitly by the WTO 26

27

28

Cf the panel report in Korea – Measures Affecting Government Procurement, WT/DS163, adopted on 19 June 2000, at para 7.96. These decisions are more extensively discussed than is possible here in Pauwelyn’s paper supra and in Pauwelyn (2003), 478 et seqq.; see also Bartels (2001) for a discussion of earlier rulings. Cf the Appellate Body report, European Communities – Regime for the Importation, Sale and Distribution of Bananas, WT/DS27/AB/R, para 142 (confirming the reasoning of the panel report at paras 7.26 ff); Appellate Body report, Korea – Dairy, WT/DS98//AB/R at para 123; see also Pauwelyn (2003), 447-449 with further references regarding WTO and ICJ practice and Vranes (2003), 48 et seqq. regarding pertinent WTO decisions. In the same sense Pauwelyn (2003) ibidem.

Comments on Joost Pauwelyn’s Paper

93

Members. As there are no other relevant indications, such contracting-out from rules of general international law must be discernible from the WTO agreements themselves. Since such a contracting-out constitutes the exception to the rule, advocates of the existence of this exception bear the argumentative burden of having to advance sufficient reasons in support of their view. Otherwise, there is a strong presumption for the standard rule, which implies (i) that WTO law can be modified under the conditions ensuing from the principles of general international law, and (ii) that WTO adjudicating bodies have to take account of such modifications, both in merito and when their jurisdiction is affected. This would imply, in other words, that WTO adjudicating bodies, when faced with a given non-WTO law rule conflicting with WTO law, have to ascertain whether the WTO Members bound by this rule have intended to modify WTO law29 and whether this is lawful. C. Substantive Law: Restricted Capacity of Panels and WTO Members? Marceau has argued on the one hand that panels and the Appellate Body ‘do not seem to have the constitutional capacity to reach any standard recommendations in situations where another treaty provision has superseded … a WTO provision’.30 Marceau concludes that a panel may decline jurisdiction in such cases, or should ‘[a]t best... declare that in international law the WTO provision has been superseded by another treaty’s provision but that the WTO dispute settlement mechanism is prohibiting from acting further on that specific claim’.31 On the other hand, Marceau has taken a quite restrictive stance regarding the capacity of WTO Members to modify WTO law. In her view, inter se modifications in line with Article 41 VCLT may not be permissible, since ‘WTO obligations are always the same for all WTO Members’.32 In her approach, there29 30 31 32

This is where in cases of doubt the lex specialis and lex posterior principles come into play. Marceau (2001), 1104. Ibidem, 1107-1108. Cf again Marceau (2001), 1105 (‘If the WTO obligations are always the same for all Members – and it can be argued they are – such bilateral modification of WTO rights and obligations may simply not be possible without affecting the rights of other third WTO Members’).

94

Erich Vranes

fore, provisions like Article XX of GATT become central, as she expressly admits.33 In dealing with this view, which is opposed to the position presented by Pauwelyn and which appears to be preponderantly based on Article 3.2 of the DSU,34 it is appropriate to recall the actual wording of this provision: ‘The dispute settlement system of the WTO is a central element in providing security and predictability to the multilateral trading system. The Members recognize that it serves to preserve the rights and obligations of Members under covered agreements, and to clarify the existing provisions of those agreements in accordance with customary rules of interpretation of public international law. Recommendations and rulings of the DSB cannot add to or diminish the rights and obligations provided in the covered agreements.’ Article 19.2 adds: ‘In accordance with paragraph 2 of Article 3, in their findings and recommendations, the panel and Appellate cannot add to or diminish the rights and obligations provided in the covered agreements.’ Articles 3.2 and 19.2 can arguably be read in two ways. These provisions can, first, be read, and are in fact usually read, as setting out limits on interpretation.35 Looked at from this angle, it would be difficult to infer from the clause ‘add to or diminish’ that panels must not, in the exercise of their Kompetenzkompetenz, take account of modifications of WTO law that are brought about lawfully, in terms of international law, by WTO Members themselves. In other words, on this reading, it would be difficult to sustain that the Kompetenzkompetenz of WTO adjudicating bodies has been restricted by Articles 3.2 and 19.2 of the DSU. 33 34

35

Ibidem, 1107. By way of introduction of the pertinent considerations, Marceau refers to Articles 1.1, 4.2, 4.4, 7 and 11 as well as to Articles 3.2 and 19.2 of the DSU (cf ibidem 1102). The ensuing inferences appear to be based on Article 3.2 and the wording ‘add to or diminish’ which forms part of Articles 3.2 and 19.2 of the DSU (cf ibidem, 1102 ff). Cf also Bartels (2001), 507, who however submits that these provisions limit the applicable law as well.

Comments on Joost Pauwelyn’s Paper

95

However, and this brings us to the second possible reading, it could be argued that the third sentence can not only be read in the context of the second part of the second sentence of Article 3.2, which refers to the interpretative task of adjudicating bodies, but also in the context of the first part of the second sentence of Article 3.2, according to which WTO Members ‘recognize that [the WTO dispute settlement system] serves to preserve the rights and obligations of Members under covered agreements’. In other words, on the first reading, the question was whether the compétence de la compétence of WTO adjudicating bodies has been limited. On the second reading the question is: does it follow from this context of the clause ‘must not add to or diminish’ that WTO Members have restricted their own capacity to modify WTO law, and that, consequently, WTO adjudicating bodies must not take account of inter se modifications in conflict with WTO law? This second reading, which apparently also finds support in academic writing,36 has to be rejected: it would imply that WTO law would be exempt from normal international law rules on treaty modification, thereby making it a sort of lex superior. Moreover, in certain cases the pivotal link to other international law would then be clauses of the type of Article XX of the GATT. This would imply that non-economic policies could in certain cases only be pursued by the least trade-restrictive means possible (e.g. if Article XX(b) of the GATT is applicable). It has to be emphasized, in contrast, that there are no indications in these technical DSU provisions that WTO Members intended to restrict their ability to undertake inter se modifications more than would be permitted under established international law principles, in particular the pacta tertiis, lex specialis and lex posterior principles.37 D. WTO Dispute Settlement: An (Absolutely) Exclusive Forum? Concerning the jurisdictional issues addressed by Pauwelyn, Marceau and Tomazos regard must also be had to Article 23 of the DSU. According to Article 23.1 of the DSU, WTO Members ‘shall have recourse to’ the DSU in disputes concerning WTO issues. Pursuant to Article 23.2, WTO Members must not unilaterally determine that their WTO rights have been infringed. Article 23 is 36 37

Cf Marceau’s view just outlined, in particular at 1104-1105. Similar doubts are also expressed by Pauwelyn (2003), 352 et seqq., 354.

96

Erich Vranes

generally referred to as an ‘exclusive jurisdiction’ clause.38 The question is, however, whether Article 23 is to be understood as even excluding any inter se modifications by WTO Members themselves, as has in fact been submitted by Kwak / Marceau and Marceau / Tomazos,39 or at least as excluding the power of WTO adjudicating bodies to take account of such conflicting international law norms. According to these authors, Article 23 has the effect that a WTO complainant ‘arguably cannot even agree to take its WTO dispute to another forum, even if that other forum appears to be more relevant’.40 In the view of Kwak and Marceau, Article 23 ‘reflects the clear intention of WTO Members to ensure that WTO adjudicating bodies can always exercise exclusive jurisdiction’.41 Moreover, ‘[i]n order to change this, Members would have to negotiate amendments to Article 23’.42 The arguments submitted by Kwak / Marceau and Marceau / Tomazos appear confined too strongly to the ‘four corners’ of WTO law. Their stance that, pursuant to Article 23, the alleged exclusivity of the WTO dispute settlement system can only be changed by explicit amendments amounts to claiming that WTO Members have themselves reduced their capacity to modify WTO law. It is questionable whether one can infer this from Article 23, since the wording of this provision, in particular paragraph 2, as well as the negotiating history underline that the parties to the negotiations were more concerned with the prevention of unilateralism43 than with possible (lawful) inter se modifications of Article 23 e.g. through the bi- or plurilateral establishment of overlapping international procedures, if these do not impinge on the rights of third

38

39 40 41 42 43

Cf e.g. Shany (2003), 183 with further references; see also the panel report US – Sections 301-310 of the Trade Act of 1974, WT/DS152/R, adopted on 27 January 2000, at para 7.43; GonzalezCalatayud / Marceau (2002), 281; Kwak / Marceau (2002), 8 and passim. Cf in the following. Kwak / Marceau (2002), 12; see also Marceau / Tomazos in this volume. Kwak / Marceau (2002), 8; cf the similar submissions of Marceau / Tomazos in this volume. Ibidem, 8. Shany (2003), 185.

Comments on Joost Pauwelyn’s Paper

97

WTO Members.44 The fact that the reading of Kwak and Marceau is improbable is also underlined by the modest integrative ambitions of the WTO, in which ‘the value of jurisprudential coherence in international trade is important, but not as crucial as in regional economic integration’; therefore, in the words of Shany, the establishment of a ‘watertight closed legal subsystem’ does not appear necessary in the case of the WTO.45 Hence, the arguments of Marceau, Kwak and Tomazos arguably do not suffice to call into question the proposition that WTO Members themselves are capable under international law to establish competing fora as well as to exclude the jurisdiction of WTO adjudicating bodies inter se, as long as this is done in accordance with international law. There remains the related question whether one has to interpret Article 23 as meaning that WTO adjudicating bodies have lost their implied competence to take into account modifications of WTO law that take away their jurisdiction. If this were true, the WTO Members would have foregone the possibility to raise a pertinent objection in WTO proceedings. Consequently, WTO Members would have to go through the whole of WTO proceedings and would only be in a position to argue ex post (ie outside of WTO legal proceedings) that the WTO adjudicating bodies are not competent. Such a reading of Article 23 is hardly convincing without weighty arguments. As has been shown, however, such arguments can arguably not be inferred from Article 23 of the DSU. IV. Conclusions These comments have focussed on the areas of disagreement in the preceding papers of Joost Pauwelyn on the one hand and Gabrielle Marceau and Anastasios Tomazos on the other. This paper has argued that, since the WTO treaty constitutes a normal international agreement, inter se modifications are permissible in line with standard rules of general international law, unless it can be shown that WTO Members have contracted out from these principles. The same is true for the other relevant default rule, namely that adjudicating bodies have the authority and obligation to assess their own 44

45

The requirement that the rights of third WTO Members are not infringed may entail the consequence that their interests have to be taken into account in non-WTO proceedings in a way similar to WTO standards (cf Article 10 of the DSU). Cf Shany (2003), 185-186.

98

Erich Vranes

competence. Consequently, authors arguing that these legal starting points do not apply would have to advance sufficient legal grounds in both regards. The preceding brief analysis has tried to make it clear that the stance taken by Marceau and Tomazos, which constitutes the exception to the (general international law) rule, does not appear convincing, as sufficient arguments cannot be inferred from the WTO provisions to which these authors refer in support of their arguments. This, on the other hand, militates in favour of Pauwelyn’s approach which essentially constitutes a restatement of the rules of general international law: inter se modifications of the WTO treaty are permissible in line with the principles of international law; if such modifications are invoked as a defense, they must be respected by WTO adjudicating bodies.

References Klaus Adomeit (1986), Normlogik – Methodenlehre – Rechtspolitologie: Gesammelte Beiträge zur Rechtstheorie 1970-1985 (= Schriften zur Rechtstheorie Vol. 120), Berlin (Duncker & Humblot) 1986. Robert Alexy (1994), Theorie der Grundrechte, Frankfurt a. M. (Suhrkamp) 21994. Lorand Bartels (2001), Applicable Law in WTO Dispute Settlement Proceedings, in: Journal of World Trade 35 (2001), 499519. Jeremy Bentham (1970), Of Laws in General (1782, ed. posthumously by H. L. A. Hart), London (The Athlone Press) 1970. Wladiyslaw Czaplinski / Gennady M. Danilenko (1990), Conflicts of Norms in International Law, in: Netherlands Yearbook of International Law 21 (1990), 3-42. Walter Dubislav (1981), Die Definition, Hamburg (Felix Meiner Verlag) 41981 (reprint of the 3rd edition of 1931). Karl Engisch (1977), Einführung in das juristische Denken, Stuttgart et al. (Kohlhammer) 71977. Alexandra Gonzalez-Calatayud / Gabrielle Marceau (2002), The Relationship between the Dispute-Settlement Mechanisms of

Comments on Joost Pauwelyn’s Paper

99

MEAs and those of the WTO, in: Review of European Community and International Environmental Law 11 (2002) No. 3, 275-286. Herbert Lionel Adolphus Hart (1971), Bentham’s ‘Of Laws in General’, in: Rechtstheorie 2 (1971), 55-66. Herbert Lionel Adolphus Hart (1992), The Concept of Law, Oxford (Clarendon Press) 1992. Peter Hilpold (2002), Aktuelle Rechtsfragen zum WTOStreitbeilegungsverfahren, in: Favorita Papers 2/2002, 46-61. Wilfred C. Jenks (1953), The Conflict of Law-Making Treaties, in: British Year Book of International Law 30 (1953), 401-453. Wolfram Karl (1984), Treaties, Conflicts between, in: Rudolf Bernhardt (ed.), Encyclopedia of Public International Law Vol. 7, Amsterdam / London / New / Tokyo (North Holland) 1984, 468-473. Hans Kelsen (1979), Allgemeine Theorie der Normen (edited by Kurt Ringhofer and Robert Walter), Wien (Manz) 1979. Friedrich Klein (1962), Vertragskonkurrenz, in: Karl Strupp / Hans-Jürgen Schlochauer (eds.), Wörterbuch des Völkerrechts Vol. 3, Berlin (de Gruyter) 21962, 555. Ulrich Klug (1982), Juristische Logik, Berlin (Springer) 41982. Kyung Kwak / Gabrielle Marceau (2002), Overlaps and Conflicts of Jurisdiction between the WTO and RTAs. Paper presented in the Conference on Regional Trade Agreements, WTO 26 April 2002 (available at http://www.wto.org/english/tratop_e/ region_e/sem_april02_e/sem_april02_ reading_e.htm9). Hans Lenk (1974), Konträrbeziehungen und Operatorengleichungen im deontologischen Sechseck, in: Hans Lenk (ed.), Normenlogik. Grundprobleme der deontischen Logik, Pullach bei München (Verlag Dokumentation) 1974, 198. Gabrielle Marceau (2001), Conflicts of Norms and Conflicts of Jurisdiction: The Relationship between the WTO Agreement and MEAs and other Treaties, in: Journal of World Trade 35 (2001) No. 6, 1081-1131. David Palmeter / Petros Mavroïdis (1998), The WTO Legal System: Sources of Law, in: American Journal of International Law 92 (1998), 398-413.

100

Erich Vranes

Joost Pauwelyn (2003), Conflict of Norms in Public International Law. How WTO Law Relates to Other Rules of International Law, Cambridge (Cambridge University Press) 2003. August Reinisch (2002), Der Streit ums Forum – oder: Was gehört eigentlich vor WTO-Panels?, in: Recht der Internationalen Wirtschaft 48 (2002), 449-456. Richard Robinson (1968), Definition, Oxford (Clarendon Press) 1968. Klaus F. Röhl (1995), Allgemeine Rechtslehre. Ein Lehrbuch, Köln et al. (Heymann) 1995. Yuval Shany (2003), The Competing Jurisdictions of International Courts and Tribunals, Oxford (Oxford University Press) 2003. August Thon (1878), Rechtsnorm und subjectives Recht, Weimar (Böhlau) 1878. Joel Trachtman (1999), The Domain of WTO Dispute Resolution, in: Harvard International Law Journal 40 (1999) No. 2, 333377. Erich Vranes (2003), The Banana Dispute. Fundamental Issues under WTO Law, in: Fritz Breuss / Stefan Griller / Erich Vranes (eds.), The Banana Dispute. An Economic and Legal Analysis, Vienna / New York (Springer) 2003, 39-111. Ota Weinberger (1989), Rechtslogik, Berlin (Duncker & Humblot) 2 1989. Wolfgang Weiss (2003), Security and Predictability under WTO Law, in: World Trade Review 2 (2003) No. 2, 183-219. Wilhelm Heinrich Wilting (1996), Vertragskonkurrenz im Völkerrecht, Köln et al. (Heymann) 1996. Rüdiger Wolfrum / Nele Matz (2003), Conflicts in International Environmental Law (= Beiträge zum ausländischen öffentlichen Recht und Völkerrecht Vol. 164), Berlin (Springer) 2003. Manfred Zuleeg (1977), Vertragskonkurrenz im Völkerrecht. Teil I: Verträge zwischen souveränen Staaten, in: German Yearbook of International Law 20 (1977), 246-276.

William J. Davey*

The Quest for Consistency: Principles Governing the Interrelation of the WTO Agreements I. II.

The WTO Agreements and the Treatment of Conflicts WTO Jurisprudence and the Treatment of Conflicts A. Conflicts in Public International Law B. Resolving Conflicts in the WTO Agreements 1. The EC Bananas Case: GATT Article XI and the ATC 2. The Indonesia Autos Case: GATT and the Subsidies Agreement 3. GATT and the TBT Agreement 4. Summary III. Overlapping Obligations in the WTO Agreement IV. The Consequences of the Cancún Failure on the Coherence of International Economic Law V. Conclusion References

105 108 108 110 110 111 115 120 121 123 125 126

Ensuring consistency and avoiding conflicts between agreements have been longstanding problems in the GATT/WTO system. For the most part, GATT – the General Agreement on Tariffs and Trade – itself seems to be basically internally consistent.1 However, in the *

1

Guy Raymond Jones Professor of Law, University of Illinois College of Law. I would like to thank Lorand Bartels and the other participants in the conference for their comments. There appear to be a few potential conflicts that may arise from the basic provisions of GATT. See the cases mentioned under ‘Scope of alternative GATT articles’ in Pescatore / Davey / Lowenfeld (1991), Index 1/15. For example, there are aspects of the relationship between Articles III and XI that have never been definitively established. Should a measure that bans certain types of products, whether domestic or imported, be analyzed under Article XI if it is applied at

102

William J. Davey

Tokyo Round, the coherence of GATT was undermined by the negotiation of side agreements, which dealt with GATT-related topics – such as dumping, subsidies, valuation and standards, but to which many GATT members were not parties.2 This resulted in the socalled GATT à la carte. From a legal perspective, it raised difficult issues as to when GATT members, who were not parties to a side agreement, might nonetheless benefit from it because of GATT’s most-favored-nation clause.3 The creation of the World Trade Organization in 1995 largely eliminated this problem through what was called its ‘single under-

2

3

the border to bar entry to a product? Or under Article III? Or both? While it is often said that Article III applies to internal measures and Article XI applies to border measures, the distinction is not always clear and the language at issue is subject to more than one interpretation. The Ad note to Article III provides that such measures are subject to Article III, but does not explicitly exclude application of Article XI as well. Generally speaking, however, the potential conflicts among GATT articles did not raise major problems for dispute settlement panels in their application of GATT rules. The parties to the Tokyo Round agreements tended to be limited to developed country GATT parties, but the membership varied on an agreement-by-agreement basis. The members of each agreement were listed annually in GATT Activities 19xx, the annual review of the work of the GATT. Because some of the Tokyo Round agreements dealt with matters that fell within the scope of Article I and its most-favored-nation (MFN) obligation, it would seem that all GATT members would benefit from concessions in respect of those matters made by a member to other parties to a Tokyo Round agreement, notwithstanding the fact that the GATT member not a party to the agreement would not be subject to any obligations under the agreement. The matter was never definitively decided, although the issue was raised in a case brought by India against the United States. In that case, India noted that the United States only applied countervailing duties (CVDs) on products from parties of the Tokyo Round subsidies agreement if imports of those products caused material injury to US industry and claimed that the US would violate the MFN clause if it applied CVDs on India products without considering whether they caused any injury. The case was settled (GATT, GATT Activities in 1981, at 5051). Some of the Tokyo Round agreements, such as the one on government procurement, could not be analyzed in this way since Article I would not seem to apply.

Principles Governing the Interrelation of the WTO Agreements 103

taking’ approach, pursuant to which all WTO members are required to adhere to all WTO agreements, with a couple of minor exceptions.4 Unfortunately, while this approach brought virtually all WTO commitments under a single umbrella, it did not succeed in creating one coherent system of legal obligations. This failure was due in large part to two factors: First, although some of the WTO agreements expressly deal with the problems of conflict between WTO agreements, not all of the agreements deal with this fundamental issue. Second, while the agreements were all negotiated in one round of negotiations, they were negotiated in many separate negotiating groups, which made the possibility of conflicts more likely than it might have otherwise been since negotiators in one group were not necessarily aware of the details of the negotiations in other groups. The problems of consistency and conflicts in the WTO agreements may well increase, at least in the short term. The recent failure of the WTO’s Cancún ministerial meeting,5 which was intended to give direction and impetus to the ongoing Doha Development Agenda negotiations, does not bode well for improving the coherence of international economic law based on the WTO agreements. While it is too early to know what the precise consequences of the Cancún failure will be, two results seem likely. First, in order to 4

5

WTO Members must accept all of the so-called multilateral trade agreements annexed to the WTO Agreement. Adherence to the plurilateral agreements is optional. As of 2004, there were only two such agreements – on government procurement and on civil aircraft. The fifth WTO ministerial conference held at Cancún in midSeptember 2003 was intended to define more precisely the negotiating parameters for the Doha Development Agenda, the round of negotiations launched at the fourth WTO ministerial conference held in Doha in November 2001. The meeting was terminated following an inability to reach consensus on whether to start negotiations on the so-called Singapore issues – competition, investment, trade facilitation and transparency in government procurement. A decision to start such negotiations at Cancún had been taken at Doha, subject to agreement on modalities at Cancún, but the disagreement at Cancún seemed to involve more fundamental issues than modalities. There was also broad disagreement over the basis for proceeding with negotiations on agricultural issues, which could well have led to a breakdown at Cancún even if the Singapore issues had been dealt with satisfactorily.

104

William J. Davey

gain agreement to negotiate on issues that are new to the WTO system, such as competition and investment, it is quite possible that the single undertaking approach of the Uruguay Round will be abandoned, with the result that the interrelationship of the generally applicable WTO agreements and WTO plurilateral agreements may become more important and complex. Second, as has been typical when GATT or WTO negotiations are significantly delayed, many WTO members – such as the US, the EC, Brazil and such Asian countries as Japan and Korea – have already announced their intention to pursue additional regional trade agreements while the multilateral process is blocked.6 To the extent that this actually occurs, it also will undermine and complicate the coherence of international economic law. This paper will focus on the current problems of conflict and overlap among WTO agreements. It will leave to one side the related and very interesting problem of how conflicts between the WTO agreements and other international agreements should be dealt with.7 It begins with an examination of the structure of the WTO agreements and the way that they deal explicitly with conflicts and overlaps. It then considers how conflicts between WTO agreements that are not dealt with explicitly in those agreements should be resolved, particularly in light of international law rules on conflicts, taking into account any special circumstances applicable because of the particular nature of WTO agreements. In that connection, WTO jurisprudence dealing with this issue is examined. Third, it considers the related problem of overlapping obligations and how WTO dispute settlement has approached that issue. Finally, it briefly treats problems of consistency and coherence in in-

6

7

See, e.g., Robert Zoellick, Op-ed: America will not wait for the won’t-do countries, The Financial Times, Sept. 22, 2003. Since Cancún, the US has in fact completed FTA negotiations with Australia, five Central American countries, the Dominican Republic and Morocco. See also Pascal Lamy, Result of the WTO Ministerial Conference in Cancún, Speech to a Plenary Session of the European Parliament, Strasbourg, Sept. 24, 2003, at 5 (‘The second question: do we remain attached to the priority for multilateralism? … Are bilateral or regional agreements still a complement to multilateral disciplines or must come to substitute for them in the event of a standstill [at the WTO]?’). This topic is dealt with in Joost Pauwelyn’s paper, supra.

Principles Governing the Interrelation of the WTO Agreements 105

ternational economic law that may arise from the failure of the Cancún ministerial meeting. I. The WTO Agreements and the Treatment of Conflicts The potential for problems of overlap and internal conflicts within and among the WTO agreements is significant, in large part because of the basic structure of WTO obligations. At the top of the structure is the Agreement Establishing the World Trade Organization (the ‘WTO Agreement’) – a relatively brief agreement that establishes the WTO and its principal organs and deals with various institutional issues such as decision-making and accession.8 Annexed to this agreement are three general agreements: First, in Annex 1A, the General Agreement on Tariffs and Trade 1994, simply referred to as GATT, which is essentially the 1947 GATT text modified by a half dozen understandings on specific GATT provisions.9 Second, in Annex 1B, the General Agreement on Trade in Services, known as GATS.10 Third, in Annex 1C, the Agreement on Trade-Related Aspects of Intellectual Property Rights, known as the TRIPS Agreement.11 Annex 1A, which contains GATT 1994, also contains twelve other agreements concerned with trade in goods. In many cases, these additional agreements deal with issues also dealt with in GATT 1994, including those in respect of agriculture, sanitary measures, textiles and clothing trade, technical barriers, import licensing, customs valuation, dumping, subsidies, safeguards and trade-related investment measures. A few of these agreements deal with subjects that are largely not dealt with in GATT rules, such as the agreements on rules of origin and pre-shipment inspection. This structure means at a minimum that there will be significant overlaps. That is obviously the case where both GATT and one of the other agreements on trade in goods deal with the same subject matter. The GATT rules on safeguards in GATT Article XIX will obviously be impacted by the Agreement on Safeguards. However, the potential for overlap is in fact much broader. For example, rules on intellectual property may affect the sale of goods, and 8 9 10 11

The WTO Agreement is available at the WTO website – www.wto.org – and is also found in WTO (1995), 6-18. Ibidem, at 21-38. The text of GATT 1947 is at 485-558. Ibidem, at 325-364. Ibidem, at 365-403.

106

William J. Davey

thereby bring into play GATT rules that limit a Member’s rights to affect the sale of goods. The broad reach of GATT Article III:4, which requires national treatment in respect of laws, regulations and requirements affecting the internal distribution of goods, may well affect distribution services, which are for most part subject to regulation under GATS. Despite these overlaps and potential for conflict, the WTO agreements generally do not recognize, let alone deal with, the problem of overlapping obligations and contain relatively few rules resolving conflicts. Moreover, even those agreements which purport to resolve conflicts by specifying the agreement that has priority in the event of a conflict, do not define what is meant by the term ‘conflict’. The principal rules that do exist are as follows: First, as to the WTO Agreement itself, it provides in Article XVI:3 that in the event of a conflict, which as noted above is a concept that is not defined, between the WTO Agreement and any of the annexed trade agreements, the WTO Agreement shall prevail to the extent of the conflict. Given that the WTO Agreement deals mainly with institutional issues, not many such conflicts – however defined – are likely to arise. Second, in respect of the agreements in Annex 1A dealing with trade in goods, a general interpretative note to Annex 1A specifies that in the event of a conflict, here too undefined, between a provision of GATT 1994 and a provision of any other Annex 1A agreement, the provision of the other agreement will prevail.12 Thus, if there is a conflict between a provision of the Agreement on Safeguards and a provision of GATT 1994, the Safeguards Agreement provision applies. Third, a number of Annex 1A agreements – that is, those dealing with trade in goods – have specific provisions dealing with the priority of agreements. For example, Article 21.1 of the Agreement on Agriculture stipulates that GATT 1994 and the other Annex 1A agreements apply subject to the provisions of the Agriculture Agreement. This provision clarifies the priority and effectiveness of certain provisions of the agriculture agreement, which allow measures such as certain subsidies and safeguards that would violate the otherwise applicable rules. A second example is found in the Agreement on Sanitary and Phytosanitary Measures – the so-called SPS Agreement, which provides that it applies to SPS measures to the exclusion of the Agreement on 12

Ibidem, at 20.

Principles Governing the Interrelation of the WTO Agreements 107

Technical Barriers to Trade, which would otherwise seem to overlap in part with the SPS Agreement. Finally, the Understanding on the Settlement of Disputes – known as the DSU – provides that its rules apply subject to certain special or additional rules and procedures that are contained in other WTO agreement and that are specified in an Annex to the DSU. It is specifically provided that to the extent there is a ‘difference’ (as opposed to a conflict) between the DSU rule and the special or additional rule, the latter shall prevail. It is obvious from what I have outlined, that there is much regarding overlaps and conflicts that is not dealt with in the WTO Agreement. One can ask why are the rules so incomplete? Why is there no definition of conflict? Why are there rules governing the relationships of some agreements specified but not others? There seems to be no clear answer to these questions beyond the obvious problem of uncoordinated negotiating groups alluded to above. Of course, some of the most obvious conflict problems were treated – I would include in that category the relationship between GATT 1994 and the more specialized agreements on GATT-related issues in Annex 1A and the relationship of the DSU and the special provisions on dispute settlement in the various WTO agreements. But other obvious problems were not addressed – such as the interrelationship between GATT, GATS and TRIPS and the relationship of the Agreement on Trade-Related Investment Measures and the Agreement on Subsidies, both of which deal with local content requirements, albeit in different ways. Thus, for better or worse, there are difficult outstanding issues with regard to the principles governing the interrelationships of the WTO agreements.

108

William J. Davey

II. WTO Jurisprudence and the Treatment of Conflicts13 A. Conflicts in Public International Law Given the paucity of clear rules for treating conflicts in the WTO agreements, it is not surprising that WTO jurisprudence on the issue has not developed in an entirely satisfactory and coherent matter. To analyze this issue, it is first necessary to define what is meant by ‘conflict’. Since the WTO agreements provide no real guidance on the issue, it is useful to consider how that concept is handled in public international law.14 Of course, the first rule when dealing with an alleged conflict of obligations is that the treaty interpreter should try to interpret the relevant obligations so that they are not in conflict.15 The second rule is that as between the same parties the most recent obligation prevails over an earlier obligation.16 But neither of these rules is particularly helpful in the event of an actual conflict in the WTO context – an actual conflict cannot be interpreted away; the rule that the last in time prevails is not useful since the whole complex of WTO agreements entered into force at the same time. What more then can we learn from public international law on the definition and resolution of conflict problems? Quite interestingly, for me at least, is that an examination of what has been written on this issue in public international law seems to be rather unhelpful. It appears that most of the eminent authorities in public international law have defined ‘conflict’ as meaning what some call a ‘true’ conflict. By that they mean that a conflict exists only where a country has undertaken two obligations and that it is impossible

13

14 15 16

I will focus in this section on conflicts among the Annex 1A agreements – the agreements involving trade in goods. For an analysis of the conflict issues presented at the level of the general agreements, i.e. GATT vs. GATS vs. TRIPS, see Davey / Zdouc (2003). For the most part, I will not consider panel and Appellate Body decisions regarding the relationship between the DSU and other dispute settlement provisions in the WTO agreements. For those cases, see generally Lorand Bartels, infra this volume. See generally Pauwelyn (2003). Ibidem, at 240-244. Ibidem, at 335.

Principles Governing the Interrelation of the WTO Agreements 109

for it to comply with both at the same time.17 For example, it has undertaken in one case to do an act ‘X’ and in another case to not do ‘X’.18 While no one would dispute that this situation represents a conflict, the harder question is whether this exhausts the meaning of 17

18

Various authorities are cited for this proposition in the Panel Report on Indonesia – Certain Measures Affecting the Automobile Industry, WT/DS54/R, adopted by the WTO Dispute Settlement Body on 23 July 1998. For example, in footnote 649 at page 329-330, the report quotes Wolfram Karl, Conflicts Between Treaties, in: R. Bernhardt et al. (eds.), Encyclopedia of Public International Law, vol. 7, Amsterdam (North Holland) 1984, at 468: (‘[T]echnically speaking, there is a conflict when two (or more) treaty instruments contain obligations which cannot be complied with simultaneously’). See also Jenks (1953), 430 (‘[A] conflict in the strict sense of direct incompatibility arises only where a party to the treaties cannot simultaneously comply with its obligations under both treaties’). There is an interesting US Supreme Court case that considers the idea of true conflict in an international setting. In Hartford Fire Insurance Co. vs. California, 509 U.S. 764 (1993), a group of British insurers had engaged in a certain activity in London that was lawful in the United Kingdom, but that was arguably unlawful in the United States because of the effect it had on insurance markets in the US. While the insurers did not challenge the jurisdiction of the United States, they did argue that US courts should decline to exercise their jurisdiction because there was a conflict between the US and UK rules and the insurers’ behavior was lawful under the rules of the place where it occurred. The Supreme Court rejected the argument on the grounds that there was no true conflict – in its view, the insurers could have complied with US law without violating UK law. They simply had to refrain from the activity that was illegal in the US. To make the example clearer – what was presented was not the situation, which could easily arise, where a foreign defendant was prohibited from doing what a US court or regulator required it to do. Accordingly, the Supreme Court declined to require US courts to decline to exercise jurisdiction under comity principles. Pauwelyn cites a number of different formulations of the definition of conflict that have been advanced in doctrinal writing in public international law. See Pauwelyn (2003), at 166-169. To me, his review tends to establish that the prevailing view is to limit the concept of conflict to mutually exclusive obligations, although a number of authorities he quotes are susceptible of being interpreted less narrowly. Pauwelyn himself takes a broad view of conflict (ibidem, at 169188).

110

William J. Davey

conflict. In particular, what if a country has undertaken not to do ‘X’ under one agreement, but it has entered into another agreement that explicitly permits it to do ‘X’. B. Resolving Conflicts in the WTO Agreements 1. The EC Bananas Case: GATT Article XI and the ATC Perhaps the true-conflict rule makes sense in some contexts, but its application would raise serious problems in the WTO context. To take a concrete WTO example – GATT Article XI bans the use of quotas, while the Agreement on Textiles and Clothing permits certain WTO members to maintain quotas on textile and clothing products until 2005. If the concept of ‘conflict’ is limited to ‘true’ conflicts, as described earlier, then there is no conflict between these provisions. The WTO members authorized to impose quotas under the ATC – the Agreement on Textiles and Clothing – are not required to do so. Thus, they can comply with that agreement and the GATT rule by not imposing quotas. But does that make sense in the WTO context? It clearly does not. There is one treaty framework to which all WTO members are parties. To require those parties authorized by the ATC to forego their right under that agreement to impose quotas because quotas are banned by GATT would effectively defeat the entire purpose of the ATC, which was designed, subject to specified conditions, to provide a ten-year phase-out for such quotas.19 Such a result would fly in the face of basic principles of treaty interpretation. As the WTO Appellate Body has said, a treaty interpreter is not permitted through interpretation to reduce a provision to inutility, a result that it characterized in the Gasoline case to be in violation of the principles of treaty interpretation.20 Rendering a whole agreement inutile could never be countenanced. While this particular issue involving Article XI and the ATC has never arisen, it was used by the panel in the famous Bananas case to inform its discussion of what constituted a conflict.21 As such, that panel reached the 19 20

21

See Croome (1999), at 88-92, 192-199, 268-270. Appellate Body Report on United States – Standards for Reformulated and Conventional Gasoline, WT/DS2/AB/R, at 23, adopted by the WTO Dispute Settlement Body on 20 May 1996. Panel Report on European Communities – Regime for the Importation, Distribution and Sale of Bananas, WT/DS27/R, paras. 7.1587.162, adopted by the WTO Dispute Settlement Body on 25 Septem-

Principles Governing the Interrelation of the WTO Agreements 111

conclusion that limiting the notion of conflict to true conflict, as described above, would be inappropriate. In was, in the panel’s view, necessary to include within the scope of the definition of conflict the situation where one agreement proscribed what another agreement explicitly permitted. Not to do so would clearly frustrate the intentions of the parties, as it could not be argued seriously that they intended one agreement to have no effect. The resolution of the meaning of conflict was not crucial to the Bananas case. Indeed, it was dealt with by the panel in a footnote, and the discussion was not among the numerous issues appealed by the parties. 2. The Indonesia Autos Case: GATT and the Subsidies Agreement A year later in the 1997 Indonesia Autos case, the definition of conflict was crucial.22 In the Autos case, which was not appealed, Indonesia argued that there were conflicts between GATT Article III and the TRIMs Agreement, i.e., the Agreement on Trade-Related Investment Measures, on the one hand, and the Subsidies Agreement, on the other hand. Focusing on the TRIMs – Subsidies Agreement conflict, Indonesia noted that subsidies conditioned on meeting minimum local content requirements were normally prohibited by the Subsidies Agreement, but that developing countries such as Indonesia were not subject to that prohibition during a transition period that was still in effect at the time of the Autos case. At the same time, Indonesia noted that the TRIMs Agreement prohibited local-content measures generally, whether they were a condition for subsidies or not. For Indonesia, this meant that applying the TRIMs Agreement prohibition would render the Subsidies Agreement transition period inutile. The analogy to the GATT-ATC situation discussed in the Bananas case seems clear. The panel, however, refused to recognize this situation as a conflict. Rather, it took the position that for a conflict to be found, there had to be a true conflict and that this situation was not a true conflict, since Indonesia could comply with its obligations under both agreements by not providing a subsidy conditional on meeting a local-content requirement. Moreover, the panel argued that the

22

ber 1997. The discussion of conflicts was not appealed and consequently was not discussed in the Appellate Body’s report. Panel Report on Indonesia – Certain Measures Affecting the Automobile Industry, WT/DS54/R, adopted by the WTO Dispute Settlement Body on 23 July 1998.

112

William J. Davey

two agreements covered different types of obligations and different subject matters. To quote the panel: ‘In the case of the Subsidies Agreement, what is prohibited is the grant of a subsidy contingent on the use of domestic goods, not the requirement to use domestic goods as such. In the case of the TRIMs Agreement what is prohibited are TRIMs – trade-related investment measures – in the form of local content requirements, not the grant of an advantage, such as a subsidy.’23 The panel went on to note that a violation of the Subsidies Agreement could be cured even if a local content requirement was maintained, while the TRIMs violation could be removed by simply changing the conditions for the subsidy, which could remain in effect. While the panel’s arguments are not at all unreasonable, they are not completely persuasive if one accepts, as I do, the correctness of the Bananas example of the GATT-ATC relationship. However, for a number of reasons, the result of the Indonesia panel seems correct to me. I would not base the result on a mechanical application of the no true conflict rule. Rather, I would recognize the possibility of a conflict under the broader definition of conflict. But I would ultimately find no conflict because of consideration of such factors as the structure of the agreements and the obligations therein, i.e. in the terms of the Vienna Convention on the Law of Treaties – the context in which the obligations appear and the object and purpose of the agreements24 – which I think can be examined to ensure a result that is consistent with the likely intent of the parties. In that vein, I would note first, as noted earlier in the discussion of the Bananas panel’s analysis, the key reason for giving effect to the Textiles Agreement over GATT’s quota ban is to give effect to what seemed to be the clear intent of the parties – to fail to do so would render an entire agreement meaningless. That would not be

23 24

Ibidem, at para. 14.50. Article 31 of the Vienna Convention provides a general rule of interpretation as follows: ‘A treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose.’ It is accepted that DSU Article 3.2 specifies that this rule shall be used in interpreting the WTO Agreement.

Principles Governing the Interrelation of the WTO Agreements 113

the result, however, in the Indonesia Autos case if the TRIMs Agreement were given effect. Second, in examining the overall structure of WTO obligations on TRIMs and subsidies that are conditioned on a local content requirement, it is important to note that for developing country TRIMs of any kind existing as of the date of WTO Agreement’s effectiveness, if notified to the WTO, those TRIMs would be valid for five years – exactly the period for which the Subsidies Agreement rule on prohibited subsidies containing local content requirements would not apply to developing countries.25 Thus, to a considerable degree the provisions of the two agreements are coordinated and it is only as to new measures that were implemented after 1995 or measures that were not notified that a potential conflict exists. To the extent of this coordination, it is arguable that the agreements should not be viewed as being in conflict such that the TRIMs obligation is trumped by the Subsidies Agreement exception. Third, it is instructive to consider that once the relevant transition periods have expired, the measures at issue in Indonesia Autos would violate both the TRIMs Agreement and the Subsidies Agreement. However, the consequences of the violations would be different. For the part of the measure that violated the Subsidies Agreement, the provisions of the Subsidies Agreement on expedited dispute settlement and the requirement of prompt withdrawal of the subsidy would apply, while for the part of the measure that violated the TRIMs Agreement, the normal DSU rules would apply. Seen in that way, one could conclude that a no-conflict finding would not at all render an agreement or provision inutile. That is to say that in the case of new or non-notified TRIMs, during a five year period they would be subject to normal dispute settlement, but would not be subject to the expedited procedures and prompt withdrawal requirement of the Subsidies Agreement. Thus, all of the relevant provisions would have significant meaning at all times, suggesting that the potential minimal conflict existing after the above-described coordination provisions were taken into account would arguably disappear. Finally, and somewhat related to the concept of lex specialis, I would note that it would be odd to think that a longstanding GATT

25

TRIMs, art. 5.2; Subsidies Agreement, art. 27.3.

114

William J. Davey

rule – the prohibition of local-content requirements26 – could be negated by what I would characterize as a minor provision in a wholly unrelated agreement. It seems acceptable to conclude that the Textiles Agreement, which deals only with the administration for ten years of textile and clothing quotas, was intended to supplant temporarily the GATT quota ban. Likewise, if the TRIMs Agreement were to provide a right to provide TRIMs for a specified period (as in fact it does), then I would have no problem concluding that Article III would not apply during that period. It seems strange though that a long-standing ban on TRIMs in two other agreements – one dealing explicitly and only with TRIMs – should be negated by a time-limited exception to the ban on prohibited subsidies in the Subsidies Agreement, since that agreement deals principally with quite distinct issues that are unrelated to TRIMs in general. In any event, I find the Indonesia Autos result to be appropriate. As outlined above, however, I would not base the decision so much as on defining ‘conflict’ as including only ‘true’ conflicts. I would accept potentially a broader conflict definition, but would ultimately find no conflict based on an examination of the structure of the agreements and the effects of the various provisions. Thus, a flexible approach to ‘conflicts’ is needed – both in terms of interpreting agreements so as to avoid finding conflicts no matter how defined and in defining the notion of conflict so as effectuate the intent of the parties, as demonstrated by the structure and operation of the agreements in question. Ascertaining that intent seems to me to be the key of dealing with alleged conflicts and to do so it will be necessary to examine closely the structure and interaction of the relevant agreements, as well as their wording.27

26

27

The obligations in the TRIMs Agreement are generally viewed as restating GATT jurisprudence, not creating new obligations. See GATT Panel Report on Canada – Administration of the Foreign Investment Review Act, BISD 30S/140, adopted by the GATT Council on 7 February 1984. For conflicting views on the correctness of the Indonesia Autos case, see Pauwelyn (2003), at 193-194 and Marceau (2001), at 1085. The Appellate Body has not spoken clearly on the issue, although I tend to think that its jurisprudence to date under the DSU suggests that it would support a true-conflict approach. Accord, Pauwelyn (2003), at 194; but see Bartels, infra.

Principles Governing the Interrelation of the WTO Agreements 115

3. GATT and the TBT Agreement The most interesting conflict situation that has become apparent to date involving GATT and the other agreements relating to goods trade probably involves GATT Articles III and XX, on the one hand, and Article 2 of the Agreement on Technical Barriers to Trade – the TBT Agreement, on the other. Although this conflict has been implicitly presented in two cases – US Gasoline and EC Asbestos – it was avoided by the panels, which resolved the cases on the basis of Article III.28 The first potential conflict is between GATT Article III – the basic national treatment rule, which is subject to the exceptions of Article XX (health and safety, conservation, etc.) – and TBT Article 2.1, which imposes a national treatment obligation in respect of technical requirements, which is not subject to any exceptions since the TBT Agreement has no general exceptions clause. Assuming, as I think is necessary, that GATT Article III and TBT Article 2.1 impose the same national treatment obligation in respect of technical requirements,29 then the two agreements are usually in harmony, 28

29

The panel in US Gasoline simply noted after having considered the GATT issues raised that it was not necessary to consider the TBT issues (Panel Report on United States – Standards for Reformulated and Conventional Gasoline, WT/DS2/R, para. 6.43, adopted by the WTO Dispute Settlement Body on 20 May 1996). The panel in Asbestos concluded that the challenged measure did not fall within the coverage of the TBT Agreement, a conclusion reversed by the Appellate Body (Appellate Body Report on European Communities – Measures Affecting Asbestos and Asbestos-Containing Products, WT/DS135/AB/R, paras. 59-83, adopted by the WTO Dispute Settlement Body on 5 April 2001). The Appellate Body has generally taken the view that the more specific agreement should be examined first (see Appellate Body Report on European Communities – Regime for the Importation, Distribution and Sale of Bananas, WT/DS27/AB/R, para. 204, adopted by the WTO Dispute Settlement Body on 25 September 1997) (panel should dispose of claims under the Import Licensing Agreement prior to examining GATT issues related to licensing). Thus, it seems that the Appellate Body would take the position that the more appropriate approach in dealing with cases raising GATT and TBT issues would be to consider the TBT issues first, so it is likely that the GATT/TBT conflict will not be so easily avoided in the future as it was in Gasoline and Asbestos. It could be argued that a different notion of like product should be applied under Article 2.1 than under Article III, but I see no justifi-

116

William J. Davey

except where GATT Article XX excuses a violation of GATT Article III. In that case, the challenged measure would not violate GATT, but would violate the TBT Agreement.30 This would occur despite the fact that the same national treatment obligation was present under each agreement. While not a true conflict in the sense explored above – if it is viewed as a conflict for WTO purposes, the WTO Agreement’s conflict rule would give priority to the TBT Agreement obligation over a conflicting GATT provision. That would mean that the same provision in the two agreements would effectively have two consequences – what one permits as a consequence of the Article XX exception, the other forbids. In my view, such a result should be avoided by effective interpretation. Would that be consistent with the parties’ intent? That is not clear,31 but one can speculate that the intent of the parties in including TBT Article 2.1, which dates from the Tokyo Round, was not to expand in some undefined fashion the reach of GATT’s nondiscrimination rules, but rather to make provision for the admission of parties to the TBT Agreement who were not GATT contracting parties and who therefore would not be subject to the general GATT nondiscrimination rules that would ordinarily cover technical regulations. That is confirmed by the structure of the Tokyo Round TBT Agreement, as Articles 15.2 and 15.3 thereof specifically provide for non-GATT parties. In fact, there was a non-GATT party to the TBT Agreement from early on: Tuni-

30

31

cation for such an approach and no basis for choosing any particular different definition of like product for purposes of Article 2.1. Indeed, if one should first analyze the measure under the TBT Agreement, the measure would be found to violate the TBT Agreement and the issue of its GATT compatibility would presumably not even be considered under the usual notions of judicial economy. One could argue that discrimination in technical requirements was viewed as always inexcusable by negotiators and thus no exceptions were needed or appropriate. Even if that were an accurate description of their views, one could ague that they had not considered fully the consequences of the broad definition of like products that GATT/WTO jurisprudence has often produced. As such, one could argue that their intent not to have any exceptions was based on a false premise and that effective interpretation is needed. As noted above (see note 30), I do not think that adjusting the definition of like products is the answer, although it is a possible approach to avoiding a finding of a TBT violation.

Principles Governing the Interrelation of the WTO Agreements 117

sia was a party to the TBT Agreement as of 1981, although it was not formally a GATT contracting party at the time.32 In any event, there is no evidence that the TBT national treatment clause was intended to negate Article XX (or Article XXI, for that matter). Indeed, the preamble to the TBT Agreement itself explicitly states the contrary and largely incorporates the ideas of Article XX – that no country should be prevented from taking measures necessary, inter alia, for the protection of life, health or the environment, provided that they are not applied in a manner that would constitute a means of arbitrary or unjustifiable discrimination or a disguised restriction on international trade. Thus, I would argue that the appropriate result is that a TBT measure that violates Article III:4, but which is permitted by Article XX, should not be found to violate Article 2.1 of the TBT Agreement. In my view, the context of Article 2.1 (i.e., the preamble and Article 15) demonstrates that it was not intended to expand the non-discrimination requirement beyond what was found in GATT. The second potential conflict is between GATT Articles III and XX, and Article 2.2 of the TBT Agreement. Article 2.2 prohibits technical regulations that are unnecessary obstacles to international trade and specifies that such regulations should not be more trade restrictive than necessary to fulfill a legitimate objective – several of which are listed. If a measure violates Article III and is not excused by Article XX and at the same time violates Article 2.2 32

Report (1981) of the Committee on Technical Barriers to Trade, BISD 28S/34. The 1980 committee report noted that Tunisia had deposited its acceptance in 1980, but that certain terms had not yet been worked out at that time (Report of the Committee on Technical Barriers to Trade, BISD 27S/37). Tunisia had provisionally acceded to GATT in 1959, a status that continued until 1989. The declaration conferring that status provided for tariff negotiations between Tunisia and GATT contracting parties, but provided that Tunisia would have no direct rights with respect to the concessions contained in GATT tariff schedules (GATT, BISD 8S/15-16). Presumably, since it was not formally a contracting party, Tunisia was not subject to dispute settlement and recommendations under Article XXIII. While Tunisia seems to be the only party to the Tokyo Round TBT Agreement that was not a GATT contracting party, other Tokyo Round agreements (e.g., Bovine Meat, Dairy and Valuation) had other non-GATT contracting parties as members, e.g., Botswana, Bulgaria, Guatemala, Lesotho and Paraguay (GATT, GATT Activities in 1986, 95).

118

William J. Davey

there is obviously no conflict. Similarly, an Article III violation that is excused by Article XX is consistent with a finding of no Article 2.2 violation. The third possibility is an Article III violation excused by Article XX, combined with a TBT Article 2.2 violation. Such a situation could arise if a measure were viewed as nondiscriminatory and not a disguised restriction on international trade in terms of the chapeau of Article XX, but as being more trade-restrictive than necessary to satisfy a legitimate objective under Article TBT 2.2. Obviously, not many measures would fall into this category since in the typical case a measure that is found not to be a disguised restriction on international trade would probably also be found not to be more trade restrictive than necessary to achieve one of the TBT Agreement’s specified legitimate objectives, but that result is not logically compelled. If the true conflict definition were applied, no conflict would be found in this situation and a violation of the TBT Agreement would be the result. If a broader notion of conflict were applied and a conflict found, under the WTO Agreement’s conflict rule, the TBT Agreement would prevail and a TBT violation would be found. While the situation giving rise to this sort of situation would not often occur, the latter result would be appropriate in my view. This is because, unlike the GATT III-TBT 2.1 case, where two similar if not identical nondiscrimination obligations are at issue, there are two quite different obligations involved in the GATT III-TBT 2.2 case – GATT’s nondiscrimination or national treatment rule and the TBT’s ban on unnecessarily trade restrictive measures. In my mind, the second obligation could be violated without calling into question the first. The fact that a measure is nondiscriminatory or not a disguised restriction on international trade would not preclude a finding that it is an unnecessary obstacle to trade. The lack of a GATT violation should not excuse a violation of an unrelated TBT obligation. The last possibility is a measure that violates Article III (and is not excused by Article XX) but is permitted under TBT Article 2.2. Such a situation could arise if a measure were found to be discriminatory for purposes of Article III and the chapeau of Article XX, although not creating unnecessary obstacles to international trade or being unnecessarily trade restrictive. While it may be likely that such a discriminatory measure would be viewed as an unnecessary obstacle to trade or unnecessarily restrictive, it would not necessarily have to be so. If the true conflict rule were applied, and a viola-

Principles Governing the Interrelation of the WTO Agreements 119

tion of GATT but not of the TBT Agreement were found, there would be no conflict, so the GATT violation would stand notwithstanding the absence of a TBT violation. If a broader view of conflict were taken, then it might be argued that a conflict exists and that under the WTO Agreement’s conflict rule, the TBT Agreement prevails over GATT, which would mean that no violation would be found. For me, it would not be appropriate to find a conflict in this situation. As in the immediately preceding case, the concern of GATT – discrimination – is distinct from the concern of the TBT Agreement, which is focused on the trade effect of the measure and whether it is necessary. In any event, of course, such a measure would presumably violate TBT Article 2.1 (since it violates GATT Article III), but the point to be made is that compliance with TBT Article 2.2 should not excuse unrelated violations, such as the prohibition on discriminatory measures. The analysis is made more complicated if GATT Article XI is considered. It is arguable, of course, that in the situation we are considering – the application of a technical regulation to both domestic and imported products – Article XI should not be applied, but rather the analysis should be conducted under Article III and that is the approach that I would adopt.33 If that is not the approach adopted, then it would be necessary to consider situations where application of GATT Articles XI and XX, on the one hand, and TBT Article 2.2, on the other, led to different results. In examining the structure of GATT and the TBT Agreement, I think that the TBT Agreement could be viewed as dealing with a special case dealt with under Article XI – prohibitions on imports as a result of technical regulations. As such, the TBT Agreement is lex specialis in this situation, such that a measure permitted by the TBT Agreement should be deemed permissible, notwithstanding any analysis under Articles XI and XX. The main impact would be that a measure not justifiable under GATT Article XX, but meeting the terms of TBT Article 2.2 would be permissible. This is different than the result of the similar situation in the Articles III and XX versus TBT Article 2.2 case, but it is explainable because in the latter case, GATT is concerned with discrimination – something not a concern of TBT Article 2.2, but in the former case it is only concerned with a restriction on imports – something that, for technical regulations,

33

See note 2 supra.

120

William J. Davey

is the primary concern of Article 2.2. Thus, the different result is justified. These issues are obviously difficult, particularly as to Articles III and XX and TBT Article 2.1, which explains why panels have avoided them. It is arguable that neither the narrow or broad definition of conflict necessarily leads to the right result. I find it preferable to start with the broad definition, however, and use interpretative methods to find no real conflict in appropriate cases.34 4. Summary One final thought on conflicts: For whatever reason, the issue of conflicts between WTO agreements has not much arisen in WTO disputes. Indeed, in two of the cases I have discussed – the GATT/ATC conflict and the TRIMs/Subsidies conflict, the issue will disappear as the relevant transitional periods expire. Nonetheless – even if the issue arises infrequently – which probably reflects the fact that the WTO agreements are in general internally consistent – it will undoubtedly arise from time to time. One area where such conflicts seem likely to be inevitable involves the interaction of the three general agreements – GATT, GATS and TRIPS – because the coverage of each of these agreements is expressed in very broad terms. The coverage of GATT has long been interpreted expansively, particular with respect to its MFN and national treatment obligations. The scope of GATS has been defined broadly as applying to measures affecting trade in services, while the MFN and national treatment provisions of TRIPS apply generally to the protection of intellectual property.35 It seems clear that in the future the uncertain rules on treating conflicts in the WTO agreements will be tested regularly, even if not frequently. For the moment, the treatment of conflicts remains an area where the rules are quite uncertain. However, as to the problem of overlapping obligations – it has arisen more often and seems likely to continue to arise on a more regular basis.

34 35

For a somewhat different view of how to resolve GATT-TBT conflicts, see Marceau / Trachtman (2002). On this topic, see Davey / Zdouc (2003).

Principles Governing the Interrelation of the WTO Agreements 121

III. Overlapping Obligations in the WTO Agreement In considering the question of overlapping obligations, there is first a problem of terminology. Such obligations are generally thought of as not conflicting, but rather consisting of two obligations, both of which could be complied with, but where it can be argued that one of the obligations is more basic and somehow replaces the other obligations. This sort of situation arose early in WTO dispute settlement, and, generally speaking, the Appellate Body has concluded that overlapping obligations should be viewed as cumulative. Several examples will serve to illustrate the kind of issues that arise. The first case of this sort was the Canada Periodicals case.36 There, Canada argued that since it had made no specific commitment in GATS on advertising services, it could not be held responsible under GATT for a measure regulating advertising that indirectly affected trade in goods. The Appellate Body held, however, that the creation of GATS did not carve out an exception to the preexisting coverage of GATT. Moreover, it questioned where the measure at issue was really directly aimed at advertising services that incidentally affected periodicals since the measure was a tax that was levied on periodicals. Soon thereafter, the relationship of GATT and GATS arose in the context of the Bananas case.37 The EC argued in essence that a measure taken to regulate trade in goods could not be inconsistent with that member’s GATS obligations. The argument was rejected given the broad scope of the GATS obligations, which apply generally to measures ‘affecting’ trade in services. This broad definition does not suggest that only regulatory measures aimed at a particular service sector are covered. In the view of the Appellate Body, there could be three different types of measures – those affecting only trade in goods, those affecting only trade in services and those affecting both, such as measures involving a good that relates to a particular service or a service supplied in conjunction with a particular good. In its view, in the third case, the measures at issue 36

37

Appellate Body Report on Canada – Certain Measures Concerning Periodicals, WT/DS31/AB/R, at pp. 17-20, adopted by the WTO Dispute Settlement Body on 30 July 1997. Appellate Body Report on European Communities – Regime for the Importation, Distribution and Sale of Bananas, WT/DS27/AB/R, paras. 217-222, adopted by the WTO Dispute Settlement Body on 25 September 1997.

122

William J. Davey

could be examined under both GATT and GATS. The Appellate Body noted, however, that the specific focus of the examination of the measure under each agreement would likely be different. Under GATT, the focus would be on how the measure affects the goods involved, while under GATS the focus would be on how the same measure affects the supply of the service or the service supplier involved. This notion that overlapping obligations should be viewed as cumulative has found expression in other cases as well. Perhaps its most controversial application has been in the line of decisions where the Appellate Body has ruled that a requirement of GATT Article XIX on safeguards – that they may be justified only if increased imports have resulted from unforeseen developments – must still be complied with even though it is the only requirement of Article XIX that was not carried over into the 1994 Agreement on Safeguards.38 The result is controversial because it seems that many negotiators thought that the decision not to mention the requirement in the new agreement meant that it was no longer applicable. Technically, however, it is difficult to argue with the Appellate Body’s analysis, especially given the reasoning in the Periodicals and Bananas cases. For me, these results that impose overlapping or cumulative obligations seem appropriate. They essentially go back to the idea that a treaty interpreter should avoid reading provisions out of a treaty. Where there is no question of true conflict, nor any question of an obligation such as an explicit prohibition in tension with a right such as an explicit permission, it seems appropriate to apply each obligation. Of course, as the Appellate Body found in Bananas,39 it may be appropriate and sufficient to find only a violation of only one agreement as a matter of judicial economy, but that 38

39

Appellate Body Report on Argentina – Safeguard Measures on Imports of Footwear, WT/DS121/AB/R, paras. 76-97, adopted by the WTO Dispute Settlement Body on 12 January 2000; Appellate Body Report on Korea – Definitive Safeguard Measure on Imports of Certain Dairy Products, WT/DS98/AB/R, paras. 68-90, adopted by the WTO Dispute Settlement Body on 12 January 2000. Appellate Body Report on European Communities – Regime for the Importation, Distribution and Sale of Bananas, WT/DS27/AB/R, para. 204, adopted by the WTO Dispute Settlement Body on 25 September 1997.

Principles Governing the Interrelation of the WTO Agreements 123

does not and should not imply that the other obligation was inapplicable. IV. The Consequences of the Cancún Failure on the Coherence of International Economic Law This volume of papers is particularly concerned with the effect of the failure of the WTO’s Cancún ministerial conference on the future of the WTO.40 Thus, I will conclude with a few thoughts on my view of the consequences of that failure on the coherence of international economic law. As noted earlier, two possible consequences of the failure could be a rise in the use of plurilateral agreements on controversial topics such as competition and investment and an increase in the number of regional trade agreements. Both results would obviously undermine the coherence of international economic law as represented by the WTO system since WTO members would be subject to different obligations within the WTO system and there would be even more variation likely in the regional trade agreements that were exempt from many WTO rules. To the extent that new plurilateral agreements are in truly new areas that are distinct from current WTO rules, I think that it could be argued that they would not have a major impact on the WTO, but it is not at all clear that the two most likely plurilateral agreements – competition and investment – would be so distinct. In respect of competition, one could imagine that many of the benefits of an agreement would have to be extended to non-participants, even though they undertook no obligations, because of GATT’s most-favored-nation requirement, which would apply to laws and regulations affecting the sale of goods and services – as at least some rules on competition would certainly be. Similarly, given the extent to which the TRIMs Agreement and, more significantly, the Services Agreement deal with investment, any agreement on investment may present overlap problems and potential conflict problems. It is too early to speculate on specific problems, however, as the structure and content of any agreements is yet unknown. At the moment, the number and variety of regional trade agreements seems to be increasing rapidly and the Cancún failure will likely accelerate that trend. The direct effect of such agreements on the coherence of WTO rules may not be so great, but the huge vari40

See note 6 supra.

124

William J. Davey

ety in rules that are contained in regional agreements will certainly undermine the coherence of international economic law. Regional trade agreements by their very nature involve rules of origin that are likely to vary on an agreement-by-agreement basis since different products will likely be viewed differently in terms of their sensitivity under the different agreements. Application of those rules of origin greatly complicates the business of international trade – they are often very complex and difficult to understand since they are designed to protect certain producers; and they are difficult to comply with – sometimes substantively and always procedurally, such that the expected benefits of regional trade agreements are often not fully achieved. Moreover, although entities like the United States and the EC tend to use past regional trade agreements as models, they all tend to have some distinct rules, incorporated because of a special issue that the parties want to deal with or because they are viewed as improvements over the old model agreement. The result is the opposite of the goal of trade facilitation that is supposedly desired by the major players in the WTO. I am hopeful that the rush to regional trade agreements will not undermine the multilateral system and I think it will not principally for three reasons. First, a proliferation of regional trade agreements will make doing business internationally more difficult because of the disparate rules that will be found in those regional trade agreements. These added complexities will make the uniformity of multilateral approaches that much more attractive. Second, while bilateral or regional trade agreements can and do make progress on market access issues, they usually fail to deal with truly difficult trade issues, such as agriculture and dumping – to name only two. Those difficult issues will remain unresolved and suitable for resolution only in more broad-based negotiations in the WTO. This is particularly true for the major players – such as the Quad (the US, EC, Canada and Japan) and the leading developing countries – for whom the WTO is the main arena in which they manage their trade relations with each other. Third, those same major players are the main users of the WTO’s dispute settlement system, even where, as is the case between Canada and the United States, they have a free trade agreement between them with its own WTO-like dispute settlement system. For them, the existence of an effective dispute settlement is essential to ensure that agreements are respected. Yet, dispute settlement in free trade areas is typically not so effective, as is demon-

Principles Governing the Interrelation of the WTO Agreements 125

strated by the general disuse of NAFTA’s general dispute settlement system.41 Thus, the failure of Cancún will likely lead to more regional trade agreements and thereby increase incoherence, but the long term strength and advantages of the multilateral trading system – relative simplicity over complexity, the ability to deal with the hard issues and more effective dispute settlement – will ultimately reverse this trend away from coherence. V. Conclusion In conclusion, in discussing the consistency of the WTO system, I would say that the problem of conflicts within the system presents very interesting and difficult issues, but does not seem to have arisen often enough to present basic difficulties for the overall effectiveness of the WTO dispute settlement system. The problem of overlaps is more serious, but given the Appellate Body’s clear direction that overlaps are not conflicts and that overlapping obligations must all be complied with, the system seems to dealing with the overlap problem. Finally, the Cancún failure will ultimately lead to more plurilateral and regional trade agreements, which inevitably will undermine the coherence of international economic law, but I think that in the end the multilateral system will remain the most attractive alternative which suggests that, notwithstanding the Cancún setback, more coherence can be expected over time.

41

The US has initiated four cases against Canada and five against Mexico as of March 16, 2003 at the WTO; Canada has initiated 11 against the US (many involving their lumber dispute), while Mexico has initiated five against the US. See data on WTO website – www.wto.org – under ‘Disputes’. Most of the cases have resulted in panel proceedings. In contrast, NAFTA’s general dispute settlement mechanism – the Chapter 20 procedure – has not been much used. There has been one case by the US against Canada, the result of which was inconclusive from the US point of view. There have also been two cases by Mexico against the US. Thus, there have been 25 proceedings in the WTO, but only three in NAFTA. See data on Chapter 20 disputes at NAFTA Secretariat website – www.nafta-secalena.org.

126

William J. Davey

References John Croome (1999), Reshaping the World Trading System: A History of the Uruguay Round, Den Haag (Kluwer) 1999. William J. Davey / Werner Zdouc (2003), The Triangle of TRIPS, GATT and GATS, in: Thomas Cottier / Petros C. Mavroidis (eds.), Intellectual Property: Trade, Competition and Sustainable Development, Ann Arbor, MI (The University of Michigan Press) 2003, 53-84. Wilfred Jenks (1953), Conflict of Law-Making Treaties, in: British Yearbook of International Law 30 (1953), 401-453. Gabrielle Marceau (2001), Conflicts of Norms and Conflicts of Jurisdiction: The Relationship between the WTO Agreement and MEAs and other Treaties, in: Journal of World Trade, 35 (2001) 6, 1081-1131. Gabrielle Marceau / Joel P. Trachtman (2002), The Technical Barriers to Trade Agreement, the Sanitary and Phytosanitary Measures Agreement, and the General Agreement on Tariffs and Trade: A Map of the World Trade Organization Law of Domestic Regulation of Goods, in: Journal of World Trade, 36 (2002) 5, 811-881. Joost Pauwelyn (2003), Conflict of Norms in Public International Law, Cambridge (Cambridge University Press) 2003. Pierre Pescatore / William J. Davey / Andreas F. Lowenfeld (1991), Handbook of GATT Dispute Settlement, Ardsley-on-Hudson / New York (Transnational Juris Publications) 1991. World Trade Organization (1995), The Results of the Uruguay Round of Multilateral Trade Negotiations: The Legal Texts, Geneva 1995. WTO Cases Argentina – Safeguard Measures on Imports of Footwear, WT/DS121, adopted by the WTO Dispute Settlement Body on 12 January 2000 Canada – Certain Measures Concerning Periodicals, WT/DS31, adopted by the WTO Dispute Settlement Body on 30 July 1997.

Principles Governing the Interrelation of the WTO Agreements 127

European Communities – Measures Affecting Asbestos and Asbestos-Containing Products, WT/DS135, adopted by the WTO Dispute Settlement Body on 5 April 2001. European Communities – Regime for the Importation, Distribution and Sale of Bananas, WT/DS27, adopted by the WTO Dispute Settlement Body on 25 September 1997. Indonesia – Certain Measures Affecting the Automobile Industry, WT/DS54, adopted by the WTO Dispute Settlement Body on 23 July 1998. Korea – Definitive Safeguard Measure on Imports of Certain Dairy Products, WT/DS98, adopted by the WTO Dispute Settlement Body on 12 January 2000. United States – Standards for Reformulated and Conventional Gasoline, WT/DS2, adopted by the WTO Dispute Settlement Body on 20 May 1996.

Lorand Bartels*

Treaty Conflicts in WTO Law – A Comment on William J. Davey’s Paper ‘The Quest for Consistency’ I. II. III. IV. V.

Introduction Two Definitions of Treaty Conflicts WTO Panels on Treaty Conflicts Problems with a Narrow Definition of Conflicts Appellate Body Jurisprudence A. A Right Subject to Cumulative Conditions (Argentina – Footwear (EC)) B. A Right Subject to a Condition and a ‘Noncondition’ (Guatemala – Cement I) C. A Right Subject to Cumulative Conditions (again) (US – FSC) D. Conflict between an Obligation and a ‘Nonobligation’ (US – Hot Rolled Steel) E. Conflict between an Obligation and a Right (Brazil – Aircraft; US – FSC) 1. Brazil – Aircraft 2. US – FSC VI. Conclusions References

*

130 130 131 132 135 135 136 139 140 141 141 143 144 145

University of Cambridge / UK, currently Humboldt Fellow, Max Planck Institute for International Law, Heidelberg. I am grateful to Erich Vranes and Joost Pauwelyn for discussion and comments on this article.

130

Lorand Bartels

I. Introduction This contribution concentrates on how to determine when there is a conflict between WTO norms, such that it is necessary to apply the rules of international law on treaty conflict.1 II. Two Definitions of Treaty Conflicts In his contribution, Professor Davey draws attention to two distinct methods of defining a ‘conflict’ between norms in different treaties. These are as follows: 1. Conflict can only exist between obligations. In this case it is impossible for a party to act in compliance with all applicable obligations. 2. Conflict may also exist between an obligation and an express right. In this case, it is possible for a party to act in compliance with its obligations, but only at the expense of not exercising a right expressly granted under some applicable legal provision. This is, of course, an oversimplification of the issue, given that not all legal relations can be encapsulated in a rights-obligations dichotomy. In addition to these, legal norms include at least powers and norms regulating other norms.2 Furthermore, as will be illustrated later in this contribution, it is strongly arguable that a condition on the exercise of a right should not be treated as an obligation, given that it does not actually require any performance but merely limits either the exercise of the right (by establishing express preconditions) or its terms (be incorporating these preconditions in the

1

2

This contribution does not discuss the rules applicable to resolve conflicts. These rules are constituted predominantly by the principles of lex specialis (the more specific norm prevails) and lex posterior (the norm later in time prevails), subject to the legitimate interests of third parties and subject to the express terms of the relevant treaties themselves. The Vienna Convention on the Law of Treaties codifies certain of these principles but is by no means conclusive on the issue, nor on the consequences of any breach of obligations as a result of treaty conflict. In particular, the Vienna Convention fails to account for the principle of lex specialis, which has a definite, though perhaps ambiguous, history in the jurisprudence of international tribunals and academic writings. For a list, see Pauwelyn (2003), 158 et seq.

Treaty Conflicts in WTO Law

131

definition of the right).3 Nevertheless, though simplified, this dichotomy does represent the terms of the debate in the academic literature, and in particular, in relevant WTO cases. The following will give an account of the current state of play, before making some suggestions on the appropriate test that should be adopted in the WTO legal system. III. WTO Panels on Treaty Conflicts The question whether a narrow or wide definition of treaty conflict should be adopted in WTO law has not yet been determined. At present, one panel has opted for the wide view, another for the narrow view, and the Appellate Body has not yet ruled on the issue, though some indications of its views may be gleaned from some of its reports. The issue of treaty conflicts was first addressed by the panel in EC – Bananas III, which considered that there could be a situation of conflict between a right and an obligation.4 The Panel gave the example, in a footnote, of the express right of a WTO Member under Article 2 of the Agreement on Textiles and Clothing to impose quantitative restrictions that would be prohibited by Article XI:1 of the GATT.5 By contrast, in the context of analyzing the relationship between the prohibition on local content rules in the TRIMS Agreement (an obligation) and the transitional period for local content rules in the SCM Agreement (a right), the panel in Indonesia – Automobiles rejected this approach. This Panel limited the notion of ‘conflict’ to a situation in which it is impossible for a party to comply with different obligations at the same time. Specifically, the panel said as follows: ‘In international law for a conflict to exist between two treaties, three conditions have to be satisfied. First, the treaties concerned must have the same parties. Second, the treaties must cover the same subject matter. … Third, the 3 4

5

Cf ibidem, 160. WTO Panel Report EC – Bananas III (US), WT/DS27/R/USA, adopted as modified by the Appellate Body report on 25 September 1997, para 7.159. Ibidem, fn 728.

132

Lorand Bartels

provisions must conflict, in the sense that the provisions must impose mutually exclusive obligations.’6 The panel went on to adopt the following quotation from the Encyclopædia of Public International Law: ‘Technically speaking, there is a conflict when two (or more) treaty instruments contain obligations which cannot be complied with simultaneously. … Not every such divergence constitutes a conflict, however, … Incompatibility of contents is an essential condition for conflict.’7 In his paper, Professor Davey aligns himself with the view of the Bananas panel, although for a number of reasons he supports the result of the panel in Indonesia – Automobiles.8 This contribution will explore further the validity of the narrow definition of treaty conflicts adopted by the panel in Indonesia – Automobiles, and for that matter subsequently approved in Turkey – Textiles.9 IV. Problems with a Narrow Definition of Conflicts There are a number of problems with the narrow definition of conflicts adopted by the panel in Indonesia – Automobiles. In the first 6 7 8

9

WTO Panel Report, Indonesia – Automobiles, WT/DS54/R, adopted 23 July 1998, para 14.28, fn 649. Encyclopædia of Public International Law, Vol 7 (North-Holland, 1984) at 468, quoted ibidem, para 14.28, fn 649. Professor Davey distinguishes between the rights in the ATC (protected by the Bananas III panel) and those in the SCM Agreement (unprotected by the Indonesia – Automobiles panel), and argues that whereas the former are a clear expression of the intent of the parties, the latter are at best a minor part of that agreement, and therefore could cede to the more important rules set out in the TRIMs Agreement. This is supported, according to Davey, by the temporal coordination between the SCM Agreement and the transitional period in the TRIMS agreement. In addition, Davey notes the different consequences of violation of the TRIMS and SCM Agreement after the expiry of the transitional period. Finally, with reference to the principle of lex specialis, Davey refers again to the minor nature and different purpose of the transitional period in TRIMS Agreement. WTO Panel Report, Turkey – Textiles, WT/DS34/R, adopted as modified by the Appellate Body report on 19 November 1999, paras 9.92-9.95.

Treaty Conflicts in WTO Law

133

place, the panel’s proposition that international law mandates a narrow definition of ‘conflict’ is not unambiguously true. While some eminent authorities, most notably Wilfred Jenks, take the narrow view, there are others equally eminent that have taken the wide view, including Humphrey Waldock, Hans Aufricht and Hersch Lauterpacht.10 Even more importantly, perhaps, the jurisprudence of international tribunals is, at the very least, undecided on the issue.11 This is perhaps unsurprising, because there is a very strong policy argument in favour of a wide definition of conflict. This is well explained by Joost Pauwelyn, who, in the context of discussing the narrow view of Wilfred Jenks, says as follows: ‘[B]y refusing to recognise certain situations as conflicts – such as a contradiction between a prohibition to do X and permission to do X) – Jenks’ strict definition indirectly resolves a number of contradictions in favour of the strictest norm, in casu, in favour of the prohibition to do X, since not invoking the right to do X under the permissive norm will avoid breaching the prohibition to do X in the prohibitive norm. In those situations, the alleged conflict is then not solved by a rule on how to solve conflict but by the very definition of conflict.’12 These observations are particularly pertinent in the context of the rights and obligations contained in the various WTO agreements, which, as the Appellate Body said in Argentina – Footwear (EC), form part of ‘one treaty, the WTO Agreement’.13 The consequence of this, as the Appellate Body said in the same paragraph, is that ‘an appropriate reading of this “inseparable package of rights and disciplines” must, accordingly, be one that gives meaning to all the relevant provisions of these two equally binding agreements.’ Allied to this understanding of the WTO ‘package’ is the proposition, accepted by the Appellate Body, that the provisions of 10 11 12 13

For a comprehensive discussion see Pauwelyn (2003) 166-169 and Vranes (2004). See Pauwelyn (2003), 340 et seq.; and Vranes 2004. Pauwelyn (2003), 170 et seq. WTO Appellate Body Report, Argentina – Footwear (EC), at para 81 (referring to Article XIX GATT and the Safeguards Agreement).

134

Lorand Bartels

the WTO agreements represent a carefully negotiated ‘balance of rights and obligations’ between WTO Members that must be respected.14 Indeed, in US – Shrimp the Appellate Body went so far as to refer to a balance between the competing rights of different WTO Members when it would have been more natural to refer to a balance between the rights and obligations of the same WTO Member. In the context of describing the Chapeau to Article XX GATT, the Appellate Body said the following: ‘Exercise by one Member of its right to invoke an exception, such as Article XX(g), if abused or misused, will, to that extent, erode or render naught the substantive treaty rights in, for example, Article XI:1, of other Members.’15 In the specific context of dispute settlement, this emphasis on preserving the rights of WTO Members in dispute settlement is reinforced by Article 3.2 DSU, which states that ‘[r]ecommendations and rulings of the DSB cannot add to or diminish the rights and obligations provided in the covered agreements’ and Article 19.2 DSU, which states that ‘[i]n accordance with paragraph 2 of Article 3, in their findings and recommendations, the panel and Appellate Body cannot add to or diminish the rights and obligations provided in the covered agreements.’ If both WTO rights and WTO obligations are equally deserving of protection, then it seems fair to say that, in a situation in which it is not possible to protect both at the same time, at the very least a reasoned choice must be made between equally applicable rights and obligations. This possibility would be precluded by the application of a conflicts rule that ipso facto diminishes WTO rights in favour of WTO obligations. This speaks very much in favour of the adoption of a wide definition of treaty conflicts in WTO law.

14 15

See also Pauwelyn (2003) 197 et seq. WTO Appellate Body Report, US – Shrimp, at para 156. Later in this paragraph the Appellate Body reformulated this sentence in more conventional terms. The Appellate Body similarly refers to the balance of rights and obligations in its reports on US – Underwear, at para 13 (concerning Article 6.10 ATC); EC – Bananas III, at para 136 (concerning a Member’s right to bring a claim); and Brazil – Aircraft, at para 139 (in relation to Article 3.1(a) and 27.2 and 27.4 SCM).

Treaty Conflicts in WTO Law

135

It might also be added that a similar view seems to be taken by the WTO Membership. It is otherwise very difficult to know what to make of the notorious ‘sequencing’ problem, which concerns a conflict between the right of one Member to request authorisation from the Dispute Settlement Body to suspend concessions under Article 22 DSU and the right of another Member to have its implementation of the recommendations and rulings of the DSB determined by a panel in accordance with Article 21.5 DSU.16 It simply makes no sense to see this conflict in terms of a conflict between two obligations. If WTO Members did not recognize that there could be a conflict between the rights of WTO Members under the covered agreements, then it is probably not going too far to say that the problem of sequencing would never have arisen. V. Appellate Body Jurisprudence The Appellate Body has on a number of occasions considered how to reconcile apparently competing norms, but it has never actually found it necessary to pronounce on the question whether a wide or narrow definition of conflict should prevail in WTO law. Nevertheless, a survey of its practice tends to indicate that the Appellate Body would be unlikely to adopt the narrow view of the panel in Indonesia – Automobiles. A. A Right Subject to Cumulative Conditions (Argentina – Footwear (EC)) One may begin with the two simultaneously adopted reports in Argentina – Footwear (EC) and Korea – Dairy.17 Both cases concerned the question whether the exercise of a right to take safeguards under the Agreement on Safeguards is subject to the condition, set out in Article XIX GATT, that safeguard measures may only be adopted if the increased imports are an ‘unforeseen development’. In both cases, the Appellate Body was able to deal with this question without much difficulty, stating that the Agreement on 16

17

See WTO Appellate Body Report, US – Certain EC Products, at paras 91-2, where the Appellate Body called these provisions ‘not a model of clarity’. See, however, Mavroidis (2004) for an interpretation reconciling these two provisions. WTO Appellate Body Report, Argentina – Footwear (EC), and WTO Appellate Body Report, Korea – Dairy.

136

Lorand Bartels

Safeguards applied to measures that were both defined in and conformed to Article XIX GATT.18 This was justified on the basis that Article 1 of the Safeguards Agreement clearly stipulates that: ‘This Agreement establishes rules for the application of safeguard measures which shall be understood to mean those measures provided for in Article XIX of GATT 1994.’ This was not therefore a situation in which a Member has a right to take an action under one agreement that would be nullified by a condition or obligation in another agreement. The right only ever existed subject to the condition in Article XIX. Though less elaborated, the same reasoning underlies the earlier ruling in Brazil – Desiccated Coconut, confirmed in US – Softwood Lumber, that countervailing duties may be imposed only in accordance with both the SCM Agreement and Article VI of GATT.19 B. A Right Subject to a Condition and a ‘Non-condition’ (Guatemala – Cement I) A slightly more complicated situation arose in Guatemala – Cement I.20 This case involved the express conflicts clause in Article 1.2 DSU, according to which the ‘special or additional rules and procedures [in other covered agreements] shall prevail’ over the provisions of the DSU ‘[t]o the extent that there is a difference between’ the two sets of provisions. On the notion of ‘difference’, the Appellate Body said this: ‘In our view, it is only where the provisions of the DSU and the special or additional rules and procedures of a covered agreement cannot be read as complementing each other that the special or additional provisions are to prevail. A special or additional provision should only be found to prevail over a provision of the DSU in a situation where adherence to the one provision will lead to a viola-

18 19

20

Argentina – Footwear (EC), para 83; Korea – Dairy, para 77. WTO Appellate Body Report, Brazil – Desiccated Coconut, 13 et seq.; WTO Appellate Body Report, US – Softwood Lumber, at para 134. WTO Appellate Body Report, Guatemala – Cement I.

Treaty Conflicts in WTO Law

137

tion of the other provision, that is, in the case of a conflict between them.’ 21 The Appellate Body added, one paragraph later, that: ‘It is, therefore, only in the specific circumstance where a provision of the DSU and a special or additional provision of another covered agreement are mutually inconsistent that the special or additional provision may be read to prevail over the provision of the DSU.’22 The allegedly ‘different’ provisions at issue were the special dispute settlement rules and procedures in Article 17 of the Antidumping Agreement (ADP), on the one hand, and the normal dispute settlement rules and procedures in the DSU, on the other. The Appellate Body began its analysis of the relationship between these provisions by making the following statement: ‘Clearly, the consultation and dispute settlement provisions of a covered agreement are not meant to replace, as a coherent system of dispute settlement for that agreement, the rules and procedures of the DSU. To read Article 17 of the Anti-Dumping Agreement as replacing the DSU system as a whole is to deny the integrated nature of the WTO dispute settlement system established by Article 1.1 of the DSU. … The Panel’s conclusion is reminiscent of the fragmented dispute settlement mechanisms that characterized the previous GATT 1947 and Tokyo Round agreements; it does not reflect the integrated dispute settlement system established in the WTO.’23 What was critical, therefore, was that there is only one right to request the establishment of a panel, and that right is found in the DSU.24 Consequently, it was open for the Appellate Body to find that any conditions on the exercise of this right in either the DSU or the ADP could be treated as cumulative, as was done in Argentina – Footwear (EC). 21 22 23 24

Ibidem, para 65 (italics added). Ibidem, para 66. Ibidem, para 63. Presumably in Article 6.1 DSU, which states that ‘[i]f the complaining party so requests, a panel shall be established’, though the Appellate Body never actually referred to this provision.

138

Lorand Bartels

But this left another question: what happens when the presence of a condition in one agreement is met with the absence of that same condition in the other? This was the question to be decided in this case. Article 6.2 DSU requires a complainant to ‘identify the specific measures at issue’, a condition that is not reflected in the special procedures under the Antidumping Agreement. Article 17.5 of this agreement provides instead for these other conditions: ‘The DSB shall, at the request of the complaining party, establish a panel to examine the matter based upon: (i) a written statement of the Member making the request indicating how a benefit accruing to it, directly or indirectly, under this Agreement has been nullified or impaired, or that the achieving of the objectives of the Agreement is being impeded, and (ii) the facts made available in conformity with appropriate domestic procedures to the authorities of the importing Member.’ The question was this: was the absence of the condition to identify specific measures in the Antidumping Agreement significant? The panel held that it was;25 the Appellate Body disagreed. This was probably the correct result, given the Appellate Body’s earlier finding on the ancillary nature of the procedures in Article 17 ADP. Unfortunately, however, the Appellate Body came to its conclusion for what were arguably irrelevant reasons. Instead of arguing that the absence of a condition in Article 17.5 does not render inapplicable the presence of the condition in Article 6.2, the Appellate Body based its analysis on the ‘additional requirements’ in Article 17.5 (concerning the written statement and facts provided to domestic authorities). It reasoned as follows: ‘The fact that Article 17.5 contains these additional requirements, which are not mentioned in Article 6.2 of the DSU, does not nullify, or render inapplicable, the specific requirements of Article 6.2 of the DSU in disputes brought under the Anti-Dumping Agreement. In our view, there is no inconsistency between Article 17.5 of the AntiDumping Agreement and the provisions of Article 6.2 of the DSU. On the contrary, they are complementary and 25

WTO Panel Report, Guatemala – Cement I, WT/DS60/R, adopted 25 November 1998, as modified by the Appellate Body report.

Treaty Conflicts in WTO Law

139

should be applied together. A panel request made concerning a dispute brought under the Anti-Dumping Agreement must therefore comply with the relevant dispute settlement provisions of both that Agreement and the DSU. Thus, when a “matter” is referred to the DSB by a complaining party under Article 17.4 of the Anti-Dumping Agreement, the panel request must meet the requirements of Articles 17.4 and 17.5 of the Anti-Dumping Agreement as well as Article 6.2 of the DSU.’26 This is somewhat beside the point. It is true that the existence of these additional requirements in Article 17.5 may be another reason for finding Article 17.5 not to prevail over Article 6.2 DSU. But it is not entirely irrelevant to ask whether these additional conditions might ‘nullify, or render inapplicable, the specific requirements of Article 6.2’. For this reason, this passage cannot simply be taken at face value.27 How then are we to understand the refusal of the Appellate Body to draw any normative inferences from the absence in the Antidumping Agreement of the condition in the DSU to ‘identify the specific measures at issue’? The most likely explanation is that the Appellate Body gave no normative weight to the absence of a condition in the Antidumping Agreement. In other words, it decided that a ‘difference’ between the provisions must be based on a positive norm, not simply the absence of a norm. C. A Right Subject to Cumulative Conditions (again) (US – FSC) A further case of potential conflict arose in US – FSC.28 In this case, there was a question whether there was a relevant ‘difference’ between Article 4.4 DSU, which requires a request for consultations to ‘give reasons for the request, including identification of the 26 27

28

Guatemala – Cement I, para 75. The Appellate Body also supported its reasoning by reference to Article 17.4 ADP, which lists three types of antidumping measure which can be challenged under the dispute settlement procedures in Article 17 ADP: Ibidem, para 79. But Article 17.4 merely states which measures may be the subject of a claim under the Antidumping Agreement; it does not require these measures to be specifically identified in such a claim. WTO Appellate Body Report, US – FSC.

140

Lorand Bartels

measures at issue and an indication of the legal basis for the complaint’ and Article 4.2 SCM, which states that ‘[a] request for consultations under paragraph 1 [of Article 4] shall include a statement of available evidence with regard to the existence and nature of the subsidy in question.’ The question before the Appellate Body was whether a request for consultations had to comply with all three of these conditions. Quoting its test in Guatemala – Cement I, the Appellate Body said that ‘[i]t is clear to us that Article 4.4 of the DSU and Article 4.2 of the SCM Agreement can and should be read and applied together, so that a request for consultations relating to a prohibited subsidy claim under the SCM Agreement must satisfy the requirements of both provisions.’29 This is a clear reiteration of the proposition that there is only one right to request consultations. Once this was accepted, then it followed from the above cases that such a right must be subject to all applicable conditions. D. Conflict between an Obligation and a ‘Non-obligation’ (US – Hot Rolled Steel) In US – Hot-Rolled Steel the Appellate Body was again faced with the more complicated situation arising in Guatemala – Cement I, although this time the alleged ‘difference’ was not between a condition and a ‘non-condition’, but between an obligation and a ‘nonobligation’. Specifically, the ‘difference’ was between Article 11 DSU, under which panels are required to make ‘an objective assessment of the matter’, and Article 17.6 ADP, under which panels are under obligations to review the factual findings (Article 17.6(i)) and legal findings (Article 17.6(ii)) of domestic antidumping authorities.30 The question, as in Guatemala – Cement I, was whether the absence of the obligation to ‘make an objective assessment of the matter’ in Article 17.6 ADP was significant. In both cases, the Appellate Body held that it was not. Speaking of the first of these special provisions, the Appellate Body held that even though it contained no reference to the standard set out in Article 11 DSU, ‘it is inconceivable that Article 17.6(i) should require anything other than that panels make an

29 30

Ibidem, at para 160. WTO Appellate Body Report, US – Hot Rolled Steel.

Treaty Conflicts in WTO Law

141

objective “assessment of the facts of the matter”.’31 As to the latter, the Appellate Body said, similarly, that ‘[n]othing in Article 17.6(ii) of the Anti-Dumping Agreement suggests that panels examining claims under that Agreement should not conduct an ‘objective assessment’ of the legal provisions of the Agreement, their applicability to the dispute, and the conformity of the measures at issue with the Agreement.’32 Following on from Guatemala – Cement I, this is perhaps not a surprising result. As in that case, the Appellate Body decided that there cannot be a ‘difference’ within the meaning of Article 1.2 DSU, consisting of a norm in one set of dispute settlement rules and the absence of that norm in the other. E. Conflict between an Obligation and a Right (Brazil – Aircraft; US – FSC) Two apparently more conventional cases are Brazil – Aircraft, in which the Appellate Body was once again confronted with a potential ‘difference’ between a special dispute settlement procedure and the DSU, and US – FSC, which involved the substantive provisions of the GATT and the Agreement on Agriculture. In both cases, the Appellate Body decided that an obligation in one agreement prevailed over a right in the competing agreement. However, it is not possible to draw any broader conclusions from this – for instance, that the Appellate Body has decided in favour of a narrow definition of conflicts – because in both cases the agreement containing the obligation was itself expressed to prevail over the agreement containing the right. 1. Brazil – Aircraft The two provisions at issue in Brazil – Aircraft were Article 4.7 of the Agreement on Subsidies and Countervailing Measures and Article 21.3 of the DSU. Article 4.7 SCM provides that: ‘If the measure in question is found to be a prohibited subsidy, the panel shall recommend that the subsidizing Member withdraw the subsidy without delay.’ Article 21.3 DSU states that:

31 32

Ibidem at para 55. Ibidem, at para 62.

142

Lorand Bartels

‘If it is impracticable to comply immediately with the recommendations and rulings, the Member concerned shall have a reasonable period of time in which to do so.’ Article 21.3 provides three methods of determining the reasonable period of time, and paragraph (c) states that if this period is determined by arbitration, it should normally be no longer than 15 months. Also relevant is Article 4.12 SCM, which halves the time period normally prescribed under the DSU for cases brought under the SCM Agreement, except where a time period is expressly prescribed in Article 4 SCM. The allegedly ‘different’ provisions as issue were the special dispute settlement rules and procedures in Article 17 of the Antidumping Agreement (ADP), on the one hand, and the normal dispute settlement rules and procedures in the DSU, on the other. The question in this case was whether the panel was correct to give Brazil 90 days to withdraw its subsidy, under Article 4.7 SCM, as opposed to half of the longer period specified in Article 21.3(c) DSU (as read in accordance with Article 4.12 DSU). Unlike in the cases discussed above, the Appellate Body decided here that: ‘there is a significant difference between the relevant rules and procedures of the DSU and the special or additional rules and procedures set forth in Article 4.7 of the SCM Agreement.’33 Despite the fact that the Appellate Body did not refer to Article 1.2 DSU, this language that it uses is clearly drawn from this provision. Indeed, it is only on this basis that the Appellate Body could have gone on to say that: ‘… the provisions of Article 21.3 of the DSU are not relevant in determining the period of time for implementation of a finding of inconsistency with the prohibited subsidies provisions of Part II of the SCM Agreement.’34 The Appellate Body can be faulted for not mentioning Article 1.2 DSU, when this provision was clearly being applied, and for not explaining in more detail why Article 4.7 SCM is relevantly ‘different’ from Article 21.3 DSU. What is more interesting, though, are 33 34

Brazil – Aircraft, at para 192. Ibidem, at para 192. The Appellate Body also noted in this paragraph that Article 4.12 was not applicable in this case.

Treaty Conflicts in WTO Law

143

the types of legal norms involved, and the Appellate Body’s choice between these norms. Both Article 4.7 SCM and Article 21.3 DSU establish time-periods within which a Member must comply with its obligation to implement DSB rulings. Under Article 4.7 SCM a panel must require the measure to be withdrawn ‘without delay’, which means that the affected Member has an obligation to withdraw its measure without delay. By contrast, under Article 21.3 DSU a Member has not only a right to a reasonable period of time for implementation if it is impractical to comply immediately,35 but in addition a procedural right, in Article 21.3(c), to have this reasonable period of time determined by arbitration. The alleged conflict was not therefore between two obligations, but between an obligation (to withdraw the measure without delay) and a right (to a reasonable period of time for implementation). Concretely, then, the Appellate Body should have asked whether the obligation flowing from the panel’s duty in Article 4.7 SCM could be read as erasing the express rights set out in Article 21.3 DSU. The fact that it did not is regrettable, but there can be little controversy as to the result: the express wording of Article 4.7 SCM makes it clear that this provision does, in fact, erase any other rights in Article 21.3 DSU. 2. US – FSC In US – FSC, the Appellate Body contrasted the ‘disciplines’ on export subsidies under the Agreement on Agriculture with the permissive exception for such subsidies on ‘primary products’ under Article XVI:4 GATT. It said: ‘Unquestionably, the explicit export subsidy disciplines, relating to agricultural products, contained in Articles 3, 8, 9 and 10 of the Agreement on Agriculture must clearly take precedence over the exemption of primary products from export subsidy disciplines in Article XVI:4 of the GATT 1994.’36 Although the Appellate Body did not refer to tem, the result is justifiable on the basis of two conflicts provisions. First, the General Interpretive Note to Annex 1A states that: 35

See Vranes (2004).

36

US – FSC, at para 117.

144

Lorand Bartels

‘In the event of conflict between a provision of the General Agreement on Tariffs and Trade 1994 and a provision of another agreement in Annex 1A to the Agreement Establishing the World Trade Organization (referred to in the agreements in Annex 1A as the ‘WTO Agreement’), the provision of the other agreement shall prevail to the extent of the conflict.’ Second, Article 21 of the Agriculture Agreement states that: ‘The provisions of GATT 1994 and of other Multilateral Trade Agreements in Annex 1A to the WTO Agreement shall apply subject to the provisions of this Agreement.’ On this basis, this case resolves into a conflict between an obligation and a right, in which the former prevailed over the latter. The weakness of the Appellate Body’s reasoning, as in Brazil – Aircraft, is that it did not state what it was doing.37 VI. Conclusions If one looks at these reports as a whole, a number of propositions emerge. First, the Appellate Body stated in Argentina – Footwear (EC) that where there is only one applicable right, it will be subject to all conditions relevant to the exercise of that right. Second, the Appellate Body has indicated that there is no normative conflict in situations when a condition or obligation is present in one agreement (Guatemala – Cement I; US – Hot-Rolled Steel). In this sense a condition or obligation will necessarily prevail over an equally applicable ‘non-condition’ or ‘non-obligation’. And third, an obligation will prevail over a right, at least in case where the obligation is contained in an agreement which is expressly stated to prevail over the agreement containing the right (Brazil – Aircraft; US – FSC). From this we can draw the following general conclusions on the definition of treaty conflicts in WTO law. While the Appellate Body has not ruled expressly on whether a narrow or wide definition of conflict should prevail in the WTO legal system, it does recognize (sometimes implicitly) that there is a wide variety of norms in the WTO legal system. Whether the Appellate Body will establish a hierarchy of obligations over other norms is hard to 37

See also Pauwelyn (2003), 197.

Treaty Conflicts in WTO Law

145

predict, but on the basis of what has been done so far, it seems unlikely. This would be a pleasing result. Not only is it questionable whether public international law mandates any narrow test for conflicts, but a narrow test has the disadvantage that it ipso facto determines any conflict between a right and an obligation in favour of the obligation. This is inadequate as a matter of general policy, but even more particularly it is problematic in the case of the WTO legal system, which is based on a ‘balance of rights and obligations’. It is therefore to be hoped, and indeed it might even be expected, that, in an appropriate case, the Appellate Body will clearly state that a right can prevail over an inconsistent obligation or condition, and thereby lay to rest the misunderstanding that obligations are somehow the most superior of legal norms.

References Petros C. Mavroidis (2004), Proposals for Reform of Article 22 of the DSU: Reconsidering the ‘Sequencing’ Issue and Suspension of Concessions (Taking Care of Business in Geneva), in: Federico Ortino / Ernst-Ulrich Petersmann (eds.), The WTO Dispute Settlement System, London / Den Haag (Kluwer) 2004. Joost Pauwelyn (2003), Conflict of Norms in Public International Law: How WTO Law Relates to Other Rules of International Law, Cambridge (Cambridge University Press) 2003. Erich Vranes (2004), The Definition of ‘Norm Conflict’ in International Law and Legal Theory, unpublished manuscript, 2004, on file with the author.

Peter-Christian Müller-Graff*

Protectionism or Reasonable National Regulation? The Protection of Non-Economic Interests as Barriers to the Free Movement of Goods: A Comparison of EC Law and WTO Law I. II.

Introduction Relevant Common Core and Differences in the Basic Legal Framework of EC and WTO Law A. Substantial Content B. Legal Character C. Summary III. The Basic Legal Requirement of the Free Movement of Goods A. EC Law B. WTO Law IV. The Criteria for Justification of Trade Barriers and Imminent Restrictions of the Prohibition A. EC Law B. WTO Law V. Means of Overcoming Justified Restrictions A. EC Law B. WTO Law References

147 148 148 150 152 152 152 154 161 161 163 165 165 166 166

I. Introduction Protectionism or reasonable national regulation? By the question of protection of non-economic interests as barriers to the free movement of goods in comparison of EC law and WTO law, the long standing and well-known problem is articulated under which legal conditions non-tariff barriers to the free border crossing flow of *

Professor Dr. Dr. h.c. Peter-Christian Müller-Graff, University of Heidelberg.

148

Peter-Christian Müller-Graff

goods can be justified by non-economic national requirements (such as health or environmental protection) within the legal orders of the European Community (EC) and the World Trade Organization (WTO).1 The following observations are split into four parts thereby dividing the overarching problem into subquestions. Firstly, it shall be asked how the basic legal framework for comparing the relevant legal standards for this problem coincide or differ (II.); then the overriding question of whether the approach of WTO rules differs from that of EC law will be scrutinised under three aspects: the basic legal requirement of the free movement of goods (III.), the criteria of justification of trade barriers stemming from the protection of non-economic interests (IV.) and eventually the means for overcoming justified barriers (V.). The substantially inherent question to be looked at is whether the fabric of WTO rules when compared to EC law has a systemic tendency to favour free trade at the expense of non-economic interests. II. Relevant Common Core and Differences in the Basic Legal Framework of EC and WTO Law Concerning the first question – namely the relevant common core and differences in the basic legal framework of EC and WTO law, two basic criteria for comparison should be recalled at the outset: the substantial content (A) and the legal character (B) of the two different bodies of international rules. A. Substantial Content With regard to the substantial content, it is well known that the core of EC law is orientated towards the establishment and functioning of an internal market, which is defined as comprising an area without internal frontiers in which the free movement of goods, persons, services and capital is ensured in accordance with the EC Treaty2 and which is based upon a customs union.3

1 2

3

See, e.g., Müller-Graff (2000), 120 et seqq. See article 14 para. 2 TEC; confirmed by the Treaty establishing a Constitution for Europe agreed upon by the Intergovernmental Conference 2004 (June 18, 2004) in articles I-4 and III-130 para. 2. See article 23 TEC; confirmed by the Treaty establishing a Constitution for Europe agreed upon by the Intergovernmental Conference 2004 (June 18, 2004) in article III-151.

Protectionism or Reasonable National Regulation?

149

Already in this basic respect, the legal framework of the substantially closest piece of the WTO array of agreements, namely the General Agreement on Tariffs and Trade (GATT)4 together with the Technical Barriers to Trade Agreement (TBT)5 and the Agreement on the Application of Sanitary and Phytosanitary Measures (SPS),6 is differently shaped. It is true that the underlying economic idea in the EC Treaty as well as in the GATT is inspired by the same trade theory of comparative cost advantage as is elaborated upon by David Ricardo in his book on ‘Principles of Political Economy Taxation’ in 18297 (despite its idealistic approach and its adaptation in view of the empirical conditions; headwords: neoclassical theory of Heckscher-Ohlin8 and Leontieff-paradoxon,9 new trade theory of economies of scale10). However, the common economic theoretical point of orientation of the EC Treaty and the GATT has not led to identical consequences in binding rules for the exchange of goods between the participating states as a result of the obvious differences of the economic, political and historical settings in the emergence of the two treaties. It is obvious that the GATT does not aim to establish an internal market, a customs union or a free trade area – in the sense of classical customs theory or in the form of the EFTA with the abolition of all tariffs between the participating states. Moreover, the objective of free trade is not even explicitly mentioned in the first part of the preamble of the WTO Agreement, where the raising of living standards is emphasised as well as full employment, high and continuously increasing income, effective demand, extension of production and trade with goods and services, optimal use of resources in conformity with a sustainable development and protection and preservation of the environment.11 As a

4 5 6 7 8 9 10 11

For the text see, e.g. Hummer / Weiss (1997), 553 et seqq. For the text see ibidem, 929 et seqq. For the text see ibidem, 888 et seqq. For the EC see, e.g., Molle (1990), 9; for the WTO see, e.g., Weiß / Herrmann (2003), 7, 10. See, e.g. Weiß/Herrmann (2003), 11. Ibidem, 12. Ibidem, 13. See Scholl / Schorkopf (2002), 31.

150

Peter-Christian Müller-Graff

means to these ends the reduction of tariffs and other trade barriers only is introduced in the third part of the preamble.12 Despite the linguistic structure of the objectives, it has to be maintained that the means, in other words the paths for realizing the general objectives, are the characteristic features of the WTO Agreement and the GATT. They are doubtlessly shaped by the aim and philosophy of free trade. This is also true for the EC Treaty, where the basic article 2 lists the establishment of a common market only as a means, but also as the most characteristic feature for promoting a broad array of substantive objectives (such as a harmonious, balanced and sustainable development in economic activities, a high level of employment and of social protection, equality between men and women, sustainable and non-inflationary growth, a high degree of competitiveness and convergence of economic performance, a high level of protection and improvement of the quality of environment, the raising of the standard of living and quality of life, the economic and social cohesion and solidarity among Member States). However, these similarities in basic economic philosophy did not hinder that the means of fostering the border-crossing trade in goods are differently unfolded. On the one side, the establishment of an internal market with – among many more devices – the free movement of goods, on the other side, no prohibition of tariffs – but only the principle of the most-favoured nation treatment (article I GATT 1994) and the principle of national treatment in internal taxation and regulation (article III), as well as the prohibition of any non-tariff trade restriction (article XI). B. Legal Character The legal characters of the two bodies of international rules once again show very distinct features. It is common wisdom that under EC law, in particular, the provisions of the free movement of goods which are contained within primary law are so called ‘self executing’ norms.13 They are directly applicable and hence have to be applied by national courts, contain subjective rights of individuals and enjoy supremacy over deviating national law.14 At the same time the institutions of the 12 13 14

Ibidem, 32. Leading case: ECJ, case 26/62, Van Gend & Loos [1963] ECR 1. Leading case: ECJ, case 6/62, Costa/ENEL [1964] ECR 1251.

Protectionism or Reasonable National Regulation?

151

Community are attributed legislative powers to establish or promote the free movement of goods.15 In most cases of the approximation of national laws, to these ends single Member States can be outvoted in the qualified majority procedure in the Council.16 For the problem at hand, it is important that justified national restrictions of the free movement of goods can be overcome by the preemption doctrine in favour of EC legislation.17 In contrast to this European framework, the rules of the WTO Agreement in general and the GATT in particular, contain no supranational elements. Their provisions are neither directly applicable nor supreme over contradicting national law.18 In addition, if a non-tariff barrier which roots in non-economic interests is found to be justified under GATT rules, the WTO is neither empowered to harmonize the laws of the Contracting States which pose obstacles to the free flow of goods, nor to implement any non-economic policies by supranational action.19 The absence of those healing powers may well cause higher pressure in the GATT-framework than in EC law on the position that a national rule is justified by non-economic interests. In the Hormones Case the Panel Report treated the recommendations of the Codex Alimentarius Commission, a common suborganization of the FAO and the WHO, concerning the composition, quality and labelling of products as the yardstick for evaluating the necessity of measures for life and health protection in the sense of article 3.1 of the SPS Agreement and the GATT.20 Hence the EC regulation which prohibited the import of beef from hor15 16 17

18

19 20

In particular articles 94, 95 TEC. See article 251 TEC. See, e.g., ECJ, case 5/77, Tedeschi, [1977] ECR 1555; case 72/83, Campus Oil, [1984] ECR 2727; case C-323/93, Crespelle, [1994] ECR I-5077, para. 31; Müller-Graff (2003a), Art. 30 EG para. 13 et seq. See, e.g., ECJ, joined cases 21 to 24/72, International Fruit Company [1972] ECR 1219; case 280/93, Germany v Council, [1994] ECR I4973, 5039; case 469/93, Amministrazione delle Finanze dello Stato v Chiquita Italia SpA [1995] ECR I-4533, 4558; see also MüllerGraff (2000), 127; as analysis Hilf (1998), 21 et seqq. See Müller-Graff (2000), 128 et seq. Panel Report, EC – Measures concerning meat and meat products (Hormones) – complaint by the United States, WT/DS26/R/USA, para. 8.57.

152

Peter-Christian Müller-Graff

mone-treated cattle was strictly scrutinised because it laid down a higher level of health protection than necessary under the Codex Alimentarius recommendations.21 However, in view of the lack of any power of the Codex Alimentarius Commission or the WTO to harmonize health protection standards with binding legal force, this approach is doubtful. Article 3 of the SPS Agreement does not contain an obligation for the participating states to have food quality standards which are identical to those recommended by international commissions or organizations. Hence such international nonbinding standards can only be taken as one source or persuasive authority to exclude higher protection standards, but not as the final say as made clear by article 3.3 SPS Agreement. C. Summary Summarizing the comparison of the basic frameworks it is obvious that EC law on free movement of goods and the WTO framework on trade in goods are legally and politically distinctly different. Concerning the legal nature, elementary differences show up. Nevertheless, in terms of interstate trade theory, both EC law and the WTO framework draw from the same schools of thought. III. The Basic Legal Requirement of the Free Movement of Goods Turning to the set of questions whether the approach of WTO rules differs from that of EC law in dealing with the problem of solving conflicts between the requirements of the free movement of goods and the protection of non-economic interests, the first aspect to be compared is the normative basis and content of the requirement of the free movement of goods. A. EC Law Concerning the basis of prohibitions of trade barriers stemming from the national pursuit of non-economic interests in EC law, it is well known that article 28 TEC prohibits quantitative restrictions on imports and all measures having equivalent effect without distinguishing, in principle, between the purpose of such a measure22 or

21 22

Ibidem, para. 8.75 and 8.76. See, e.g., Müller-Graff (2003a), Art. 28 EG, para. 55.

Protectionism or Reasonable National Regulation?

153

between discriminatory and non-discriminatory measures.23 This is rooted in the famous broad Dassonville-definition – given by the European Court of Justice – of measures having equivalent effect as any measure which is capable of directly or indirectly, actually or potentially, hindering trade between Member States.24 As a consequence, the prohibition covers measures having a protectionist purpose as well as, in principle, those without such a purpose.25 This is not unreasonable, as indistinctly applicable general rules for noneconomic purposes have fairly often been discovered as being disguised protectionist measures.26 On the other hand, it is also obvious that indistinctly applicable rules very often serve to protect non-economic national interests even beyond those embodied in the exception clause of article 30 TEC, where the founders listed: public morality, public policy and public security; the protection of health and life of humans, animals or plants; the protection of national treasures possessing artistic, historic or archaeological value and the protection of industrial and commercial property. Hence within the scope of the prohibition of article 28 TEC, the question arises as to how a line could be drawn between rules which aim at establishing or which generate illegitimate protectionism on the one side and the pursuit of legitimate non-economic interests on the other. The famous, but in its delineating value doubtful Keck-restriction27 of the Dassonville-formula does not aim at this distinction, but has been created for solving a different problem: namely drawing a borderline between cases of interference of national measures with the border-crossing freedom of market access and cases of interference with general economic freedom.28 In contrast the so-called Cassis-jurisprudence29 of the 23 24 25 26 27 28 29

See, e.g., Epiney (2002), Art. 28, para. 14, et seq.; ECJ, case 178/84, Commission v Germany [1978] ECR 1262. ECJ, case 8/74, Dassonville [1974] ECR 837. See, e.g., ECJ, case 178/84, Commission v Germany [1987] ECR 1262. See e.g., Müller-Graff (2003a), Art. 28 EG, para. 211. ECJ, case C-267/91 and C-268/91, Keck and Mithouard [1993] ECR I-6097. See, e.g., Müller-Graff (2003a), Art. 28 EG, para. 237 et seq. ECJ, case 120/78, Rewe-Zentral AG v Bundesmonopolverwaltung für Branntwein [1979] ECR 649.

154

Peter-Christian Müller-Graff

European Court of Justice began to restrict the scope of the prohibition in relation to those indistinctly applicable measures which are found to be necessary to fulfil so called non-economic mandatory requirements such as environmental protection,30 fair trade,31 consumer protection,32 efficient tax control33 and numerous other requirements.34 Since it quickly turned out that, notwithstanding two specific features, the legal conditions for such a restriction of the prohibition parallel the basic pattern of the criteria of justification inherent in the explicit exception clause of article 30 TEC, they will be later discussed in conjunction with them. B. WTO Law Turning to WTO law, the basis for a prohibition of trade-barriers stemming from indistinctly applicable national rules does not seem to be as clear as in EC law. As a result of a less concise and more abstract wording than in the EC Treaty, several provisions can be taken into consideration, in particular articles XI and III of the GATT.35 1. Article XI GATT bars any prohibition or restriction, be it in the form of: quantitative restrictions, import or export licences, or of ‘other measures’. From the outset, the term ‘other measures’ could well be conceived as including any measure capable of hindering trade36 in the sense of the Dassonville-formula. Alternatively, in view of the explicit emphasis on quantitative restrictions and import or export license requirements, the wording of article XI could also be restricted to specific rules on border-crossing trade, thereby excluding all those

30 31 32 33 34 35 36

See, e.g., ECJ, case 302/86, Commission v Denmark [1988] ECR 4630, para. 8; Müller-Graff (2003b), 273. See, e.g., ECJ, case C-30/99, Commission v Ireland [2001] ECR I4660, para. 29. See, e.g., ECJ, case C-239/90, Boscher [1991] ECR I-2039, para. 17; case C-390/99 [2002] ECR I-607, para. 34. See, e.g., ECJ, case 120/78, Rewe-Zentral AG v Bundesmonopolverwaltung für Branntwein [1979] ECR 649. See as an overview Müller-Graff (2003a), Art. 28 EG, para. 225. See for the dispute on the scope of application of these provisions, e.g., Weiß / Herrmann (2003), 185 para. 470 et seqq. In this direction Weiß / Herrmann (2003), 184 para. 465.

Protectionism or Reasonable National Regulation?

2.

37 38

155

measures which pursue non-economic interests.37 If not, an overlap with article III GATT is inherent.38 Article III GATT in several paragraphs requires national treatment of imported goods in internal taxation and regulation. Par. 1 stipulates that internal taxes and other internal charges, laws, regulations and requirements affecting the internal sale, offering for sale, purchase, transportation, distribution or use of products and internal quantitative regulations requiring the mixture, processing or use of products in specified amounts or prohibitions, shall not be applied so as to afford protection of domestic goods. In comparison, par. 4 concentrates on the rules themselves. It states that imported products shall be accorded treatment no less favourable than that accorded to like products of national origin in respect of all laws, regulations and requirements affecting their internal sale etc. It is evident that rules on sale, offering, purchase, transportation, distribution or use can root in economic as well as in non-economic interests. Hence the question arises as to whether par. 4 can be understood as prohibiting national rules, which restrict market access independent from their purpose and also independent from explicit discrimination, or whether par. 4 is restricted to open protectionist measures. In view of the overall objective of the GATT to reduce trade barriers to tariffs, only an interpretation either of Article III:4 GATT or of Article XI GATT would be consequent, which includes any national measure capable to restrict the market access of goods from another country in the sense of the Dassonville-formula. Some elements in the application of Article III:4 GATT point in this direction. In 1989 in the case dealing with US section 337, article III:4 GATT was held to require effective equality of opportunities for imported products and, thus, can be violated if national law – even though neutral on the outset

As examples see, e.g., Senti (2000), para. 534, 549. For a distinction between article III GATT as a prohibition of discriminating measures concerning the product itself and its marketing on the one side and article XI GATT as a prohibition of restrictions concerning process and production methods on the other Weiß / Herrmann (2003), 187 para. 475.

156

Peter-Christian Müller-Graff

– puts illegitimate burdens on imported products.39 In the Korean Beef Case, the question arose as to whether a so-called dual-retail system – in which imported and domestic beef products had to be sold in separate shops – constituted a violation of the national treatment principle.40 Contrary to the view of the Korean government, the formal differentiation on the basis of the origin of beef itself was a distinguishing treatment as the Panel convincingly recognised.41 In EC law this would, in principle, suffice to brand such a rule as a measure having an equivalent effect, since it would be assumed that the result of a split of distribution in an EC Member State would usually force retailers to make a choice between domestic and imported products at the disadvantage of the latter, although in any given case this should be dependent upon the country and the type of product. The Appellate Body gave a more refined reasoning when maintaining that a formal discrimination is neither necessary nor sufficient to constitute a violation of article III:4 GATT and emphasized as the decisive aspect the impact of a rule on competition.42 It can be drawn from this that article III:4 GATT is violated if the national measure modifies the conditions of competitors between domestic and imported products to the detriment of the latter. In its core, this reasoning is not far from the Dassonville-criterion of a trade-hindering effect, because impediments to trade can obviously flow from disadvantageously distorted conditions of competition caused by national rules. In the Korean Beef Case, this distortion was clearly shown by the fact that the split system of distribution led to a considerable reduction in the number of outlets for imported beef.43 Substantial differences to the Dassonville-formula can emerge from a protectionist ‘aims or effect” criterion which

39 40 41 42 43

Panel Report, United States – Section 337 of the Tariff Act of 1930, L/6439-36S/345, para. 5.11. Panel Report, Korea – Measures affecting imports of fresh, chilled and frozen beef, WT/DS161/R, para. 488 et seq. Ibidem, para. 504. Appelate Body Report, Korea – Measures affecting imports of fresh, chilled and frozen beef, WT/DS161/AB/R, para. 137. Ibidem, para. 145.

Protectionism or Reasonable National Regulation?

157

arises from an interpretation of article III:2 and III:4 GATT in the light of article III:1 GATT,44 in short, from the inclusion of the principle that the application of national rules on imported goods should not lead to the protection of similar domestic goods. The specific question of whether the protectionist criterion of article III:1 GATT can be generalised beyond the scope of article III:2 GATT, second sentence, at all, has not to be discussed here. But it has to be emphasized that the criterion of protectionism has its flaws. It has never been taken up as a limiting criterion in article 28 TEC45 and has even been factually abandoned by the European Court in favour of a competition oriented approach in the interpretation of article 90 para. 2 TEC despite46 its explicit inclusion in this provision. The flaws stem from the risk created by the word ‘protectionist” which enables to mix objectives and effects and even the criterion of ‘similar” products and, as a result, excludes any predictability. The ‘protectionist aims or effects” approach led the Panel in the US-Automobile-Tax-Case to the conclusion that the zerotax-rate on domestic light trucks with low fuel economy did not violate article III:2 GATT, although imported automobiles with the same fuel economy were subjected to a high tax rate.47 A protectionist aim was not found and a protectionist effect was denied because of lack of similarity of light trucks and regular automobiles.48 It is doubtful whether under EC law the applicability of article 90 TEC could be denied by the Court of Justice in such a case. First, the objective of a measure is not decisive for defining a trade restriction. Second, the similarity criterion has sharpened in favour of the competition-orientated criterion of whether a product is a substitute to the product

44

45 46 47 48

Panel Report, United States – Measures affecting alcoholic and malt beverages, DS23/R, para. 5.25; Panel Report, United States – Taxes on automobiles, DS31/R, para. 5.7. Due to the broad concept of capability of measure to hinder trade as laid down in the Dassonville-formula. See ECJ, case 168/78, Commission v France [1980] ECR 347, para. 6. See supra note 44, para. 5.35, 5.36 and 5.37. Ibidem, para. 5.35.

158

3.

49

50 51 52 53 54 55

Peter-Christian Müller-Graff

from the viewpoint of the buyer.49 In this aspect, the Court of Justice follows a broad concept of potential competition as shown, for example, in cases concerning different taxation of low-powered cars and high-powered cars in France50 or of beer and wine in Britain.51 In view of the purpose of article III:4 GATT and in the interest of a clear doctrinal approach, it is preferable to abandon the test of protectionist aims or effects on like products as done by the Panel in the case of Japanese taxes on alcoholic beverages.52 On the assumption that either article III:4 GATT or article XI GATT or both seen together can be understood in the sense of the Dassonville-formula,53 the Cassis-question54 could also arise, i.e. the question whether an immanent restriction of the prohibition in the case of the legitimate pursuit of mandatory non-economic requirements is necessary. However, such a necessity is limited since article XX GATT offers a broad array of non-economic reasons for justifying trade barriers. It can be added that the development of the Cassis-restriction was not a necessity either, but an escape from the strict interpretation given to the GATT-inspired article 30 TEC exception-clause by the Court of Justice.55 Be it as it is, the aspect of avoiding the prime labelling of a national measure adapted for non-economic purposes as being incompatible with the law (notwithstanding a justification), seems to be a common concern both on the European level and on the world trade level. It somehow

See, e.g., supra note 46, para. 5: Interpretation of the concept of ‘similar products’ with ‘sufficient flexibility’ including meeting ‘the same needs from the point of view of consumers’ and ‘comparable use’. See ECJ, case 112/84, Humblot v Directeur des Services Fiscaux [1985] ECR 1367. See ECJ, case 170/78, Commission v United Kingdom [1983] ECR 2265. Panel Report, Japan – Taxes on alcoholic beverages, WT/DS8/R, para. 6.16, 6.17 and 6.18. See supra note 24. See supra note 29. See Müller-Graff (2003a), Art. 28 EG, para. 184; Art. 30 EG, para. 28.

Protectionism or Reasonable National Regulation?

159

resembles the uneasiness of doctors with tort law in Germany, which is interpreted by courts in the way that any vaccination by a needle or any surgery constitutes a tort (violation of physical integrity),56 although it is justified when applied for healing purposes with the consent of the patient and in compliance with the state of art in medical practice. After the Panel report in the case of Japanese taxes on alcoholic beverages57 (in which the protectionist test was set aside58), some environmentalists feared that reasonable taxes and regulations for the protection of the environment could be assessed as violating article III GATT, e.g., by stating that a non-refillable container is a like product to a refillable container, or a biodegradable package is a like product to a non-biodegradable package. The question of restricting the prohibition in cases of the legitimate pursuit of non-economic interests seems to be treated sometimes in a disguised form under article III:4 GATT, namely under the heading of the likeness or similarity criterion. In the Asbestos Case59 the question arose as to whether a complete ban on the manufacture and the sale of asbestos-based fibres (for the purpose of protecting the health of workers and consumers) violated article III:4 GATT insofar as non-asbestos based fibres remained admitted on the French market.60 Under article 28 TEC, such a ban could loose its character as a trade restriction if it fulfilled the mandatory requirement of public health protection (as long as this is part of the Cassis-jurisprudence, which is doubtful in view of the system of article 30 TEC61). In the Asbestos Case this question took the form whether asbestos fibres and non-asbestos fibres were like 56 57 58 59

60

61

§ 823 para. 1 BGB. Panel Report, Japan – Taxes on alcoholic beverages, WT/DS8/R. Ibidem, para. 6.18. Panel Report, European Communities – Measures affecting asbestos and asbestos containing products, WT/DS135/R; Appellate Body Report, European Communities – Measures affecting asbestos and asbestos containing products, WT/DS135/AB/R. Panel Report, European Communities – Measures affecting asbestos and asbestos containing products, WT/DS135/R, para. 8.155, 8.157 and 8.158. See Müller-Graff (2003a), Art. 28 EG, para. 208; ECJ, cases C-1/90 and C-179/90, Aragonesa [1991] ECR I-4151, 4183, para. 11.

160

Peter-Christian Müller-Graff

products in the sense of article III:4 GATT.62 In previous cases a broad spread of factors had been developed to decide the likeness question: physical properties, nature of the product, quality, end-use, tariff classification, consumers’ perception and conduct (tastes and habits)63 – hence rather stiff and static criteria. In the Asbestos Case the central question asked was whether the potential health risk of asbestos was a factor to be taken into consideration. While the Panel thought that this aspect would be best considered within the exception of article XX(b) GATT,64 the Appellate Body took a different view highlighting the fact that the risk of asbestos being cancerogenic, was not only a characteristic of the product, but might also influence the choice of the consumer and hence exclude the likeness of the products compared.65 However, this is very ambiguous reasoning, since it implies the existence of potential competition between the two types of fibres and its restriction by the ban of asbestos-based fibres. Taking into account the non-economic interest of health protection for the presumed perception of the consumer, the likeness of the two types of products and hence the violation of article III:4 GATT was denied. While this result is acceptable, it is difficult to consider the reasoning a convincing legal approach to grasp the problem.

62

63

64

65

Panel Report, European Communities – Measures affecting asbestos and asbestos containing products, WT/DS135/R, para. 8.117 to 8.150. Report of the Working Party, Border Tax Adjustments, BISD/18S/97, para. 18; Panel Report, European Communities – Measures affecting asbestos and asbestos containing products, WT/DS135/R, para. 8.112; Appellate Body Report, European Communities – Measures affecting asbestos and asbestos containing products, WT/DS135/AB/R, para. 133. Panel Report, European Communities – Measures affecting asbestos and asbestos containing products, WT/DS135/R, para. 8.129, 8.130 and 8.132. Appellate Body Report, European Communities – Measures affecting asbestos and asbestos containing products, WT/DS135/AB/R, para. 115 and 122.

Protectionism or Reasonable National Regulation?

161

IV. The Criteria for Justification of Trade Barriers and Imminent Restrictions of the Prohibition Concerning the criteria for justification of trade barriers, the EC law (A) and WTO law (B) have developed on their respective normative tracks. A. EC Law The criteria for justification have their basis in article 30 TEC. In addition, they have been expanded to the (imminent) restriction of the basic prohibition in article 28 TEC by the jurisprudence of mandatory requirements.66 1. In line with the doctrine of market law,67 the restriction of the prohibition and the justification of a hindrance as interpreted by the Court of Justice have five conditions in common: first, both can be invoked only for the pursuit of non-economic interests,68 second, in both cases, the concrete measure which poses an impediment to free trade has to be capable of serving the relevant non-economic interest,69 third, the measure must be necessary to achieve the pursued goal, which means that no less restrictive means must be available for realizing the purpose,70 fourth, the achievement for the non-economic interest must be proportional in relation to the trade restriction and hence constitute a gain for the overall common good in the Community,71 and eventually, fifth, according to article 30 second sentence TEC, such prohibitions and restrictions shall not constitute a means for arbitrary discrimination or disguised

66 67 68

69

70 71

See supra note 29. See Müller-Graff (1984), 280. See, e.g., ECJ, case C-120/95 Decker [1998] ECR I-1831, 1884, para. 39 and ECJ, case 7/61 Commission v Italy [1961] ECR 659, 698. See, e.g., ECJ, case C-317/92 Commission v Germany [1994] ECR I2039, 2061, para. 17 and ECJ, case C-39/90 Denkavit [1991] I-3069, 3108, para. 23. See, e.g., ECJ, case 261/81 Rau [1982] ECR 3961 and ECJ, case 104/75 de Peijper [1976] ECR 613. See, e.g., ECJ, case C-189/95, Franzén [1997] ECR I-5909, 5976, para. 76 and ECJ, case 155/73 Sacchi [1974] ECR 409, 428.

162

2.

72 73 74 75

76

Peter-Christian Müller-Graff

restriction on trade between Member States.72 Whether the fifth condition is at all relevant when assessing a concrete restriction in a systematic approach remains an open question. Usually the necessity-test or the proportionality-test will already unveil arbitrary discrimination or disguised restriction. There is a plentitude of detailed jurisprudence on all these questions73 which need not be outlined in this context. Only the most significant overarching characteristic in dealing with the conflict of free movement of goods and the protection of non-economic interests has to be emphasised. EC law has developed a fine-tuned and dense systematic and doctrinal approach to solving such conflicts, thereby permitting rather reliable predictions for the possible outcome in terms of legal rules. This is slightly less true for the two criteria which distinguish the (imminent) restriction of the prohibition from the requirements of the justification in the exception clause. Firstly, while the justification clause of article 30 TEC can pertain to indisdictly applicable as well as to discriminatory provisions and restrictions, the restriction of article 28 TEC was conceived by the Court of Justice to only apply to measures which are indistinctly74 applicable to domestic and imported goods such as, e.g., purity rules in foodstuff and beverages. However, beginning with the Wallonian-Dump-decision,75 the clear judicial borderline to open discriminatory treatment has been blurred, though for good reasons, since the strict interpretation given by the Court of Justice to the explicit interests of article 30 TEC excluded such objectives as consumer protection or fair trade.76 Hence it seems that the Court See, e.g., ECJ, case 27/80 Fietje [1980] ECR 3839, 3855 and art. 30 TEC. See, e.g., Müller-Graff (2003a), Art. 28 EG, para. 186 et seq.; Art. 30 EG, para. 22 et seq., para. 47 et seq. See, e.g., ECJ, case 113/80 Commission v Ireland [1981] ECR 1625, 1639, para. 10. See ECJ, case C-2/90 Commission v Belgium [1992] I-4431, 4480, para. 34, 36; for recent developments see Müller-Graff (2003a), Art. 28 EG, para. 197. See, e.g., ECJ, case 177/83 Kohl v Ringelhan [1984] 3651, 3663, para. 19.

Protectionism or Reasonable National Regulation?

163

was forced to invent a new legal tool to give way to the compatibility of proportional measures based on such purposes with internal market law. Thus it is only consequent that the Court has often contemplated the possibility of restricting the prohibition of article 28 TEC, also for differently applicable measures in the interest of mandatory requirements (e.g. of environmental protection77 or the social security system78). The reverse side of this development is presently a lack of predictability whether the Court will accept a discriminatory measure on the grounds of mandatory requirements. The second distinction between restriction and justification concerns the scope of protectable non-economic interests. While article 30 TEC contains an exhaustive list which enjoys, as mentioned, a strict interpretation by the Court,79 the array of mandatory requirements as a creation of the Court is neither limited nor clearly predictable. However, the manifest tendency to expand or detail the spectrum of protectable non-economic interets is unavoidable in the light of the broad scope of the Dassonville-formula. B. WTO Law Turning to WTO law, it has already been illustrated that a distinction between an immanent restriction of a prohibition and a justification is not developed in a way comparable to EC law. However, the justification of non-tariff trade barriers on non-economic grounds is part of the founding philosophy of the GATT as laid down in article XX GATT. Compared to the jurisprudence of the European Court on article 30 TEC, it can easily be recognised that the scheme of a similar pattern in the interpretation and the application of article XX GATT has developed, although perhaps not always elaborated upon with the same doctrinal rigidity and conciseness as in EC law.

77 78 79

See, e.g., ECJ, case C-389/96 Aher Waggon [1998] ECR I-4473, 4488, para. 18, 4490, para. 26. See ECJ, case C-120/95 Decker [1998] ECR I-1831, 1884, para. 39. Rule of strict interpretation; see, e.g., ECJ, case 7/61 Commission v Italy [1961] ECR 695, 720.

164

Peter-Christian Müller-Graff

As a clear common basis for justifying a trade barrier, the noneconomic interest has to be invoked.80 Second, it is also common ground that the measure can be justified only if it can be considered to promote the respective noneconomic interest (e.g. Appellate Body in the Shrimp-Turtle Case: import of shrimps and shrimp production in relation to the protection of sea turtles81). The criterion is also touched by the requirement of a reasonable relation between measures to the legitimate end.82 Third, the necessity test is well established under article XX GATT (as well as under articles 2 and 5 SPS). Here, however, it seems that at least in one respect, more is required by WTO Law than under EC law: namely the requirement of exhausting all options reasonably available through measures consistent with the GATT before taking national legislation, especially serious efforts of negotiations of an international agreement83. In contrast, this is not required from Member States within article 30 TEC. The GATT requirement can be understood as a reaction of the necessity-criterion to the lack of any competence of the WTO to overcome justified restriction by way of supranationally harmonizing national rules. A second specific feature (in comparison to EC law), is a certain vagueness in the approach to necessity. It is a somewhat sweeping and not a very operationable statement when necessity is considered to relate to a range of degrees of necessity (Appellate 80

81 82 83

E.g. the protection of sea turtles in the Shrimp-Turtle Case (WT/DS58/R), the protection of marine mammals, namely dolphins, in the Tuna-Dolphin Case (DS21/R); the protection of human health in the Cigarettes Case (DS10/R-37S200). Ibidem, WT/DS58/R para. 137 et seq. Ibidem, para. 141. See e.g. Panel Report, United States – Restrictions on imports of Tuna, WT/DS21/R, para. 5.28; Panel Report, United States – Import prohibition of certain shrimp and shrimp products, WT/DS58/R, para. 7.55; Appellate Body Report, United States – Import prohibition of certain shrimp and shrimp products, WT/DS58/AB/R, para. 166; Appellate Body Report, United States – Import prohibition of certain shrimp and shrimp products, Recourse to Article 21.5 of the DSU by Malaysia, WT/DS58/AB/RW, para. 123 and 134, where the Appellate Body finds, that negotiations are required but not the conclusion of an agreement.

Protectionism or Reasonable National Regulation?

165

Body in the Korean Beef case84), thereby hinting that the more vital and important the interests are at stake the easier it is to accept an enforcement measure as being necessary85. It is already difficult to see how – beyond the life and health of humans – the vitality of interests can be graded on an abstract scale. In any case, it may raise the barrier for justification by denying the necessity of interests considered to be less vital. The necessity-criterion in the interpretation of article XX GATT also comprises the requirement of proportionality between the extent to which the measure contributes to the realization of the end pursued on one side and the extent to which the measure produces impeding effects on international trade on the other side. It is an open question whether, in this respect, the interpretation of article XX GATT requires more promotion of the non-economic interest than under article 30 TEC. Last a clear parallel exists between article XX GATT and article 30 TEC in so far as both exceptions are subject to the so-called Chapeau-Clause, in other words to the condition that they do not amount to arbitrary discrimination or a disguised restriction on international trade. V. Means of Overcoming Justified Restrictions A. EC Law A specific characteristic of the approach of EC law in dealing with the conflict of free trade and the protection of non-economic interests concerns the means for overcoming justified restrictions of trade. As mentioned already, EC law empowers Community institutions to legislate for the purpose to establish free trade in goods (in particular articles 94 and 95 TEC). Although orientated towards the functioning of the common or internal market, these competencies imply the harmonization of national standards for the protection of non-economic interests. As a consequence the Court of Justice has developed the so called ‘preemption rule’ which means that

84 85

See Appellate Body Report, Korea – Measures affecting imports of fresh, chilled and frozen beef, WT/DS161/AB/R, para. 161. Ibidem.

166

Peter-Christian Müller-Graff

Member States, in principle,86 are barred from unilaterally taking measures (with restrictive effects on trade) by invoking the protection of non-economic interests by their own standards after the Community has issued a binding conclusive standard to the relevant question.87 B. WTO Law The lack of a supranational competence of the WTO to overcome justified non-tariff trade barriers by way of harmonizing national standards demonstrates the limits of the basic GATT-concept. The objective of abolishing non-tariff trade barriers faces serious obstacles as long as it is not accompanied by the respect for and the harmonization of the justified protection of non-economic mandatory requirements which cause impediments to trade. This is the genuine European experience with the establishment of the internal market. On the WTO level, TRIPS is a good example that the extension of free trade depends upon the implementation of common standards for the protection of mandatory requirements (in the case of TRIPS: intellectual property rights) which could be jeopardised by free trade. Hence border-crossing liberalization and deregulation depends upon the degree of harmonised regulation. The European experience in this respect can not be ignored on the WTO level if free trade is to be further expanded in the future at the global scale.

References Astrid Epiney (2002), Art. 28, in: Christian Calliess / Matthias Ruffert (eds.), EUV/EGV-Kommentar, Neuwied (Luchterhand) 2 2002. Meinhard Hilf (1998), Internationales Wirtschaftsrecht und nationale Gerichte, Bonn (Zentrum für internationales Wirtschaftsrecht) 1998.

86

87

Notwithstanding the exception of article 95 para. 4 and 5 TEC or exceptions in directives which aim at approximating the laws of the Member States. See, e.g., ECJ, case 251/78 Denkavit [1979] ECR 3369, 3388.

Protectionism or Reasonable National Regulation?

167

Waldemar Hummer / Friedl Weiss (1997), Vom GATT ’47 zur WTO ’94. Dokumente zur alten und zur neuen Welthandelsordnung, Baden Baden (Nomos) 1997. Willem Molle (1990), The Economics of European Integration (theory, practice, policy), Aldershot (Dartmouth) 1990. Peter-Christian Müller-Graff (1984), Unternehmensinvestitionen und Investitionssteuerung im Marktrecht. Zu Maßstäben und Schranken für die überbetriebliche Steuerung von Produktionsinvestitionen aus dem Recht des wettbewerbsverfaßten Marktes (= Tübinger Rechtswissenschaftliche Abhandlungen, Vol. 59), Tübingen (J.C.B. Mohr) 1984. Peter-Christian Müller-Graff (2000), Die Maßstäbe des Übereinkommens über technische Handelshemmnisse (ÜTH) als Bauelemente eines Weltmarktrechts, in: Peter-Christian MüllerGraff (Hrsg.), Die Europäische Gemeinschaft in der Welthandelsorganisation, Baden-Baden (Nomos) 2000, 111-130. Peter-Christian Müller-Graff (2003a), Art. 30 EG, in: Hans von der Groeben (ed.), Vertrag über die Europäische Union und Vertrag zur Gründung der Europäischen Gemeinschaft, BadenBaden (Nomos) 62003. Peter-Christian Müller-Graff (2003b), Umweltschutz und Grundfreiheiten, in: Werner Rengeling (ed.), Handbuch zum europäischen und deutschen Umweltrecht, Vol. 1, Köln (Heymanns) 2 2003, 239-293. Peter-Tobias Scholl / Frank Schorkopf (2002), WTO – Welthandelsorganisation und Welthandelsrecht, Köln (Heymanns) 2002. Richard Senti (2000), WTO, Zürich (Schulthess) 2000. Wolfgang Weiß / Christoph Herrmann (2003), Welthandelsrecht, München (Beck) 2003.

Piet Eeckhout*

Trade and Non-Economic Policies in the EU and in the WTO. A Comment on Peter-Christian Müller-Graff’s Paper ‘Protectionism or Reasonable National Regulation?’ In this comment I should like to formulate a few reflections on the broad and multi-dimensional subject of the relationship between free-trade rules and non-economic policies, as it is developing in the EU and in the WTO. I concentrate on recent case-law, because the judiciary is the main vehicle which moves the debate forward. It goes without saying that this comment does not permit a full analysis of all the issues.1 In EC internal market law the main interface between free movement and non-economic policies of Member States is the jurisprudential concept of ‘mandatory requirements’,2 or ‘overriding general-interest reasons’.3 Its pedigree need not be explored here. It is an open-ended concept enabling the courts to scrutinize, with a lot of deference, the reasons for which Member States obstruct or impede free movement; deference, because mandatory requirements include virtually any decent government policy, provided it is ‘non-economic’ in character, i.e. it does not aim to protect a national market. More significantly, however, the ‘mandatory requirements’ concept enables the courts to review the actual effect and application of such non-economic policies, by means of the lawyer’s classical balancing act: the principle of proportionality. That principle has become the Court of Justice’s magic wand in difficult internal-market cases. Not only does proportionality have *

1 2

3

Professor of European Law and Director, Centre of European Law, King’s College London Visiting Professor, College of Europe, Bruges See Ortino (2004). As identified in Cassis de Dijon (Case 120/78 Rewe v Bundesmonopolverwaltung für Branntwein [1979] ECR 649, para 8. See e.g. Case 8/02 Ludwig Leichte v Bundesanstalt für Arbeit, judgment of 18 March 2004, para 43.

170

Piet Eeckhout

many guises; the preliminary rulings procedure, which is the main conduit for internal-market cases, enables the Court of Justice to decide how much of the balancing act it wishes to undertake, and how much it leaves to the referring court. One could say that national courts are treated as apprentices who do not know in advance (i.e. when referring a case to the Court of Justice) precisely how much of the actual feat they will have to perform. That is a peculiar result of the preliminary-ruling mechanism, which in some cases arguably defeats its whole purpose. Gourmet International is a case in point.4 The case concerned politically sensitive restrictions on alcohol advertising in Sweden, part of Sweden’s broader policy of combating alcohol consumption. The Court of Justice decided some important issues, such as whether such restrictions constituted discrimination in fact in the meaning of Keck and Mithouard.5 However, the crucial and decisive question was whether the particular restrictions in issue could be justified and were proportionate in light of Sweden’s policy. The Court did not give guidance on this issue, and left this for the referring court to decide. The above may suffice by way of background on EC free movement law, as the general principles are well-known.6 The caselaw is not however stagnant, and in recent years a couple of new trends appear to be emerging which are worth highlighting, before turning to WTO law. First, one can notice a shift in the scope of the ‘mandatory requirements’ concept. Under traditional doctrine the concept does not include ‘economic’ policies. The Court has not reformulated this doctrine, but some of the recent case-law reveals that there can be an economic dimension to a mandatory requirement. There are for example the cases in the sphere of trade and public health, concerning freedom to undergo treatment abroad, and reimbursement of medical expenses incurred in another Member State, be it for health products or services. In such cases the Court now recognises that the financial balance of the social-security or health system may constitute a mandatory requirement. It expressly accepts that it is a legitimate aim to ensure that there is sufficient and permanent 4

5

6

Case C-405/98 Konsumentsombudsmannen v Gourmet International [2001] ECR I-1795. Joined Cases C-267/91 and C-268/91 Keck and Mithouard [1993] ECR I-6097. For further analyses see Barnard / Scott (2002).

Trade and Non-Economic Policies in the EU and the WTO

171

access to a balanced range of high-quality hospital treatment in a Member State,7 and that this aim may override restrictions on the freedom to receive hospital services in another Member State. These are clearly government aims with a much stronger economic dimension than more traditional mandatory requirements such as consumer and environmental protection. In fact some of the publichealth cases seem to put the picture on its head: they pit patients, primarily intent on improving their health rather than engaging in economic transactions, against governments concerned to maintain the financial and economic basis for their hospital systems. Another example perhaps of this shift in the economic and the non-economic, in the context of mandatory requirements, is the Court’s approach towards environmental protection policies. The particular case I am thinking of is PreussenElektra.8 I will return to this case below, because it primarily concerns the second evolution I should like to address. PreussenElektra concerned a German government instruction to electricity distributors to purchase part of their electricity from ‘green’ suppliers based in Germany. The Court did not express any concern about the fact that the policy appeared to be discriminatory towards green electricity suppliers in other Member States. The case forms part of a broader development, but it may be noted that here the Court may have been inclined not to interfere with German efforts to build up a viable green electricity industry within national borders. There is clearly an economic dimension to this approach towards environmental protection policies at Member State level, and perhaps rightly so. By discussing this case I have already noted the second trend I should like to address, i.e. the relaxation, at least in cases involving environmental protection, of the traditional principle that mandatory requirements cannot be relied upon as justification for discriminatory measures. Unfortunately the Court has not been particularly clear in this regard, declining invitations to state whether it is, or is not, abandoning that traditional principle.9 But there are now enough judgments where the Court accepts what are arguably discriminatory measures to be able to speak of a trend.10 One can 7

8 9 10

Case C-157/99 Geraets-Smits and Peerbooms [2001] ECR I-5473, para 78. Case C-379/98 PreussenElektra v Schleswag [2001] ECR I-2099. See the Opinion of Jacobs AG in PreussenElektra at para 229. Ibidem, para 227.

172

Piet Eeckhout

only surmise for what reasons the Court may be reluctant expressly to confirm what is clearly developing into a general approach. From a perspective of comparison with WTO law, however, it may indeed be interesting to attempt some educated guesswork. One of the main differences, in terms of judicial process, between the WTO dispute settlement organs and the EU judiciary is methods of interpretation.11 The Appellate Body follows the Vienna Convention on the Law of Treaties, by putting great faith in the ordinary meaning of the words, and relying less on context, and even less on object and purpose. The Court of Justice, by contrast, often prefers a teleological interpretation – tongue in cheek one could say that object and purpose are at the very core of EC law theology. Yet in the case of the Court’s reluctance formally to abandon the traditional principle we are discussing here, there may be an element of inhibition to stray too far from the text of the Treaty. After all, the only justifications which the Treaty expressly permits are those in Article 30 EC (and the corresponding exceptions for services, establishment, workers, and capital). The grand bargain of Cassis de Dijon12 was to extend the scope of Article 28 EC to ‘indistinctly applicable’ measures, but in conditional form. Where such measures could be justified on grounds of mandatory requirements they were not regarded as constituting prohibited measures having equivalent effect. This conditional extension enabled the Court to create further exceptions, without distorting the Treaty text: mandatory requirements were not in the same category as exceptions under Article 30 EC. Crucially, the bargain depends on maintaining the distinction between ‘measures having equivalent effect’ as such, and ‘indistinctly applicable measures’, which are only a subcategory. The very cornerstone of that distinction is that there may be measures having equivalent effect, namely those which are distinctly applicable, which cannot be justified on grounds of mandatory requirements. If the Court removes that cornerstone, the whole doctrinal edifice of the Court not having rewritten Article 30 EC – artificial though it no doubt is – will collapse. I should like to make a particular comment on this episode. It is open to argument whether the Court should maintain its Cassis de Dijon edifice, or abandon it by admitting that mandatory requirements can be relied upon as justification for distinctly applicable 11 12

Ehlermann (2002). Above note 2.

Trade and Non-Economic Policies in the EU and the WTO

173

measures. The Court, however, appears unwilling to decide the point either way. The fudged approach of decisions such as PreussenElektra is highly unsatisfactory, as the doctrinal hesitation prevents the Court from engaging with the real issues. If it is accepted that, on environmental policy grounds, the German authorities may take measures to promote the production of domestic green electricity, what then are the contours and boundaries of this permissive approach? Up to what point and subject to which conditions is discrimination against foreign suppliers of green electricity permitted? What kind of distinctly applicable measures does the environmental-policy exception permit? What is the role of the proportionality principle? etc. My argument is that the doctrinal hesitation itself may well be the very reason for the Court’s incapacity to engage with the real issues. I make this comment because it seems to me that we see a similar phenomenon in the WTO, where the debate about the aims-and-effects doctrine has equal capacity to detract attention from the real issues; but I turn to this below. The third evolution I should like to note in the case-law on free movement is the increasing focus on human-rights questions. The recent decisions in Carpenter and Schmidberger exemplify this evolution. In Carpenter the Court of Justice decided that a UK national providing services in other Member States could rely on the relevant Treaty provisions as against the decision to deport his wife, who was from the Philippines, and was illegally resident in the UK. The Court relied on the fact that Mrs Carpenter contributed to Mr Carpenter’s business by working at home and taking care of the children. It expressly referred to the right to family life, as protected under the European Convention on Human Rights (which forms part of the general principles of EC law).13 In Schmidberger the Court had to examine whether the Austrian authorization of a blockade by protesters of the Brenner Autobahn, a blockade constituting a trade restriction, could be justified on grounds of freedom of expression and the right of assembly. The Court decided that such basic rights could justify trade restrictions, and that in the circumstances of the case the restrictions were not disproportionate, and therefore had to be tolerated.14 Whilst I do not wish to analyse 13

14

Case C-60/00 Carpenter v Secretary of State for the Home Department [2002] ECR I-6279, para 41. Case C-112/00 Schmidberger v Austria [2003] ECR I-5659.

174

Piet Eeckhout

this development any further here, it is worth noting, in particular because a case in the WTO on the relationship between free trade and fundamental rights is waiting to happen. The general picture of EC law is one where the judiciary retains considerable freedom in balancing free-trade principles with non-economic policies. Mandatory requirements and proportionality are flexible concepts, and the Court of Justice deals with a great number of cases, all factors which enable the Court to be a very valuable actor for finding the right balance. In addition, there is an active EC legislature, a factor which allows for a dialogue between judges and policy-makers.15 Things are rather different in the WTO. For a start, the active legislature is largely missing, placing too much of the burden of developing the law on the dispute settlement system. That issue cannot be further explored here.16 But also as regards the role played by the WTO judiciary, there are significant differences with EC law. There are fewer cases in the WTO, which results in slower legal development and perhaps also in excessive focus on the particular facts and circumstances of individual cases. In addition, there are the differences in methods of interpretation. As the Appellate Body is so careful to adhere to the customary international law rules of treaty interpretation, where the ordinary meaning of treaty terms is central, there appears to be much less room for manoeuvre in the act of balancing free trade with noneconomic policies. So what is the WTO record hitherto? Within the scope of this comment I can offer but a few sketchy reflections. The first is that, notwithstanding the textual constraints under which the Appellate Body operates, it has proved astute at finding ways to strike a defensible balance. First the tax side. In cases such as Japan – Alcohol and Chile – Alcohol the Appellate Body has emphasized that the national-treatment imperative of Article III:2, second sentence, GATT, concerning directly competitive or substitutable products, involves an examination of whether the tax regime affords protection. That examination includes an analysis of the objectives of the tax regime, which of course may form part of a non-economic policy.17 Differentation in the taxation of products on, say, environ15 16 17

Eeckhout (1998). von Bogdandy (2001). Appellate Body, Japan – Taxes on Alcoholic Beverages, WT/DS8/ AB/R, WT/DS10/AB/R and WT/DS11/AB/R, adopted 1 Nov 1996;

Trade and Non-Economic Policies in the EU and the WTO

175

mental grounds, is therefore not prima facie illegal, at least not as regards directly competitive or substitutable products. Second, on the regulatory side, there is EC-Asbestos, where the Appellate Body managed to introduce public-health considerations into the likeproducts analysis under Article III:4 GATT.18 Certain categories of domestic and imported products may not be like, basically on grounds of the stark differences in their respective effects on human health. If they are not like products, there can be no lack of national treatment. Third, in the political minefield of balancing trade and environment, the Appellate Body in US – Shrimp/Turtle has managed to give precedence to environmental considerations, inter alia by interpreting the notion of exhaustible natural resources in Article XX GATT in accordance with general developments in international (environmental) law. At the same time it has construed the chapeau of Article XX in such a way that it strictly regulates and conditions the extraterritorial application of domestic environmental standards.19 The above does not of course mean that there are no problems in WTO law. My second reflection concerns the doctrinal debate about the aims-and-effects theory.20 For the uninitiated, this theory was developed in the context of Article III:2, GATT 1947, by a couple of panels which emphasized that the examination of whether domestic and imported products were ‘like’ involved looking not only at the effects of the tax regime in issue, but also at its aims. If the aims of differentation between various products were unrelated to protectionism, then such differentation did not constitute lack of national treatment. As Hudec has pointed out, the analogy with ‘mandatory requirements’ was clear.21 The reference to the aims of (tax) legislation could develop into a judge-made exception to the national-treatment principle, built into the very examination of

18

19

20 21

Chile – Taxes on Alcoholic Beverages, WT/DS87/AB/R and WT/DS110/AB/R, adopted 12 January 2000. Appellate Body, EC – Measures Affecting Asbestos and AsbestosContaining Products, WT/DS/135/AB/R, adopted 5 April 2001. Appellate Body, US – Import Prohibition of Certain Shrimp and Shrimp Products, WT/DS/58/AB/R, adopted 6 November 1998 and idem, Recourse to Article 21.5 by Malaysia, WT/DS/58/AB/RW, adopted 21 November 2001. Hudec (1998). Ibidem.

176

Piet Eeckhout

whether a measure constituted lack of national treatment. However, the aims-and-effects theory did not as such survive the transition from GATT to the WTO. In Japan – Alcohol the Appellate Body considered that the terms of Article III:2, first sentence, GATT (on like products) did not lend support to the theory. As regards the second sentence, the Appellate Body emphasized that the intentions of legislatures were irrelevant.22 In EC – Bananas the Appellate Body rejected the EC’s attempts to justify the banana legislation on grounds of its non-protectionist aims, both as regards Article III:4 GATT and Article XVII GATS.23 The broad aims-and-effects theory has thus been rejected, essentially on the ground that it cannot be brought within the terms of Article III GATT. However, where the terms of Article III so permit, the Appellate Body has been willing to pay regard to regulatory aims and objectives. In Chile – Alcohol it emphasized in its examination under the heading of ‘so as to afford protection’ (Article III:2, second sentence) that ‘a measure’s purposes, objectively manifested in the design, architecture and structure of the measure, are intensely pertinent to the task of evaluating whether or not that measure is applied so as to afford protection to domestic production’.24 In EC – Asbestos the Appellate Body was willing to give considerable weight to the health effects of the various products in order to conclude that they were not ‘like’. 25 One could certainly see that as a concealed reference to regulatory objective: the difference in health effects was of course the very reason why the French legislature had sought to ban asbestos and not other products. There are two comments I should like to make on these jurisprudential developments. The first is that the accommodation of legislative objectives in Chile – Alcohol and EC – Asbestos clearly has its limits, which would appear to preclude the development of a comprehensive doctrine comparable to ‘mandatory requirements’. As regards taxation there is no scope for consideration of legislative objective under Article III:2, first sentence, i.e. in case of like products. It is true that the notion of like products in this provision needs 22 23

24 25

Above note 17. Appellate Body, EC – Regime for the Importation, Sale and Distribution of Bananas, WT/DS27/AB/R, adopted 10 October 1997. Above note 17, para 71. Above note 18.

Trade and Non-Economic Policies in the EU and the WTO

177

to be narrowly construed,26 but even in case of like products there may be tax differentation which is justified on non-economic grounds, whereas GATT does not appear to permit such differentiation if the more highly taxed product is imported. As regards regulation, the Appellate Body emphasized in EC – Asbestos that the health effects of asbestos were so relevant, under the heading of physical characteristics, because of their stark nature. It is not at all certain that differences in physical characteristics which are less pronounced (e.g. cars with different levels of fuel consumption) could significantly influence the like-products analysis. At the conference which led to this publication it was argued that EC – Asbestos signalled the introduction of something akin to ‘mandatory requirements’ in WTO law. I am not convinced by that analysis. The test for ‘like products’ remains essentially factual, and oriented towards establishing whether products are in competition. As it stands it is not a proper basis for systematic reference to and inclusion of non-economic legislative objectives. This brings me to my second comment. In my view the doctrinal debate about aims-and-effects has been wrongly focused on the notion of ‘like products’, and has not touched the real issues. The national-treatment rule is of course no more than a particular manifestation and expression of the non-discrimination principle. In all anti-discrimination law there are issues of determining what is comparable (the comparator), and the notion therefore in GATT is ‘like products’. Whilst I admit that locating the correct comparator is not always straightforward, there is another important dimension to anti-discrimination law, i.e. defining the type of different treatment which constitutes discrimination. Reflection on this issue in WTO law has progressed little beyond the slogan that not only de iure but also de facto discrimination is prohibited. Yet the contours of this imprecise notion of ‘de facto discrimination’ remain to be determined. As I have argued elsewhere, the discrimination notion itself offers much more scope for consideration of non-economic legislative objectives than the notion of ‘like products’.27 The Appellate Body hinted at that in the famous paragraph 100 of EC – Asbestos, where it emphasized that: A complaining Member must still establish that the measure accords to the group of ‘like’ imported products ‘less favourable 26 27

Japan – Alcohol, above note 17. Eeckhout (2001).

178

Piet Eeckhout

treatment’ than it accords to the group of ‘like’ domestic products. The term ‘less favourable treatment’ expresses the general principle, in Article III:1, that internal regulations ‘should not be applied … so as to afford protection to domestic production’. If there is ‘less favourable treatment’ of the group of ‘like’ imported products, there is, conversely, ‘protection’ of the group of ‘like’ domestic products. However, a Member may draw distinctions between products which have been found to be ‘like’, without, for this reason alone, according to the group of ‘like’ imported products ‘less favourable treatment’ than that accorded to the group of ‘like’ domestic products. What this means, in my view, is that regulatory distinctions between like products are not automatically in breach of Article III:4 GATT on the mere basis that an imported product is treated less favourably than a domestic product. The Appellate Body emphasizes that the comparison is between the group of like imported and the group of like domestic products. This opens up a lot of space for debate about when regulatory distinctions are protectionist.28 The analysis could, and arguably should, include consideration of legislative objectives. Further cases may bring more clarity in this respect.

References Catherine Barnard / Joanne Scott (eds.) (2002), The Law of the Single Market. Unpacking the Premises, Oxford (Hart) 2002. Piet Eeckhout (1998), The European Court of Justice and the Legislature, in: 18 Yearbook of European Law, Oxford (Oxford University Press) 1998, 1-28. Piet Eeckhout (2001), Constitutional Concepts for Trade in Services, in: Gráinne de Búrca / Joanne Scott (eds.), The EU and the WTO – Legal and Constitutional Issues, Oxford (Hart) 2001, 211-236. Claus-Dieter Ehlermann (2002), Six years on the Bench of the ‘World Trade Court’. Some Personal Experiences as Member

28

Ehring (2001).

Trade and Non-Economic Policies in the EU and the WTO

179

of the Appellate Body of the World Trade Organization, in: Journal of World Trade 36 (2002), No. 4, 605-639. Lothar Ehring (2001), De Facto Discrimination in WTO Law: National and Most-Favored-Nation Treatment – or Equal Treatment?, Jean Monnet Working Paper 12/01. Robert E. Hudec (1998), GATT / WTO Constraints on National Regulation: Requiem for an ‘Aims and Effects’ Test, in: International Lawyer 32 (1998), 619-649. Federico Ortino (2004), Basic Legal Instruments for the Liberalisation of Trade, Oxford (Hart) 2004. Armin von Bogdandy (2001), Law and politics in the WTO – Strategies to Cope with a Deficient Relationship, in: Jochen A. Frowein / Rüdiger Wolfrum (eds.), Max Planck Yearbook of United Nations Law Volume 5, Den Haag (Kluwer Law) 2001, 605-674.

Thomas Cottier and Satoko Takenoshita*

Decision-making and the Balance of Powers in WTO Negotiations: Towards Supplementary Weighted Voting I. II.

Introduction Consensus and Majority Voting in the WTO A. Rules and Practices on Decision-Making B. The Future of Consensus III. Weighted Voting for WTO Decision-Making A. Qualifying Equality of States B. The Basic Requirements for Legitimate DecisionMaking under Weighted Voting 1. Weighted Voting under IMF and the Quota Reform 2. Weighted Voting for WTO C. The Variables 1. The Size of Trade in Goods and Services – WTO Budget Contributions 2. Gross Domestic Product (GDP) 3. Market Openness 4. Size of Population 5. Basic Votes D. The Proposed Formula 1. Formula 2. Basic Requirements and the Formula E. Calculated Voting Weight F. Assessment *

182 187 187 190 193 193 196 197 198 198 199 199 199 200 200 201 201 203 204 208

Thomas Cottier: Professor of Law, Managing Director, World Trade Institute and University of Berne, Switzerland; Satoko Takenoshita: Economist, World Trade Institute, Switzerland and University of Bologna, Italy, respectively. The paper partly draws from Cottier / Takenoshita (2003).

182

Thomas Cottier / Satoko Takenoshita

IV. Assessing of Voting Power A. SSI, BI, Inclusiveness and Decision Probability Indices B. Quantified Power Indices 1. SSI and BI 2. Inclusiveness 3. Decision Probability 4. Assessment of the Power Analysis V. Conclusion References Appendix A: Evolution of the IMF Quota Formulas Appendix B: Source of Variables

209 210 211 212 217 218 219 220 221 224 227

I. Introduction The failure of the September 2003 Ministerial Meeting of Cancún (Mexico) to reach agreement on a negotiating agenda and on refined parameters of the Doha Development Agenda adopted in November 2001 begs the question whether consensus based diplomacy in the WTO has reached its limits. Coalitions based upon variable geometry shape the process and render it unpredictable and difficult. The power to block consensus has been used to extract concessions and finally let the conference fail. Many would argue that this is not new. Trade Rounds always have been very difficult and moved from crisis to catharsis and final break-troughs based upon extraordinary efforts. The failure of the Brussels Ministerial Conference in the mid of the Uruguay Round in 1988 as well as the failure of the Seattle Conference in 2000 reflects these inherent difficulties and characteristics of multilateral trade negotiations. Failure, it will be argued, is mainly due to the substance of the negotiations, not the process. In Seattle, agreement failed on key issues such as trade and environment or labour. In Cancún, the paramount reasons may be ascribed to the fact that industrialized States were not prepared to make sufficient commitments in the process of liberalizing market access and reducing domestic support in the field of agriculture. It will be a matter of closing gaps, and then negotiations can be successful on the basis of traditional modes and procedures. Also, it may be argued that the WTO has taken on more than it can chew, and that the Organization and its Members need to find back to the traditional functionalist core business of facilitating

Towards Supplementary Weighted Voting

183

market access. However, the problems faced during the recent ministerial conferences suggest that difficulties are not merely due to unbridgeable differences in substance in market access issues, and this for a number of reasons. Firstly, there is the complexity and interrelation of regulatory issues, in particular in areas relating to trade barriers of the second and third generation: Non-tariff barriers and domestic regulation of conditions on competition. Any attempt to exclude regulatory issues from trade negotiations will rapidly show that they are inherently linked to the goals of the WTO and cannot be separated from efforts to reduce trade barriers and improve market access. It should be recalled that the importation of standards of intellectual property protection with the TRIPs Agreement was essentially and rightly so motivated by an effort to overcome de facto barriers to trade and investment due to lack of IPR protection. Equally, addressing domestic levels of support in agriculture is closely linked to market access. Similarly, liberalization of services inherently touches upon domestic regulatory issues. This argument could be further refined in order to show that even the fourth generational issues of antitrust and investment (Singapore issues) refuted at Cancún, are closely related to the core business of the Organization and its regulatory framework, and indeed, already form part of WTO law in different agreements and specific sectors. International trade regulation has come a long way, and emerged as one of the most complex regulatory areas of contemporary international law. Secondly, the structure of the law and its effectiveness has changed. The binding effect of WTO dispute settlement decisions no longer allow for unilateral exits. They are no longer based upon consensus of all Members concerned, and failure to implement decisions adopted by the Dispute Settlement Body is subject to burdensome and painful retaliatory market access restrictions. While commitments under GATT 1947 were equally binding, the price of non-compliance has increased. Inevitably, this obliges governments to be cautious in entering new obligations and to assure sufficient domestic support in the negotiating process. Thirdly, the process of negotiations has become even more complex with the advent of new Members and increasing observation of the talks by pro-active non-governmental organizations outside the traditional circles of commerce. While, depending on the subject matter, a core group of some 20-30 Members were leading the process in the Uruguay Round, participation of developing

184

Thomas Cottier / Satoko Takenoshita

countries has strongly increased. It is a very positive sign that more Members are actively involved in the negotiations than in previous rounds. Yet, it is obvious that consensus is more difficult to build with soon 150 Member States. The traditions of informal consensus-building negotiations in working groups and in the Green Room are increasingly challenged by Members, other international organizations and non-governmental organizations. Negotiations based upon like-minded groups are not new. Yet, it seems that constellations are more complex today and Members continue to pursue their diverging interests even within these groups. A system of representation and thus the possibility to work again in smaller groups has not emerged. As the subject-matter of trade inherently touches upon the realm of other international organizations, enhanced efforts at policy co-ordination are required not only domestically, but also between different international organizations. Fourthly, the system of multilateral trade rounds developed and designed for MFN tariff negotiations extends to all regulatory issues, despite their complexity. The WTO is not in a position to engage in a process of on-going negotiations even within rounds, except for well-defined left-over areas. The post-Uruguay Round negotiations on telecommunication and financial services have remained the only successful examples of inter-round work and achievements, while other efforts, e.g. in the field of rules of origins, failed. The concept of working towards comprehensive package deals is a powerful one and offers many advantages. Yet, it does not allow for incremental progress in between rounds of negotiations. Given the fact that rounds usually extend to a full decade, the system entails important back-logs and has to leave progress essentially to judicial dispute settlement. In a previous paper, we extensively addressed the relationship between negotiations and judicial dispute settlement, emphasizing the problem of legislative response.1 The absence of inter-round negotiations and thus of the opportunity to effectively address and possibly change interpretations or even rules adjudicated in dispute settlement have created a profound imbalance between the political and the judicial branches of the WTO process. While panels and the Appellate Body inherently take this into account by way of mainly textual and contextual interpretations of the rules in accordance with Article 3.2 of the DSU and Article 31 of the Vienna Conven1

Cottier / Takenoshita (2003).

Towards Supplementary Weighted Voting

185

tion on the Law of Treaties, WTO jurisprudence almost exclusively assumes a heavy burden and extensive responsibility in applying and interpreting WTO principles and rules. The lack of legislative response not only creates an imbalance, but also prevents the law from being incrementally developed beyond the narrow bounds of the case law. In response to this relationship of diplomacy and the judicial process, some have suggested rebalancing by way of rolling-back the more powerful functions of the judicial branch on the basis of concerns of legitimacy and democratic accountability. The return to a former role of panels, including the Appellate Body, to former GATT 1947 standards are based upon the assumption that the political process based upon consensus and package-deals should or cannot change in terms of realpolitik.2 Criticism is informed by the view that rulings of panels and the Appellate Body have become increasingly intrusive and put a risk at democracy and legitimacy at home. However, this view and public perceptions tend to ignore the fact that the tensions caused mainly rely upon substantive WTO rules and disciplines that are negotiated rights and obligations resulting from the political, rather than the judicial, process. Problems encountered in the causes célèbres (bananas, hormones, FSC) were not caused by dispute settlement per se but by the very norms and rules that applied to these cases. Nevertheless, apart from fundamental critique addressing the democratic legitimacy of WTO as a whole,3 only a few have so far addressed the need to reform the diplomatic process by which these rules were brought about.4 Attention has been mainly drawn to the reform of dispute settlement as a main remedy to overcome impasses in settling major, mainly transatlantic, disputes.5 We therefore suggested that main efforts should primarily be made to re-establish the balance by improving the diplomatic process. To this end, we discussed and developed models and a system of weighted voting which could replace the traditions of consensus in order to allow for effective negotiations in between rounds and for legislative response.

2 3 4 5

Barfield (2001). E.g. Krajewski (2001). Ehlermann (2002a); (2002b). Davey (2003); Petersmann (2002).

186

Thomas Cottier / Satoko Takenoshita

In this paper, we will further extend the analysis for introduction of weighted voting to WTO. We argue that the idea need not be limited to inter-round negotiations and legislative response, but may serve as a subsidiary or supplementary option for WTO negotiations in general where consensus diplomacy fails. We essentially rely on the same weighted voting model presented previously. However, a number of new elements were introduced. Firstly, we consider the European Union (hereinafter EU) with its future 28 Members,6 while the Union was limited to its 15 Members analyzed and considered in the previous paper. We present a hypothetical case in which EU28 is treated as a single unit. We do so by limiting data to foreign trade of the EU with the third countries, excluding distortions created by taking into account internal trade within the EU market. We will compare its impact on voting weight and power with the model of treating 28 EU countries individually. Secondly, the study of different quota required will be more detailed. The previous work merely analyzed voting analysis for simple and a hypothetical 70% majority rule. Within the WTO framework, the existing voting rules other than consensus are simple, 2/3rd and 3/4th majority rules. In this paper, these rules are examined by using our calculated voting weight together with a double majority rule, which imposes the second condition of simple majority rule in terms of the number of Members to each majority rule. Yet, before turning to these models, we attempt to show that a shift from consensus supplemented by the principle of One State – One Vote to weighted voting would not fundamentally alter the multilateral trading system, but prepare it instead to cope with the challenges of the 21st Century.

6

The current EU25 Members are: Austria, Belgium, Czech Republic, Cyprus, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Slovakia, Slovenia, Spain, Sweden, United Kingdom. The 3 candidate countries are: Bulgaria, Romania and Turkey.

Towards Supplementary Weighted Voting

187

II. Consensus and Majority Voting in the WTO A. Rules and Practices on Decision-Making The WTO, as the GATT 1947 before, is based upon the formal and fundamental principle of One State – One Vote, reflecting sovereign equality of all Members States, small, medium and large alike. Article XXV:3 GATT 1947 enshrined the principle that ‘Each contracting party shall be entitled to have one vote at all meetings of the Contracting Parties’. Decisions were deemed to be taken, in accordance with para. 4 of this provision, by majority of the votes cast, unless otherwise provided for. A two-third majority of votes cast was required to approve waivers in accordance with para. 5. Special 2/3rd requirements existed also under Articles XXIV:10 for approving special exceptions and deviations and XXXIII GATT for admission of new Members. Amendments to the GATT partly implied unanimity as changes to Art. I and Art. XXIV required acceptance by all contracting parties to become effective. For other provisions, it would seem that a simple majority was sufficient. The amendment, however, would only enter into force upon adoption by 2/3rd of the membership. Subsequent state practice overruled the majority rule and worked under the principle of consensus diplomacy on all levels of operation. Unlike unanimity, consensus does not require affirmative approval, but merely the absence of objections. Unless objections are voiced, the chairperson, without taking a pole, may conclude that consensus exists. Other than for waivers and accessions, all decisions, including dispute settlement, were taken by consensus in the Council and in committees ever since 1959.7 The Marrakech Agreement Establishing the World Trade Organization builds upon these practices and provisions and codified a combination of consensus and a subsidiary scheme of voting procedures. The Agreement stipulates in Art IX: 1: ‘The WTO shall continue the practice of decision-making by consensus followed under GATT 1947’. Consensus is defined as follows in footnote 1: ‘The body concerned shall be deemed to have decided by consensus on a matter submitted for its consideration, if no Member, present at the meeting when the decision is taken, formally objects to the proposed decision’. Consensus is mandatory for amending the list of 7

WTO (1995), 1098 et seq., 1108 et seq.

188

Thomas Cottier / Satoko Takenoshita

plurilateral agreements in accordance with para. 9. Otherwise, voting shall take place where consensus fails. Each Member of the WTO has one vote. The Agreement specifies this for decisions of the Ministerial Conference and the General Council only. Article IX sets forth the principle of simple majority of votes cast, and the requirement for 2/3rd for authentic interpretation in para. 2, for 3/4th for the adoption of waivers in para. 3. Proposals by Members to amend the Agreements requires 2/3rd for submission of the proposal for acceptance. In addition, a further quorum needs to be met for the implementation and entry into force of amendments to the Agreements. Those listed in Annex 1 relating to the decisionmaking (Art XI) and principles of non-discrimination in GATT, GATS and TRIPs only take effect upon acceptance by all members, according to para. 2. Acceptance by 2/3rd of the Membership is required for altering rights and obligations of Annex 1 A Agreements and the TRIPs Agreement and a 3/4th majority may decide that Members not accepting the amendment may be free to withdraw from the WTO, in para. 3. Similar provisions apply to GATS in para. 5. Except for the accession of Ecuador, voting has not taken place under WTO since 1995, nor in appointing the director-general of the Organization, nor in minor matters of house-keeping. Consensus is adhered to at all costs. Moreover, the principle also extends to the adoption and implementation of agreements. The treaties were accepted on the basis of a single undertaking (except for the plurilateral agreements) in the Uruguay Round and rules of individually accepting amendments enshrined in Article X WTO Agreement have not been operational so far. The reasons for strict adherence to consensus are obvious. The principle of One State – On Vote reflects an extensive discrepancy of real and voting powers. The voting rules represent neither the stake individual Members have in the multilateral system nor their real impact on power in shaping the rules of the multilateral system (both of which are recognized by consensus practices). The imbalance and material inequality of representation in terms of voting rights is significant when we look at the shares of financial contributions to the WTO, gross domestic product (GDP), and voting rates as shown in Figure 1. The group of industrial countries, composed of 24 members, occupying 79% of Member’s total GDP and supplying 71% of the WTO budget, represent only 17% of vote at the WTO. The group of developing countries, comprising 119

Towards Supplementary Weighted Voting

189

members, on the other hand, which supplies 21% of total GDP and contributes 29% to WTO budget, represents 83% of total vote.8 The ratio of vote representation to budget contribution is 1:4.2 for the group of industrial and 1:0.37 for the group of developing countries, respectively. GDP Developing Countries

WTO Budget Vote GDP

Industrialized WTO Budget Countries Vote 0%

20%

40%

60%

80%

Figure 1: Share of GDP, Contribution and Vote in the WTO. Source: WDI except for Cyprus, Liechtenstein, Myanmar, and Taiwan (see Appendix B); Note: Armenia, Nepal and Macedonia are not included as they are new WTO members after 2003 and contribution is not allocated for 2003. The group of industrialized countries does not include new / candidate 13 countries for the EU but these countries are placed in the group of developing countries. Contribution is the average of 2001-2003; GDP is taken from the average of 1996-2000.

These ratios depict the problem of substantial material inequality which is present under the present WTO framework, and which explains why voting procedures under Art. IX WTO Agreement are not used as a supplementary means and decision-making remains strictly reserved to consensus. Current voting rules in WTO, in other words, fail to respond to the requirement that majority voting procedures need to be able to assure that major trading 8

The number of official WTO membership as of April 2004 is 146 (with EU 147). However, Armenia, Nepal and Macedonia are not included in this figure, as they entered into the membership after 2003 and the contribution for 2003 is not allocated. At the time of writing this paper, Cambodia was accepted but Parliament refuted approval. Russia’s accession is likely to be next. For calculations, the paper is based on 147 Members (including Cambodia, but not Russia).

190

Thomas Cottier / Satoko Takenoshita

partners in the system keep a profound interest in dealing reliably with each other on the basis of the WTO law. The WTO cannot be compared to the United Nations General Assembly and its One State – One Vote principle or other international organizations operating under this principle. The General Assembly does not exercise law-making functions, and matters of international security are dealt with by the Security Council which is subject to customary veto powers of permanent members. Other international organizations operating on One State – One Vote do not exercise comparable law making functions and do not adopt package deals which were essentially mandatory for all members upon completion of the Uruguay Round. In the WTO, the principle of One State – One Vote raises fundamental problems of inequality to which we return shortly. At this stage, we merely emphasize that loss of substantial influence or control through One State – One Vote majority voting runs the risk that key stakeholders informally leave the system and turn to unilateral, bilateral or regional trade regulation which they are able to effectively control. This is not in the interest of smaller nations who strongly depend on most-favoured-nation (MFN) and non-discriminatory market access rules. For such reasons, consensus has been strongly favoured by all Members alike and occasional efforts to resolve a problem by voting have been abandoned. B. The Future of Consensus Observing the past decades of GATT, the consensus principle overall has been successfully applied. While it has often led to protraction and delay, eight rounds of multilateral negotiations were completed with far reaching effects and impact on real life and the structure of the economies of Members. Consensus successfully operates in many instances, as interests of Members are taken into account ex ante in the negotiating process. All Members present (not those absent) have the right to oppose, and from this point of view it is a method that inherently reflects formal equality of States. At the same time, consensus informally reflects differences in power and influence of Members in the organization. Real powers to shape and block consensus is not evenly spread. Some clearly are more equal than others, based upon their stake and interest in the multilateral trading system and the influence they yield on the basis of market size and geopolitical roles. In practical terms, consensus diplomacy results in largely divergent powers to effectively use the

Towards Supplementary Weighted Voting

191

veto power. While large trading entities are in a position to block consensus more frequently in the daily operations of the organizations and on ministerial levels, smaller nations and interest groups need to dose the usage of their blocking rights very carefully. In practical terms, the power for them is limited to issues of existential importance; yet even there it may be impossible to exercise veto powers in light of disadvantages to be expected from such conduct. The trend to form interest groups and coalitions, such as the group of 21 or the group of 10, may be seen to enhance effective influence and veto powers with a view to discharge individual members from solely assuming responsibilities and the political and economic costs of blocking consensus. Consensus as a model of decision-making which preserves the vital interests of both large partners and equally protects smaller nations alike is likely to continue to be successful when dealing with multifarious issues negotiated with a view to achieve an overall and balanced package deal. Enlarged membership may further protract results, but does not exclude them, in particular in the field of market access negotiations. There is a predominant view among Members to the effect that negotiations should continue to based upon consensus diplomacy. For reasons already indicated, however, there are increasing doubts since the failure of the Ministerial Meeting of Cancún whether consensus diplomacy alone and without the backing of an effective voting system will be able to cope with the complexities of future rounds which reach well beyond the realm of market access negotiations. It is increasingly difficult to cope with all the regulatory issues and tasks within the framework of a package deal within a reasonable period of time. While package deals negotiated under the pressures of consensus diplomacy may be able to hammer out basic agreement, the potentials of blocking consensus is increasing, given the divergence of economic and political interests at stake. Moreover, the quality of rule-making may suffer if package deals need to deal with all the details of regulations. Clearly, it would be beneficial if more work could be effectively left to interround negotiations. Here, for reasons discussed in our first paper, consensus has largely failed. Little progress has been achieved and the main burden has been placed on dispute settlement, leading to a system out of balance. Authentic interpretation and legislative response are virtually impossible to achieve. A winning party is not

192

Thomas Cottier / Satoko Takenoshita

likely to concede changes on an issue that may put the successful implementation of a ruling adopted by the DSB at risk. The evidence affirms this view. Up until now, not a single authentic interpretation, let alone an amendment of an agreement, has materialized following dispute settlement. Formal reactions have been limited to the granting of waivers under Article IX:3 WTO Agreement. Unlike legislative response, this however amounts to dispensation from existing agreements and is limited in effect to particular Members. It is at this stage that we recognize one of the main roots of the problem: panels and the Appellate Body are confronted with the interpretation and application of rights and obligations. The scope of law-making and shaping by panels and the Appellate Body is inextricably intertwined with the modes and processes of negotiations applied. Whilst treaty law in many areas calls for support and completion by way of adjudication and case law, the practical impossibility of bringing about legislative response obliges the Appellate Body to exercise strong restraint and refrain from more extensive interpretations, in particular of obligations incurred by the agreements. The emphasis of textual interpretation is an emanation of this constellation. Panels may have somewhat more leeway, as their views can still be tested before the Appellate Body. The same is not true for the review process. It is quasi final until the law changes in a new round of multilateral negotiations. Such restraint, at the same time, bears the risk of limiting forwardlooking, purposing interpretations and clarifications that may be used to guide Members, for example in determining the relationship to agreements outside the WTO or to other international organizations. Ideally, the fragmented and often incomplete law of the WTO would rather call for a larger scope of interpretation in order to achieve full coherence with the system. Yet, the lack of possibilities of legislative response and thus a true dialogue between the judicial and the political branches of the WTO renders this politically difficult. For these reasons we submit that consensus diplomacy should be supplemented by a workable and realistic system of voting to which the Members of the WTO could revert in cases of stalemate, melt-downs and break-downs of negotiations, both within rounds and in between rounds. We do not argue to replace consensus and its advantages by voting at all times. We suggest to understand voting as a supplementary instrument of decision-making. Its primary field of application will be work undertaken in between

Towards Supplementary Weighted Voting

193

rounds and in regular business of the WTO. In trade rounds, it will assist overcoming major difficulties and influence conduct during negotiations. The proposal thus stays within the foundations enshrined in Article IX WTO Agreement and its requirement of qualified majorities, except for the One State – One Vote principle. Importantly, decisions taken within the WTO on the basis of weighted voting and the adoption of an amendment remains subject to Art. X:3 WTO Agreement. Each Member is free to decide whether an agreement approved on the basis of weighted voting shall be adhered to. Any voting system effectively available as a supplementary means of decision-making, of course, exerts ex ante influences on the operation of consensus practices. Consensus diplomacy backed with weighted voting will not be based upon exactly the same power constellations as today. The possibility to revert to voting in case of failure will shape the modes of consensus diplomacy. Even if not practiced and used, the allocation of voting rights and of voting powers is of crucial importance – humming like an operational system in the background. Deviations from the current consensusalone rule will therefore depend upon the modes of allocation of rights and voting powers. Before turning to this crucial point, we address legitimacy of weighted voting. III. Weighted Voting for WTO Decision-Making A. Qualifying Equality of States Formal equality of States is one of the fundamental principles of international law. All States, large and small alike, enjoy the same rights and obligations under the United Nations Charter and in customary international law as a horizontal order between sovereign nations. The principle was strongly reinforced through the process of decolonization in the 20th century. The principle of One State – One Vote is expression of sovereign equality; so is decisionmaking by consensus which equally allows all States alike to voice objections and block a consensus if they wish to do so. Both, consensus and One State – One Vote therefore are perceived as an important guarantee of equality and even democracy in the international system. Formal equality of States has a strong appeal and legitimacy, and it is necessary to state the reasons why unequal treatment nevertheless should be preferred. Some of the operational reasons were already stated above. The failure of consensus and a

194

Thomas Cottier / Satoko Takenoshita

complete lack of supplementary means of decision-making due to formal equality is neither in the interest of Members, large and small alike, nor from the point of view of effective multilateralism. It is important to note at this stage that equality before the law is not limited to a formal concept. Substantially, equality requires to treat alike similar or comparable constellations of facts, and treat diverging settings differently. The essence of equality amounts to operate these distinctions in order to achieve justice and fairness. Factual differences therefore matter a great deal in law. This is equally true in allocating powers in decision-making. States and markets are vastly different in size and power, and taking such realities into account renders international law and thus the rule of law more effective. Formal equality does not exclude to take into account factual differences in allocating voting powers to individual states. There are many examples to this effect in law, and within international law as well. Voting powers of individual shareholders depend on stakes and vary in company law while all are treated alike in terms of procedural rights. Seats allocated in parliament depend on the size of population, and voting powers of citizens in federal states operating dual majority requirements substantially vary. The Member States of the European Union operate under a system of weighted voting in the Council and their people are presented in the European Parliament in accordance with the size of population. Voting rights in the Bretton Woods system, discussed shortly, are based upon weighted voting on the basis of stakes and shares. Finally, some States are vested with a permanent seat in the Security Council and with customary veto power while other Members are not entitled to such rights. Finally, it was seen that consensus rule in substance and reality rather reflects different weight and influence of Members within formal equality of States. Differences in allocating voting powers therefore explicitly and implicitly exist under the principle of formal equality and sovereignty. Equality, however, requires that differences are based upon relevant and rational factual distinctions. They need to be justified in light of the overall goal and purpose of a particular structure. They need to support its efficiency and contribute to its legitimacy. By taking into account differences in reality – such as the size of markets, size of contributions made, population, economic performance, dependence on international trade, important linkages to reality are established which allow to operate a system in real terms and to provide, at the same time, a careful balance of

Towards Supplementary Weighted Voting

195

power, avoiding abuses of rights. Shaping the criteria for distinction therefore amounts to the main challenge. It is here, less so than on the principle of unequal treatment, that main discussions should take place and efforts being made. These efforts should also be made for the WTO, for reasons discussed above. We submit, as in our first paper, that there is a mutual interest in creating models of weighted voting by which decisions can be taken on the basis of simple of qualified majority of the votes cast, allowing to overcome impending difficulties to find consensus. In constructing majority requirements, it is essential to achieve a careful balance: Firstly, no Member alone should individually be in a position to block the adoption of a decision – similar to the current state of law in dispute settlement in relation to DSB decisions based upon an Appellate Body report. Secondly, no decision should be adopted against the combined will of major stakeholders. We define major stakeholders as those countries whose relations should essentially and genuinely be based upon WTO law, and not be mediated by way of bilateral or regional preferential agreements. Direct relations of these countries therefore form the backbone of the multilateral system, or the golden triangle. This triangle has essentially entailed the relationship of the United States, the European Communities, and Japan. It is in the process of being enlarged to other major countries depending their development on international trade, in particular the People’s Republic of China. As long as their mutual relations essentially and genuinely rely upon WTO law, the increasing number of regional systems and agreements of the key stakeholders with their neighbours or allies will work as a complement, rather than a threat to, the WTO. Failure to preserve and develop the multilateral backbone, in return, will render the proliferation of unilateralism, bilateralism and regionalism in a manner detrimental to the WTO. It will inevitably lead to the decline of the WTO as the basic constitutional structure of international trade. It is therefore of paramount importance that major stakeholders keep a viable interest in WTO, and incentives to informally depart from it are countervailed. Thirdly it is important to do justice to medium and smaller Member States alike. Weighted voting for the purposes of lawmaking and legislative response to dispute settlement must assist in avoiding such effects and provide a start for strengthening longer term global integration and governance within the WTO. A system

196

Thomas Cottier / Satoko Takenoshita

would be able to assure that smaller and medium sized Members cannot be systematically outvoted. Models of weighted voting complying with these basic requirements could usefully support negotiations and rational decisionmaking within the organization. Supplementary voting will contribute to transparency and thus legitimacy of results. It would also facilitate the work governments who often are exposed to conflicting domestic interests. We submit that voting following difficulties to find consensus is beneficial to discharge responsibilities of governments vis-à-vis their constituencies at home. Today, a government forced to block consensus is confronted with challenges both at home and abroad. At home, interests seeking progress in negotiations will be frustrated. Abroad, the exercise of blocking rights amounts to be blamed for derailing negotiations and will be met with increased pressures to comply in subsequent efforts. A vote will allow governments who fail to achieve their goals, to concede defeat in a rational manner. Voters will accept in the long run when a qualified majority of Members has brought about results different from what their own government has fought for. The burden of blocking consensus is an overly burdensome one and is likely to destabilize governments more than the process of accepting a subsequent vote on the matter. In proposing an introduction of weighted voting model for the WTO, the next section first overviews the basic requirements for legitimacy of weighted voting model, variables to be used in calculating the voting weight, and the proposed formula, all of which are initially discussed in detail.9 Then the quantification of the proposed formula follows. B. The Basic Requirements for Legitimate Decision-Making under Weighted Voting Weighted voting in international law is not new. While it clearly entails an important step in rationalizing international relations, it does not leave the realm of international law and does not turn the WTO into a constitutional framework which exceeds the foundations of equal sovereignty and consent. The Bretton Woods system of the IMF and the World Bank offer an important precedent in point. For many, however, it is bad example. Weighted voting in the WTO faces the problem that many object to it because of that 9

Cottier / Takenoshita (2003).

Towards Supplementary Weighted Voting

197

precedent and how weighted voting was shaped in these institutions. It is therefore important to assess that experience and learn the lessons from it. Different models of weighted voting can be designed, and a future system in the WTO needs to follow the patterns and allocations chosen in the Bretton Woods institutions. Nevertheless, it is important to recall their foundations and the experience made in the World Bank and the International Monetary Funds. In particular, their efforts at reform under way provide important insights. 1. Weighted Voting under IMF and the Quota Reform The IMF is a prime example of multilateral organization operating under weighted voting,10 although many decisions are taken by consensus. Its voting weight is allocated to Member states in a manner in which ‘each member receives 250 (basic) votes, plus 1 additional vote for each part of its quota equivalent to one 100,000 Special Drawing Rights’ (IMF Articles of Agreement Art. XII Section 5a). The basic votes are introduced so as to ‘pay some homage to the traditional principle of the equality of states, to avoid too close an analogy to the private business corporation, and to guard against too great a concentration of voting power in the hands of one or two members’.11 The majority rule adopted depends on different issues: for example, it requires 85% negative majority for authentic interpretation of the provisions of the Agreement whereas double majority rule of three-fifths of the total membership and four-fifths of the total voting weight of members is adopted as regard to an amendment of the Articles of the Fund. The legitimacy of decision-making is currently questioned as it is often criticized that it reflects too much interest of economically strong Members. Reflecting the criticism and the change in economic relations over time, the IMF quota allocation and weighted voting are currently under a process of reform. In fact, the current voting weight of industrial countries constitutes more than 60%, although it represents only 16% of Membership. In addition, the EU 10

11

The World Bank also adopts weighted voting system, which is similar to that of IMF. The voting weight is derived from specified number of membership votes (as in IMF basic votes) and additional votes based on the number of shares of the stock held. The voting weight distribution among Membership is also similar to IMF. Gold (1974), 688.

198

Thomas Cottier / Satoko Takenoshita

and the US are provided with de facto veto power when 85% majority is adopted, as their voting weights in IMF decision-making represent almost 30% and 17% respectively.12 The situation partly stems from the fact that the number of basic votes has remained constant while the level of quota increased by 37 folds due to inflation and economic growth in general since 1944. As a consequence, the share of basic votes to total votes declined from 11.3 to 2.1 percent despite the quadrupled number of Membership.13 The initial intention to preserve equality of States has thus become ineffective and the balance of power shifted toward countries with large quota. Ultimately, the quota and voting weight reform is required to reestablish a more appropriate balance. Overall, it should bring about enhanced legitimacy and is supposed to achieve three major functions: simplicity, transparency, and robustness to changes in the international economy and to the relative change in the position of a country.14 2. Weighted Voting for WTO The goals and functions aspired for IMF quota reform should be taken into account and reflected in designing a weighted voting system in the context of the WTO in order to maintain the legitimacy as an international organization. Additional criteria, however, need to be taken into account in order to meet the basic criticism which the Bretton Woods voting system faces. Besides trade specific factors, two other elements will be added in our proposal in order to ensure the legitimacy of decision-making in the WTO. Any system of fair weighted voting needs to assure that particular countries do not have independent veto power and that the representation between two groups of countries, industrial and developing is adequately assured. All these requirements will seek to ensure enhanced legitimacy of an international organization, whose primary task is to process a wide range of interests and obligations. C. The Variables As the WTO is an international organization that primarily deals with trade and trade-related rules among Member countries, it is natural to assume that variables for voting weight allocation should 12 13 14

See Cottier / Takenoshita (2003) for details. Buira (2003), 2. IMF (2001).

Towards Supplementary Weighted Voting

199

reflect each member’s performance and power in terms of trading activities undertaken by its economy. As in our first paper, we propose four possible variables to be considered: the size of trade in goods and services, the size of GDP, the openness of the market, and population. Most suggested variables are economic in nature, reflecting the fact that the WTO is an organization dealing with rules related to economic activity although its activity tends to expand to trade-contiguous issues.15 1. The Size of Trade in Goods and Services – WTO Budget Contributions The size of trade in goods and services are a fundamental element in assessing a Member’s performance and power in the context of the WTO. Indeed, this factor is used for the allocation of WTO budget contribution of individual Member States. As the contribution is allocated in proportion to each Member’s three-year average share of international trade in goods, services and intellectual property rights, we suggest the use of WTO contribution as a variable in order to allow double assessment of performance and power – trade size and budgetary contribution to the organization. In particular, as the allocation of WTO contribution provides minimum threshold of 0.015% for Members whose trade proportion to the total is less than 0.015%, those who contribute 0.015% should be treated as such also in assessing the voting power, rather than using the raw data on trade size.16 2. Gross Domestic Product (GDP) GDP is one of the most comprehensive indicators for measuring a Member’s economic size relative to other members and the availability of data is relatively well established. The three-year average GDP is used in order to smooth the effects of short-term cyclical fluctuation yet to reflect update condition of a Member state.

15

16

There may be other options such as commitment level to, e.g. WTO rule in general, Intellectual Property Rights or Environmental standards, however, the credibility of such assessment is not yet established. On the other hand the economic indicators used here are more widely accepted. There are 35 Members subject to 0.015% of WTO contribution for 2003.

200

Thomas Cottier / Satoko Takenoshita

3. Market Openness Market openness is defined as the proportion of imports to GDP and is used as an assessment of a Member’s contribution to world trade by offering open markets, regardless of the size of the GDP. We consider this to be an important factor as it also establishes dependence and reliance of an economy on the multilateral trading system. Open economies thus have a larger stake in the system, and this should be reflected in voting rights. Further, as small countries are in general more open to trade than large countries, the insertion of this variable is effective in offsetting a bias against smaller countries whose size of trade and GDP tend to be small. 4. Size of Population The population variable can be an option to be included as a noneconomic social factor. This enables to assess the size of population affected by a decision, independent from different level of social and economic development. 5. Basic Votes The combination of these above variables for assessing the voting weight will prove effective in balancing the representation between the group of industrial and developing countries as the variables of trade and GDP are dominated by industrial and emerging countries whereas the variables of openness and population are generally privileging developing countries in general (See Table 1). However, the benefit of the latter two variables within this group is disproportionate. Hence a number of Members from developing countries may be left with very small numbers of votes. In order to ensure a more equal distribution of voting weight, the insertion of basic votes can be an effective option, following the principle adopted by IMF. Thus the insertion of basic votes will also be tested. Table 1: Regional Distribution of Variables1 Contribution Contribution GDP ø Openness WTO 20032 WTO ‘98ø 98-2000 Revised 2000

Population 2001

37 Industrial Countries

72.8

62.5

81.2

0.273

19.1

USA (1)

15.9

22.2

31.2

0.14

5.3

4

10.2

0.09

2.4

EU (28)

42.9

20.6

30.0

Japan (1)

6.4

8.8

14.9

0.12

Towards Supplementary Weighted Voting

201

Contribution Contribution GDP ø Openness WTO 20032 WTO ‘98ø 98-2000 Revised 2000 110 Devel. Countries

Population 2001

27.2

37.5

18.8

0.47

80.9

1.8

2.4

1.2

0.42

11.0

Asia (23)

17.7

24.4

9.9

0.55

58.2

Europe (7)

0.2

0.3

0.1

0.51

0.5

Middle East (8)

1.9

7.9

6.5

0.48

1.6

Americas (32)

5.7

7.9

6.5

0.48

9.6

Africa (40)

Source: Calculations by the authors from various sources; WTO, WDI, National statistics, UN Statistics Division, DOTS, IFS, Eurostat (see Appendix B and C for details); Notes: 1 Numbers in parentheses indicate the number of Members for each region. Country composition of regional area adopts classification of IFS except for Liechtenstein, Taiwan, and Cuba, which are put into category of Industrial, Asia, and Americas, respectively. 2 Excluding Members joined WTO after 2003 (Armenia, Macedonia and Nepal) and Members subject to ratification for the entry (Cambodia) 3 Excluding Liechtenstein (inclusion of the country yields 0.52 for average openness for industrial countries as its openness is exceptionally high at 2.75); 4 Excluding intra-EU28 trade. The openness for EU28 is 0.33 when intra-EU trade is included. Concerning intra-EU trade, the trade in goods and services 1998-2000 are used for intra-EU15 countries. However, the data for intra-trade concerning the candidate countries (between EU 15 and candidates as well as among candidate countries) is based on only the merchandise trade in 2001.

D. The Proposed Formula 1. Formula As a basis for deriving the voting weight of each Member, the above-discussed requirements of simplicity, transparency, and robustness to changing economic climate and Member’s relative power must be fulfilled. Upon testing a number of options, we pro-

202

Thomas Cottier / Satoko Takenoshita

pose the following formula,17 which is linear in share of vote and individual country variables: V = a + bCs + cYs + dP + eLs where V = number of votes a = basic votes (a = 0 in case of no basic votes) C = contribution (= three year average trade share in goods and services) Y = GDP over three year average P = openness L = population, and the subscript s denotes ‘share’. For actual calculation of the voting weight for 2003, variables utilized are: x Contribution 2003 whose calculation is based on international trade statistics for the years 1998 – 2000; or revised contribution 2003 (excluding intra-EU trade among EU28 countries) x GDP average over the period of 1998 – 2000; x Openness average over the period of 1998 – 2000; x Population statistics of 2001. The variables of contribution and GDP are allocated each with 500 votes whereas variables of basic votes, openness and population with 250 votes. The former two variables are provided with larger weights in respect to the latter. This is to place adequate em17

Other formulae examined were: 1. linear model in value: V = a + bC + cY + dP +eL; 2. square root model: V = a + b(Cs)1/2 + c(Ys)1/2 + d(P)1/2 + e(Ls)1/2; 3. log-log model: log(V) = log(a) + b log(C) + c log(Y) + d log(P) + e log(L) + f log(G) which is equivalent to: V = ĮCbYcPdLe; where Į is a normalization factor; 4. lin-log model: V = a + b log(C) + c log(Y) + d log(P) + e log(L) + f log(G). These formulae are dismissed either because the distribution of vote is heavily biased towards a particular country or region, and/or the attribution of weight of each variable is unclear and hence lacks transparency. Cottier / Takenoshita (2003) examined also non-linear model (V = a + bCs0.75 + cYs0.75 + dP0.75 + eLs0.75) in detail. As the voting weight and power of this model strongly favoured EU in relation to the US, we focus our study on the linear model in this paper.

Towards Supplementary Weighted Voting

203

phasis on economic power and trade performance, the latter being the most important and most legitimate element in assessing the position in power relations in the WTO context. The latter variables, on the other hand, are given lower weight and seek to avoid too strong a concentration of votes to economically strong Members and to assure a reasonable balance between the industrial and developing groups. The coefficients are designed so as to achieve as closely as possible the set number of each variable vote. The sum of all variables is finally rounded for each country. 2. Basic Requirements and the Formula The proposed formula is simple as the number of votes just amounts to the summation of share of variables. It is transparent as the formula is simple, being comprised of basic economic and social variables. It is also robust to changing economic climate and Member’s relative power for three reasons. Firstly, multiple variables are used, none of which has particularly strong dominance in determining the number of votes. Secondly, as the economic variables represent the average of most recent three-year period, wide fluctuations due to short-run cyclical effects are avoided. At the same time, it is effective in capturing the recent change in economic performance and ranking. Thirdly, by fixing the weight of variables, the problem faced by the IMF of an increasing proportion of quota variable in total votes arsing from inflation and growth over time, can be avoided. The problem of avoiding de facto veto power and the need for a balanced representation between the two groups of countries, industrial and developing, will be examined in the next subsection when the formula is quantified. In this respect, two calculations are made under different hypothesis. Under Hypothesis 1, the EU Member countries are treated individually according to current practice and hence the existing contribution allocation is used. Hypothesis 2 makes two changes to this. Firstly, the EU28 will be treated as one entity in assessing the voting weight. This affects the variable of openness of the EU. Secondly, the contribution is recalculated by excluding intra-EU28 trade.18 Under the existing alloca18

The recalculation of contribution was made according to the current WTO practice. In subtracting intra-EU28 trade, intra-EU15 trade is based on trade in goods and services during the period of 1998-2000. However, the trade concerning the 10 new and 3 candidate countries, the data is based only on merchandise trade in 2001 for trade among

204

Thomas Cottier / Satoko Takenoshita

tion of WTO contribution, which is proportional to the share of trade in goods and services, intra-EU trade is included. As shown in Table 1, the current proportion of contribution of EU28 is over 40%. This inevitably leaves the EU with a disproportionately high voting weight. In order to deal with this problem, the Hypothesis 2 is provided as an estimated case in which the EU is treated as one entity. Considering that the EU constitutes a single custom area and that the voting behaviour of the EU is completely homogeneous and controversial votes cannot be made amongst EU members, it is reasonable to assume the EU as one entity, by excluding intra-EU trade. It is in this respect that calculations made in our first paper are amended and further developed. E. Calculated Voting Weight In quantifying the formula, three models are tested under each hypothesis: x model a) without basic votes (V = bCs + cYs + dP + eLs); x model b) with basic votes (V = a + bCs + cYs + dP + eLs); x model c) with basic votes and exclusion of population variable (V = a + bCs + cYs + dP) Hence given the weight of each variables described above, the model a and c will have 1500 whereas model b with 1750 votes in total. Table 2 shows the regional representation of the voting weight. Under Hypothesis 1, whose calculation is based on the current allocation of contribution, the proportion of voting weight for the EU is as high as about 30 percent. This means that the sole EU is endowed with de facto veto power if the majority rule of three quarters were to be applied. Under Hypothesis 2, on the other hand, these effects are mitigated and eliminated by excluding intra-EU trade for the purpose of allocating budgetary contributions. Hypothesis 2 is divided into two modes in treating the EU28. Model b and model c provides the basic votes to the EU as one entity whereas the model b’ and model c’ provides them with basic votes new/candidate countries and between new/candidate and EU15 countries. Although trade in services exhibits increasing trend, the proportion of trade in services for those 13 countries are considered to be still small relative their total trade. The actual EU28 contribution as a whole will not be too different from or slightly lower than our calculation.

Towards Supplementary Weighted Voting

205

times 28 EU Members. The difference between the two modes is that the EU has slightly less voting weight than the US under models b and c. The situation is revered for models b’ and c’ where the EU exercises slightly more rights than the US.19 In comparing the two Hypotheses, there are two notable differences. Firstly, the distribution of voting weight is more favourable for developing countries in Hypothesis 2. Secondly, the difference in proportion of representation between the US and the EU is much reduced in the Hypothesis 2. Under Hypothesis 1, the EU is endowed with a voting weight almost double compared to the US. In contrast, the Hypothesis 2 balances the voting weights of the two economically influential countries. Hypothesis 2 also provides a much more balanced representation of interests between industrial and developing countries compared to Hypothesis 1. Furthermore, Hypothesis 2 allows to avoid veto powers even when 3/4th majority is required since the vote of the EU is limited up to 20% of the total vote. Under Hypothesis 1, this is not the case and would allow the EU to control critical votes in a decisive manner. It is interesting to note the effects from the variables of basic vote and population. Both variables generally provide a higher share of votes to the group of developing countries as a whole. Yet, the gain in voting weight from the insertion of these variables varies among different regions. In particular, the population variable almost exclusively favours Asian countries, since about 60% of total world population is domiciled in that region. The inclusion of population increases the voting weight of Asian region by six percent for both hypotheses (model b versus model c). As to other regions, the population variable negatively affects the voting weight, except for Africa. The basic vote, on the other hand, has the effect to increase voting weight for all developing countries in the different regions, except for Asia (model a versus model b).

19

Of course, US-EU relation as well as industrial-developing relation in voting weight is subject to change over time, depending on the future development of different variables.

206

Thomas Cottier / Satoko Takenoshita

Table 2: Regional Representation from Calculated Voting Weight (Calculated Vote Share in Percentage) Hypothesis 1 Model

a

Industrial Countries

b

c

58.60

53.97

59.75

USA

16.67

14.41

15.89

EU28

28.73

27.44

30.24

Japan

7.53

6.58

7.28

41.40

46.03

40.25

7.20

10.18

9.95

21.73

20.75

14.49

Europe

0.73

1.14

1.34

Middle East

2.13

2.69

2.87

Americas

9.60

11.26

11.62

Developing Countries Africa Asia

Hypothesis 2 Model

b

b’

c

c’

52.46

46.07

48.80

50.83

53.71

USA

18.71

16.12

16.19

18.01

17.97

EU28

18.58

16.00

18.71

17.08

20.11

Japan

8.26

7.23

7.21

8.07

8.02

47.54

53.93

51.20

49.17

46.29

7.86

11.50

10.58

11.54

10.62

24.90

24.15

23.57

18.35

17.77

Europe

1.26

1.99

1.72

2.13

1.94

Middle East

2.53

3.13

2.86

3.40

3.14

10.99

13.15

12.47

13.74

12.83

Industrial Countries

Developing Countries Africa Asia

Americas

Source:

a

Calculations by the authors.

The composition of the ten countries with the largest vote is shown in Table 3. Importantly, it shows that a high proportion of voting weight is distributed among countries from different regions. Although the first three positions are taken by a group of industrial countries – the US, the EU and Japan – in all the models, they are

Towards Supplementary Weighted Voting

207

mostly followed by China. When the population variable is taken into account, India is ranked at the fifth position. The remaining group consists of Brazil, Canada, Hong Kong, Korea, Mexico, Singapore and Taiwan. Hence, the voting weight – unlike in the Bretton Woods system – is not concentrated in industrial countries, but more balanced and shared with countries from the regions of Asia and Western Hemisphere as well. The voting weight of China deserves particular attention. By combining the voting weight of mainland China, Hong Kong, and Macao, China will have a larger voting weight than Japan and ranks third. Considering the growing trend of China’s GDP and trade, it is probable that China in the future will exercise a counterweight to the US and the EU. Table 3: Ranking of Ten Highest Vote Countries from Calculation Hypothesis 1 Model a Country

Model b

Votes

Country

Model c

Votes

Country

Votes

EU28

431

EU28

480

EU28

453

US

250

US

252

US

238

Japan

113

Japan

115

Japan

109

China

93

China

95

China

35

India

60

India

62

Canada

34

Canada

34

Canada

35

Hong Kong

25

Mexico

26

Mexico

27

Mexico

23

Hong Kong

24

Hong Kong

26

Korea

22

Brazil

24

Brazil

26

Singapore

19

Korea

22

Korea

24

Taiwan

19

Hypothesis 2 Model a Country

Model b

Votes

Country

Model c

Votes

Country

Votes

US

281

US

283

US

270

EU28

279

EU28

281

EU28

256

Japan

124

Japan

127

Japan

121

China

98

China

101

China

42

India

62

India

64

Canada

42

208

Thomas Cottier / Satoko Takenoshita Hypothesis 2 Model a

Country

Model b

Votes

Country

Model c

Votes

Country

Votes

Canada

42

Canada

44

Hong Kong

32

Hong Kong

31

Hong Kong

33

Mexico

27

Mexico

30

Mexico

32

Korea

27

Korea

27

Korea

29

Singapore

24

Brazil

26

Brazil

28

Taiwan

23

Model b’

Model c’

EU28

327

EU28

301

US

283

US

269

Japan

126

Japan

120

China

100

Canada

42

India

64

China

41

Canada

43

Hong Kong

32

Mexico

32

Mexico

27

Hong Kong

32

Korea

26

Korea

29

Taiwan

23

Brazil

27

Singapore

23

Source: Calculations by the authors

The models tested do not yet separate the effect of openness, but of population and basic votes variables, as we are mainly interested in exploring the impact of these parameters. The effect of openness in general is noted in the section of variable C.3. In order to demonstrate the impact clearly, we would need to make another calculation without openness – but it is intuitively clear from Table 1, that the openness variable strongly favours developing countries as their economies are generally less restricted than those of industrialized countries. F. Assessment It is submitted that the models designed for weighted voting system under WTO amount to a realistic and fair picture of voting rights’ allocations. In so doing, the basic requirements of simplicity, transparency, robustness to changing economic climate and Member’s relative power, balanced representation and avoidance of veto power were sought to be fulfilled. In particular Hypothesis 2, in

Towards Supplementary Weighted Voting

209

which intra-EU28 trade is excluded in allocating the contribution, demonstrates a better balance of representation between groups of industrial and developing countries as well as representation between the EU and the US. Further, unlike Hypothesis 1, Hypothesis 2 provides a better option for majority voting as individual blocking votes by individual countries are avoided. Hence we conclude that Hypothesis 2 is more suitable in fulfilling the requirements for legitimate weighted voting system. Regarding the choice of variables, the insertion of a population variable may be questionable as its effect are exclusively limited to the benefit of Asian countries, even though the variable is suggested to enhance the impact and idea of democratic representation. The introduction of basic votes proves effective in providing more representation to the group of developing countries in general, covering most regions. Evidently, the choice of model and negotiations to achieve a shared perception on factors and voting rights allocations amounts to a major challenge. The experience within Bretton Woods and within the European Union shows that this is a difficult task, but not an excluded and unrealistic one. IV. Assessing of Voting Power It is important to note that the allocation of voting rights does not exclusively determine the impact of different Members on decisionmaking process. Effective voting powers cannot be defined in a static manner, but depends on many additional factors, such as the power of the pen (persuasion), conduct and coalitions. Statistical models developed in a domestic context are employed in this section with a view to further refine the analysis. Based on the calculated voting weight, we suggest an assessment of voting powers for different majority rules. In so doing, Shapley-Shubik20 and Bahnzaf21 Power Indices (hereinafter SSI and BI, respectively) and the Inclusiveness Index22 are used for assessing the voting power of individual Members, while Decision Probability Index23 is used for assessing the efficiency of the voting body in achieving a successful passage of legislation. 20 21 22 23

See Shapley / Shubik (1954). See Straffin (1988). See König / Bräuninger (1998). See Coleman (1971).

210

Thomas Cottier / Satoko Takenoshita

A. SSI, BI, Inclusiveness and Decision Probability Indices The voting power (as different from voting rights or shares) is defined as ‘the chance [it] has of being critical to the success of a winning coalition’.24 It is the probability that a country is crucial (pivotal) in making a decision which turns a loosing coalition into a winning one. In assessing the power allocation, Widgrén provides three determinants of the outcome of the voting power: majority rule, assigned weights of the various countries, and voting behaviour.25 The difference between SSI and BI stems from assessing the voting behaviour determinant which is built on different assumption of homogeneity and independence, respectively, in these two models. The Shapley-Shubik Index (SSI) is formally defined as ( s  1)!( n  s )! Ii ¦ n! S M i , where S denotes a randomly chosen coalition, Mi is the class of minimum winning coalitions with respect to country i, n the number of countries and s the number of countries in the coalition S. It assumes a random ordering of voters and perceives that only a minimal winning coalition is required in passing a piece of legislation, as the remaining unexercised votes are not relevant as soon as a majority is achieved. This simulates coalition building and SSI is suggested to be more appropriate for a voting situation where there is strong coalition building and voters communicate with each other to a considerable amount.26 On the other hand, the Bahnzaf Power Index (BI) assumes a voting situation where each voter is independent and communication among voters is non-existent. The power is measured by the number of probabilities that the voter is able to alter the outcome by changing his or her vote. Hence the order in which voters join a coalition is not an important factor. BI is defined as: n 1

§1· ¨ ¸ ¦ SM i © 2 ¹ SSI is more appropriate for the power assessment under the WTO context, as decision-making is the outcome of repeated nego-

Ei

24 25 26

Shapley / Shubik (1954). Widgrén (1995), 114. Straffin (1998).

Towards Supplementary Weighted Voting

211

tiations and bargaining. Thus, our analysis on voting power will be focused on SSI. BI will be presented as a reference in Appendix C which shows the voting power of all individual countries. The Inclusiveness Index assesses the absolute chance that a particular Member will be part of and incorporated in potentially collective decision-making. The index can be used as a quantified expectation of a Member as to the legislative gains from potential legislation. In formal terms, the Inclusiveness Index is expressed as: ¦ v( S ) S Ž N ,iS Z i (v ) ¦ v( S ) SŽN

Finally, the Decision Probability Index assesses the ability of the voting body to reach a successful passing of legislation. The index reflects the number of all winning coalitions, divided by the number of all possible coalitions, and is expressed as: ¦ v( S ) SŽN P (v ) 2n The Decision Probability of the current consensus rule is P(v)=1/(2120), treating the EU28 as one voting entity. Hence the probability to reach a successful passage of legislation is infinitely close to zero. B. Quantified Power Indices Using the indices above, our weighted voting models are tested for three majority rules – simple majority, 2/3rd and 3/4th majority requirements. Further, each majority rule will be tested for double majority rules with which a second condition – simple majority rule (50% plus one) in terms of the number of membership – is imposed. As majority ruling requiring simple majority can be achieved based upon mere voting rights with 3 to 6 members (depending on the models tested) if the top largest voting weight-holders form a coalition, 7 to 19 members are required for 2/3rd majority and 13 to 37 members for 3/4th majority, the introduction of a double majority will provide protection for smaller countries by increasing the possibility of being included in making a determining decision. In quantifying the power indices, Monte Carlo technique is being used as the use of algorithms for computation of these power

212

Thomas Cottier / Satoko Takenoshita

indices becomes extremely complex with 120 voters (the current 148 WTO members, but counting the EU 28 as one single entity).27 1. SSI and BI Under the current consensus rule, each country is theoretically pivotal in decision-making hence both SSI and BI is 1.00 for all Members and normalized SSI and BI yields 1/120 or 0.0083 for each Member. An introduction of weighted voting system inevitably reduces the voting power of most Members at below 0.0083 for more than 100 countries. Table 4 presents the regional distribution of voting power. For example, under 2/3rd majority rule under model (a) of Hypothesis 1, a Member who is pivotal in a voting situation is 64.5% likely to be from the group of industrial countries, whereas 35.5% from the group of developing countries. On the other hand, with 2/3rd majority rule under model (a) of Hypothesis 2, the probability is 55.7% for industrial countries and 44.3% for the group of developing countries. In general, Hypothesis 2 provides a more balanced distribution of voting power between the two groups of countries than Hypothesis 1. The difference between Hypothesis 1 and 2 again reflects the concentration of voting power of the EU28. Not only would Hypothesis 1, as discussed above, allow for a concentration of voting weight with the EU28 in comparison to the US; the voting power of the EU28 would be much higher than that of the US (see Table 5). Under Hypothesis 1, the EU is given about twice as much voting weight as the US, while the voting power of the EU is 2.5 time as much as the one of the US under this model Hypothesis 2, on the other hand, smoothes out such large difference in power by balancing the two. It is interesting to find that an introduction of double majority rule makes almost no change in the power relations.

27

The figures presented in this paper are results from 3 million trials although the figures showed significant level of convergence after 2 million.

Table 4: Regional Distribution of SSI Voting Power (Consensus, Simple, 2/3rd, and 3/4th Majority Rule) Consensus Industrial (10)* 0.083 Devel. (110)

0.917

Africa (40)

0.333

Asia (23)

0.192

Europe (7)

0.058

M. East (8)

0.067

SSI: normalized figure

Americas (32) 0,267

* EU28 is treated as one Hypothesis 1

Model a SM

2/3

Model b 3/4

SM

2/3

Model c 3/4

SM

2/3

3/4

Industrial

0.636 0.645 0.670 0.594 0.598 0.622 0.654 0.673 0.676

Developing

0.364 0.355 0.330 0.406 0.402 0.378 0.346 0.327 0.324

Africa

0.062 0.062 0.056 0.089 0.089 0.082 0.085 0.081 0.080

Asia

0.193 0.185 0.176 0.184 0.182 0.173 0.125 0.118 0.117

Europe

0.006 0.006 0.006 0.010 0.010 0.009 0.011 0.011 0.011

Middle East

0.019 0.018 0.017 0.024 0.024 0.022 0.025 0.023 0.023

Americas

0.083 0.083 0.075 0.099 0.098 0.092 0.100 0.095 0.093 Hypothesis 2 Model a SM

Industrial Developing

2/3

Model b 3/4

SM

2/3

Model c 3/4

SM

2/3

3/4

0.560 0.557 0.566 0.492 0.492 0.493 0.544 0.542 0.548 0.440 0.443 0.434 0.508 0.508 0.507 0.456 0.458 0.452

Africa

0.072 0.072 0.071 0.107 0.107 0.107 0.107 0.107 0.106

Asia

0.223 0.235 0.228 0.229 0.230 0.229 0.171 0.172 0.169

Europe

0.012 0.012 0.011 0.019 0.019 0.019 0.020 0.020 0.020

Middle East

0.023 0.023 0.023 0.029 0.029 0.029 0.031 0.032 0.031

Americas

0.101 0.101 0.100 0.123 0.123 0.123 0.127 0.128 0.126 Model b’ SM

Industrial

2/3

Model c’ 3/4

SM

2/3

3/4

0.523 0.522 0.528 0.575 0.572 0.587

214

Thomas Cottier / Satoko Takenoshita Hypothesis 2 Model b’ SM

Developing

2/3

Model c’ 3/4

SM

2/3

3/4

0.477 0.478 0.472 0.425 0.428 0.413

Africa

0.098 0.098 0.097 0.097 0.098 0.094

Asia

0.222 0.223 0.219 0.164 0.165 0.159

Europe

0.016 0.016 0.016 0.018 0.018 0.017

Middle East

0.026 0.026 0.026 0.029 0.029 0.028

Americas

0.116 0.116 0.115 0.117 0.118 0.114

Source: Calculations by the authors

Table 5a: Voting Power Indices for Top Ten Countries with Highest Votes for Simple, 2/3rd, and 3/4th Majority Rule Hypothesis 1 Model 1a Country

vote share

SSI SM

2/3

Inclusiveness 3/4

SM

2/3

3/4

EU28

0.287 0.363 0.372 0.367 0.917 1.000 1.000

USA

0.167 0.153 0.163 0.195 0.584 0.959 1.000

Japan

0.075 0.071 0.061 0.064 0.579 0.610 0.862

China

0.062 0.057 0.051 0.052 0.570 0.600 0.774

India

0.040 0.036 0.035 0.033 0.541 0.565 0.666

Canada

0.023 0.020 0.020 0.018 0.522 0.537 0.594

Mexico

0.017 0.015 0.015 0.014 0.517 0.529 0.573

Hong Kong

0.016 0.014 0.014 0.013 0.516 0.526 0.567

Brazil

0.016 0.014 0.014 0.013 0.516 0.526 0.567

Korea

0.015 0.013 0.013 0.012 0.515 0.524 0.561 Model 1b Country

vote share

SSI SM

2/3

Inclusiveness 3/4

SM

2/3

3/4

EU28

27.444 0.351 0.355 0.356 0.941 1.000 1.000

USA

14.408 0.134 0.138 0.162 0.560 0.967 1.000

Japan

6.575 0.061 0.057 0.058 0.557 0.634 0.920

China

5.432 0.049 0.047 0.048 0.553 0.615 0.848

Towards Supplementary Weighted Voting

215

Hypothesis 1 Model 1b India

3.545 0.032 0.031 0.030 0.535 0.571 0.702

Canada

2.001 0.018 0.017 0.017 0.519 0.542 0.618

Mexico

1.544 0.013 0.013 0.012 0.515 0.533 0.591

Hong Kong

1.487 0.013 0.013 0.012 0.514 0.531 0.588

Brazil

1.487 0.013 0.013 0.012 0.514 0.531 0.588

Korea

1.372 0.012 0.012 0.011 0.513 0.529 0.581 Model 1c Country

vote share

SSI SM

2/3

Inclusiveness 3/4

SM

2/3

3/4

EU28

30.240 0.394 0.413 0.372 0.501 1.000 1.000

USA

15.888 0.137 0.153 0.190 0.501 0.986 1.000

Japan

7.276 0.069 0.056 0.062 0.502 0.556 0.958

China

2.336 0.020 0.019 0.019 0.501 0.534 0.609

Canada

2.270 0.020 0.018 0.019 0.502 0.532 0.604

Hong Kong

1.669 0.014 0.014 0.014 0.501 0.524 0.576

Mexico

1.535 0.013 0.013 0.013 0.502 0.521 0.571

Korea

1.469 0,013 0.012 0.012 0.502 0.521 0.569

Singapore

1.268 0.011 0.010 0.010 0.507 0.518 0.559

Taiwan

1.268 0.011 0.010 0.010 0.507 0.518 0.559

Table 5b: Voting Power Indices for Top Ten Countries with Highest Votes for Simple, 2/3rd, and 3/4th Majority Rule Hypothesis 2 Model 2a Country

vote share

SSI SM

2/3

Inclusiveness 3/4

SM

2/3

3/4

USA

18.708 0.208 0.207 0.215 0.753 0.998 1.000

EU28

18.575 0.206 0.205 0.214 0.748 0.998 1.000

Japan

8.256 0.082 0.081 0.073 0.641 0.719 0.961

China

6.525 0.063 0.063 0.059 0.597 0.657 0.892

India

4.128 0.039 0.039 0.039 0.563 0.598 0.726

Canada

2.796 0.026 0.026 0.026 0.541 0.565 0.654

216

Thomas Cottier / Satoko Takenoshita Hypothesis 2 Model 2a

Hong Kong

2.064 0.019 0.019 0.019 0.530 0.549 0.615

Mexico

1.997 0.019 0.019 0.018 0.529 0.546 0.610

Korea

1.798 0.017 0.017 0.017 0.526 0.543 0.600

Brazil

1.731 0.016 0.016 0.016 0.525 0.540 0.594 Model 2b Country

vote share

SSI SM

2/3

Inclusiveness 3/4

SM

2/3

3/4

USA

16.116 0.179 0.179 0.181 0.752 0.999 1.000

EU28

16.002 0.177 0.178 0.179 0.748 0.999 1.000

Japan

7.232 0.072 0.072 0.070 0.636 0.791 0.989

China

5.752 0.056 0.057 0.056 0.596 0.708 0.959

India

3.645 0.035 0.035 0.035 0.561 0.631 0.831

Canada

2.506 0.024 0.024 0.024 0.541 0.588 0.725

Hong Kong

1.879 0.018 0.018 0.018 0.531 0.566 0.668

Mexico

1.822 0.017 0.017 0.017 0.529 0.563 0.665

Korea

1.651 0.015 0.015 0.016 0.526 0.558 0.652

Brazil

1.595 0.015 0.015 0.015 0.526 0.555 0.644 Model 2b’ Country

vote share

SSI SM

2/3

Inclusiveness 3/4

SM

2/3

3/4

USA

18.707 0.213 0.213 0.218 0.791 1.000 1.000

EU28

16.190 0.176 0.176 0.181 0.710 0.998 1.000

Japan

7.208 0.071 0.071 0.067 0.633 0.746 0.979

China

5.721 0.055 0.056 0.054 0.595 0.675 0.934

India

3.661 0.035 0.035 0.034 0.560 0.611 0.788

Canada

2.460 0.023 0.023 0.023 0.539 0.573 0.690

Hong Kong

1.831 0.017 0.017 0.017 0.529 0.554 0.642

Mexico

1.831 0.017 0.017 0.017 0.529 0.554 0.642

Korea

1.659 0.015 0.016 0.016 0.526 0.550 0.629

Brazil

1.545 0.014 0.014 0.014 0.524 0.545 0.618

To be continued.

Towards Supplementary Weighted Voting

217

Hypothesis 2 Model 2c Country

vote share

SSI SM

2/3

Inclusiveness 3/4

SM

2/3

3/4

USA

18.012 0.203 0.202 0.208 0.764 1.000 1.000

EU28

17.078 0.189 0.188 0.195 0.736 1.000 1.000

Japan

8.072 0.081 0.080 0.074 0.681 0.823 0.997

Canada

2.802 0.026 0.027 0.026 0.542 0.576 0.785

China

2.802 0.026 0.027 0.026 0.542 0.576 0.785

Hong Kong

2.135 0.020 0.020 0.020 0.532 0.558 0.714

Korea

1.801 0.017 0.017 0.017 0.527 0.549 0.680

Mexico

1.801 0.017 0.017 0.017 0.527 0.549 0.680

Singapore

1.601 0.015 0.015 0.015 0.524 0.544 0.664

Taiwan

1.534 0.014 0.014 0.014 0.523 0.542 0.653 Model 2c’ Country

vote share

SSI SM

2/3

Inclusiveness 3/4

SM

2/3

3/4

USA

20.107 0.229 0.229 0.242 0.782 1.000 1.000

EU28

17.969 0.196 0.196 0.208 0.717 1.000 1.000

Japan

8.016 0.080 0.077 0.069 0.679 0.746 0.994

Canada

2.806 0.026 0.026 0.025 0.542 0.568 0.740

China

2.739 0.025 0.026 0.025 0.541 0.566 0.732

Hong Kong

2.138 0.020 0.020 0.019 0.532 0.551 0.679

Mexico

1.804 0.017 0.017 0.016 0.527 0.543 0.651

Korea

1.737 0.016 0.016 0.016 0.526 0.531 0.646

Singapore

1.536 0.014 0.014 0.014 0.523 0.537 0.627

Taiwan

1.536 0.014 0.014 0.014 0.523 0.537 0.627

The table shows the list of ten countries with highest voting weight, denoted as vote-share, for each models. Then the calculated results for indices (SSI and Inclusiveness), tested for three majority rules (simple majority (SM), 2/3 and 3/4) are shown for each listed Member.

2. Inclusiveness As for the Inclusiveness Index, the consensus rule ensures the absolute inclusiveness to the decision-making for all countries. The

218

Thomas Cottier / Satoko Takenoshita

introduction of a weighted voting system significantly lowers the probability to be included for all members, except for the EU and the US who enjoy absolute levels of inclusiveness in almost all models for 2/3rd and 3/4th majority requirements. This ensures that the interests of two economically influential countries to be fully reflected in the process of decision-making. For other Members, the inclusiveness is reduced gradually according to the voting weight, by up to 50%. The double majority rule slightly increases the inclusiveness index for most Members, except for the EU and the US and Japan in some instances. 3. Decision Probability The Decision Probability Index shown in Table 6, demonstrates that any move from consensus will significantly improve the decisionmaking efficiency. Self-evidently, the lower the majority rule, the higher the Decision Probability Index. The probability of decisionmaking under simple majority is about 50%, whereas they are more or less than 16% for 2/3rd and less than 5% for 3/4th majority rule under Hypothesis 2. Under Hypothesis 1, the probability is even higher. It amounts to 50% for simple majority rule, to over 20% for 2/3rd and less than 10% for 3/4th majority rule requirements. The double majority rule significantly decreases the index and the level of decrease is less for higher majority rule. This is because the higher the majority rule, the more likely it is that the larger number of Members is in any case included in the decisions made. Especially, since simple majority in terms of voting weight can be achieved by very small number of countries, the introduction of a double majority is an effective tool in safeguarding the possibility of a systematic exclusion of small countries in decision making. It should be noted that the different majority rules among the same model largely influence decision probability whereas it does not change the voting power for individual countries very much. This finding implies that once the allocation of voting weights are chosen for adoption, the attention in the choice of majority rule can be placed more on the efficiency of the voting body and inclusiveness.

Towards Supplementary Weighted Voting

219

Table 6: Decision Probability Indices Simple Hypothesis 1

2/3rd

double

3/4th

double

double

Model 1a

0.499

0.271

0.218

0.131

0.090

0.064

Model 1b

0.499

0.270

0.201

0.126

0.059

0.047

Model 1c

0.501

0.264

0.231

0.136

0.093

0.069

Hypothesis 2 Model 2a

0.499

0.285

0.154

0.100

0.044

0.035

Model 2b

0.499

0.292

0.116

0.083

0.017

0.016

Model 2b’

0.499

0.290

0.138

0.095

0.027

0.023

Model 2c

0.500

0.287

0.138

0.093

0.022

0.021

Model 2c’

0.500

0.286

0.162

0.106

0.037

0.033

Source: Calculations by the authors.

4. Assessment of the Power Analysis The power analysis in this section makes it that the Hypothesis 2 is more appropriate in balancing the voting power between the group of industrial countries and developing countries, as well as between the EU and the US. The result confirms the analysis on voting weight discussed in the previous section. The assessment of various voting power indices reveals that there is a clear trade-off when introducing a weighted voting system: the loss of voting power and of inclusiveness for a majority of Members versus improved efficiency in decision-making process. It should be reminded, however, that power relations also exist under the current consensus rule as recourse to blocking powers is not evenly spread in reality. As described above, smaller countries or groups of countries cannot readily take recourse to blocking without facing difficulties elsewhere. The introduction of weighted voting may merely make such power relations more transparent. In this sense, the loss of voting power of many countries in real terms may not be as significant as they may think and may be more than compensated by a more efficient organization protecting their MFN based market access rights. Finally, it is important to note that the statistical models employed ignore the dynamics of flexible coalition building in the WTO, and more empirical work would be required to be able to make conclusive statements on the impact of power relations

220

Thomas Cottier / Satoko Takenoshita

among Members under consensus and weighted voting. The trend towards interests groups in a variable geometry, depending on different subject matters, renders such estimations very difficult and hardly predictable, except for the fact that weighted voting clearly would render stalemates and blocking of negotiations much less likely than today under the current regime where consensus is not supplemented by a meaningful and realistic voting system in the WTO. In the same vein, it will be important to assess the impact of increasing regionalism and the formation of trading blocks in coming years which the WTO is supposed to bring and hold together under the multilateral trading system. V. Conclusion In light of an increasing complexity of WTO negotiations and the risk of ongoing and time-consuming stalemates, the finding of a clear imbalance between the negotiating process entirely dependent on consensus and dispute settlement which no longer allows for unilateral blocking, this paper suggest to supplement consensus with a system of weighted voting. Sovereign equality of States and the law of international organizations do not exclude to take into account factual differences among Members of the WTO in defining in law voting rights and thus influence in legal terms. In this paper we presented models for weighted voting system to WTO decision-making subsequent to our previous paper, based on different variables – contribution to the WTO, GDP, market openness, population and / or basic votes. A legitimate decisionmaking based on weighted voting model was sought by achieving simplicity, transparency, robustness to a changing economic climate and a Member’s relative power, more balanced representation, and no veto power to any individual voting entity. The former three were achieved by the function of the formula we proposed, and the latter two can be achieved by excluding intra-EU trade and defining the EU28 as a single unit. A meaningful model of weighted voting would require this in order to provide a reasonable balance of power between industrial and developing countries and between the US and the EU alike. (As a practical matter, it would also entail a redefinition of EU contributions to the WTO to some 20%, with the difference to be taken up by other Members) The models presented show that a reasonable balance can be achieved among the major trading blocks. Smaller countries formally lose voting powers but benefit from a higher efficiency of the process. In light of a limited

Towards Supplementary Weighted Voting

221

blocking powers of these groups, the trade off may well be worth considering their stakes in the multilateral trading system.

References Richard E. Baldwin (1994), Towards an Integrated Europe, London (Centre for Economic Policy Research) 1994. Claude E. Barfield (2001), Free Trade, Sovereignty, Democracy: The Future of the World Trade Organization, Washington/D.C. (AEI Press) 2001. Ariel Buira (2003), The Governance of the IMF in a Global Economy, Research Papers of the Intergovernmental Group of Twenty-Four on International Monetary Affairs and Development, Washington/D.C. 2003 (available at http://www.g24.org/ buiragva.pdf). James S. Coleman (1971), Control of Collectivities and the Power of a Collectivity to Act, in: Bernhardt Lieberman (ed.), Social Choice, New York (Gordon & Breach) 1971, 269-300. Thomas Cottier / Satoko Takenoshita (2003), The Balance of Power in WTO Decision-Making: Towards Weighted Voting in Legislative Response, in: Aussenwirtschaft 59 (2003), 171-214. William Davey (2003), Mini-Symposium on the Desirability of the WTO Permanent Panel, in: Journal of International Economic Law 6 (2003) No. 1, 175-261. Sara Dillon (1999), Fuji-Kodak, the WTO, and the Death of Domestic Political Constituencies, in: Minnesota Journal of Global Trade 8 (1999) No. 2, 197-248. Claus-Dieter Ehlermann (2002a), Six years on the Bench of the ‘World Trade Court’. Some Personal Experiences as Member of the Appellate Body of the World Trade Organization, in: Journal of World Trade 36 (2002) No. 4, 605-639. Claus-Dieter Ehlermann (2002b), The Evolving WTO Constitution: Tensions between Dispute Settlement Processes and the Diplomatic and Treaty-making Activities of the WTO, Paper presented at the 2nd Annual WTO Conference, Dispute Resolution and the WTO Constitution, The British Institute of Interna-

222

Thomas Cottier / Satoko Takenoshita

tional and Comparative Law, London, May 15th, 2002 (on file with the author). Joseph Gold (1974), Developments in the Law and Institutions of International Economic Relations – Weighted Voting Power: Some Limits and Some Problems, in: American Journal of International Law 68 (1974) No. 4, 687-708. IMF (2001), Alternative Quota Formula: Considerations, Prepared by the Treasurer’s and Statistics Departments, Washington/D.C. 2001, available at http://www.imf.org/external/np/ tre/quota/2001/eng/aqfc.pdf (14.7.2005). IMF (2002), Transcript of an Economic Forum ‘Governing the IMF’, September 17th, 2002, Washington/D.C., available at http:// www.imf.org/external/np/tr/2002/tr020917a.htm (14.7.2005). Konstantin J. Joergens (1999), True Appellate Procedure or only a Two Stage Process? A Comparative View on the Appellate Body under the WTO Dispute Settlement Understanding, in: Law and Policy in International Business 30 (1999) No. 2, 193-230. Thomas König / Thomas Bräuninger (1998), The Inclusiveness of European Decision Rules, in: Journal of Theoretical Politics 10 (1998) No. 1, 125-142. Markus Krajewski (2001), Verfassungsperspektiven und Legitimation des Rechts der Welthandelsorganisation, Berlin (Duncker & Humblot) 2001. Matthias Oesch (2003), Standards of Review in WTO Dispute Resolution, Oxford (Oxford University Press) 2003. Ernst-Ulrich Petersmann (ed.) (2002), Preparing the Doha Development Round: Improvements and Clarifications of the WTO Dispute Settlement Understanding, Conference Report, Robert Schuman Centre for Advanced Studies, Florence (European University Institute) 2002. Lloyd S. Shapley / Martin Shubik (1954), A method for evaluating the distribution of power in a committee system, in: The American Political Science Review 48 (1954) No. 3, 787-792. Reprinted in: Alvin E. Roth (ed.), The Shapley Value, Cambridge/MA (Cambridge University Press) 1988, 41-48. Andrew Shoyer (1998), The Future of WTO Dispute Settlement, in: Proceedings of the American Society of International Law 92 (1998), 75-79.

Towards Supplementary Weighted Voting

223

Philip D. Straffin (1988), The Shapley-Shubik and Banzhaf power indices as probabilities, in: Alvin E. Roth (ed.), The Shapley Value, Cambridge/MA (Cambridge University Press) 1988, 71-82. Alan O. Sykes (1999), Regulatory Protectionism and the Law of International Trade, in: University of Chicago Law Review 66 (1999) No. 1, 1-46. Mika Widgrén (1995), Voting power and control in the EU: the impact of the EFTA entrants, in: Richard Baldwin / Pertti Haaparanta / Jaakko Kiander (eds.), Expanding Membership of the EU, Cambridge/MA (Cambridge University Press) 1995, 113-142. WTO (1995). WTO Analytical Index. Guide to GATT Law and Practice, 2 vols., Geneva (World Trade Organization) 1995.

224

Thomas Cottier / Satoko Takenoshita

Appendix A: Evolution of the IMF Quota Formulas28 The Original Bretton Woods Formula: QC = (0.02Y + 0.05R + 0.010M + 0.10V)(1 + X/Y) where: QC = Calculated Quota Y = National income R = Gold and foreign exchange reserves X, M = Average annual exports or imports over five-year period V = Maximum fluctuation in exports defined as the difference between the highest and lowest value of exports during five-year period 1962/63 Revision of the Formula and Multi-Formula System: The change in quota calculation was made by utilizing a dual structure of ten formulae, with different sets of data for each structure, each consisting of revised Bretton Woods Formula and four derived formulae. The use of different set of data mirrors the improvements in the reporting of invisible transactions and transfers. For example, P, the current payments in Set II Data, replaced M, merchandise imports in the Set I Data. Further, the original coefficients were reduced by half for Revised Bretton Woods. Revised Bretton Woods: Q1 = (0.01Y + 0.025R + 0.05M + 0.2276V)(1 + X/Y) Q*1 = (0.01Y + 0.025R + 0.05P + 0.2276VC)(1 + C/Y) where: Y, R, X, M = as defined in original Bretton Woods formula Q1 = Quota calculated with Set I data Q*1 = Quota calculated with Set II data C, P = Average annual current receipts or payments over a recent five-year period V, VC = Variability of annual exports or current receipts, defined as one standard deviation from the centered five-year moving average, for a recent 13year period

28

Source: IMF (2001).

Towards Supplementary Weighted Voting

225

Modified Formulae on Set I Data (that use trade data): Scheme III: Q2 = (0.0065Y + 0.078M + 0.5065V)(1 + X/Y) Scheme IV: Q3 = (0.0045Y + 0.070M + 0.9622VC)(1 + X/Y) Scheme M4: Q4 = 0.005Y + 0.044M + 0.044X + 1.044V Scheme M7: Q5 = 0.0045Y + 0.039M + 0.039X + 1.304V Modified Formulae on Set II Data (that use data for trade, invisible transactions and transfers): The coefficients remain the same but V and X are replaced by VC and C, respectively, for each scheme. Scheme III: Q*2 = (0.0065Y + 0.078P + 0.5065VC)(1 + C/Y) Scheme IV: Q*3 = (0.0045Y + 0.070P + 0.9622VC)(1 + C/Y) Scheme M4: Q*4 = 0.005Y + 0.044P + 0.044C + 1.044VC Scheme M7: Q*5 = 0.0045Y + 0.039P + 0.039C + 1.304V Calculated quota: Qc = Max[ Mean (Q1, Q*i), Qˆ] where: Qˆ = Mean of the lowest two of the Mean (Qi, Q*i), i = 2 to 5 and the value of Qi (i = 2 to 5) and Q*i (i = 1 to 5) have been normalized so that their totals equal that of Q1. 1981/82 Revision of the 1962/63 Formulae (that use GDP data and a broader definition of reserves): There are four changes made. First, the five formulae of Set I Data are eliminated and those of Set II Data are utilized exclusively. Second, national income was replaced by GDP for Y. Third, measure of R, reserves, is broadened. Finally the coefficients of four derived formulae are reduced by 20 percent to soften the impact of sharp price rise due to the oil shocks of 1973/74 and 1979. Bretton Woods: Q1 = (0.01Y + 0.025R + 0.05P + 0.2276VC)(1 + C/Y) Scheme III: Q2 = (0.0065Y + 0.0205125R + 0.078P +0.4052VC)(1 + C/Y) Scheme IV: Q3 = (0.0045Y + 0.03896768R + 0.07P + 0.76976VC)(1 + C/Y)

226

Thomas Cottier / Satoko Takenoshita

Scheme M4: Q4 = 0.005Y + 0.042280464R + 0.044(P + C) + 0.8352VC Scheme M7: Q5 = 0.0045Y + 0.05281008R + 0.039(P + C) + 1.0432VC where: Y = GDP in a recent year R = Average value of gold, SDRs, ECUs, IMF reserve positions, and foreign exchange reserves in a recent year C, P, VC = as defined in 1962/63. Calculated Quota: QC = Max (Q1, Mean of lowest two of Q2, Q3, Q4, Q5) where the values of Qi (i= 2 to 5) have been normalized so that the totals of Qi equal that of Q1.

Towards Supplementary Weighted Voting

227

Appendix B: Source of Variables Contribution: WTO contribution share is determined proportional to each Member’s share (%) of total trade of all Members, including trade in goods, services and intellectual property rights for the recent past three years for which data is available. When a Member has less than 0.015 share of total trade, a minimum contribution of 0.015 per cent is allocated in such case. The contribution of 2003 is used for the calculation of voting weight, which covers the trade period of 1998–2000. The source to WTO contribution calculation is Balance of Payments statistics (BoP) from the IMF14. When IMF BoP is not available, National Accounts, IMF International Financial Statistics (IFS) or World Bank Development Indicators (WDI) are referred to. The IMF BPM5 components included in goods and commercial services are: x x

x x

Goods: (same as BPM5 goods except nonmonetary gold held as a store of value) General Merchandise, Goods for processing, Repairs on goods, Goods procured in ports by carriers, Non monetary gold (except nonmonetary gold held as a store of value); Commercial services: (same as BPM5 services except gov ernment services) Transportation, Travel, Communications services, Construction services, Insurance services, Financial services, Computer and information services, Royalties and license fees, Other business services, Personal, cultural and recreational services.

GDP: The level of GDP is calculated as average over three years (1998, 1999, 2000). The data is taken from WDI On-line, whose original data source is World Bank national accounts data and OECD National Accounts data files. The exceptions to this are for following countries:

228

Thomas Cottier / Satoko Takenoshita

x

Cyprus: National Statistics – current GDP in local currency is converted into US$ by the exchange rate of the same year x Liechtenstein: UN Statistics Division (Internet: http:// unstats.un.org/unsd/cdbdemo). The number is the UN estimate, derived only from 1990 x Taiwan: National Statistics (GNP, not GDP) x Myanmar: UN Statistics Division (Internet: http://unstats. un.org/unsd/cdbdemo). The number is the UN estimate, derived only from the 1990 The current GDP in US$ is used for voting weight calculation, rather than PPP GDP (current GDP is the GDP in local currency converted into current US dollars whereas the PPP GDP is con verted into international dollars using purchasing power parity rates). Although the latter may be more appropriate for the purpose of WTO voting weight calculation, the former is utilized for its easiness of data availability. Openness: Openness is the percentage of average import value to average GDP level both for the period of 1998–2000. The import value is taken from WDI On-line, which takes the data originally from IMF BoP and World Bank staff estimates. The exceptions to this are for following countries: x x x

x x

Democratic Republic of the Congo: DOTS taken from IMF Direction of Trade Statistics 2001 Brunei Darussalam: DOTS taken from IMF Direction of Trade Statistics 2001 Taiwan: National Statistics, the undefined imports (C.I.F.), undefined whether goods and services are included (Internet: http://www.stat.gov.tw/bs8/stat/english1.htm) Cyprus: IFS taken from IMF Direction of Trade Statistics 2001 Liechtenstein: Government website; the import figure was not available and it is substituted by the figure of industrial exports, originally denoted in CHF then converted into US$ for 1999 and 2000 (Internet: http:// llvweb.liechtenstein.li/lisite/html/liechtenstein/index.jsp). Exchange rates for 1999 and 2000 are taken from Balance of Payments Yearbook 2002, Country Tables, IMF

Towards Supplementary Weighted Voting

x x x

229

Qatar: IFS for 1998 and 1999 and DOTS for 2000 United Arab Emirates: IFS taken from IMF Direction of Trade Statistics 2001 Members of European Union (Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden and United Kingdom): The imports from other EU members are subtracted from WDI data. Such intra-EU import statistics is taken from Direction of Trade Statistics – Yearbook 2001, IMF

Population The source WDI On-line taking data originally from World Bank staff estimates from various sources including the UN Statistics Division’s Population and Vital Statistics Report, country statistical offices and Demographic and Health Surveys from national sources and Macro International. The exceptions to this are for following countries: x x

Taiwan: National Statistics Liechtenstein: Office of National Economy, the figure in 2001 is not available is taken from 2000.

Fritz Breuss

Does the ‘Development Round’ Foster Development? I. II.

Introduction Multilateral Trade Liberalization and Development A. Openness, Growth and Poverty Reduction 1. The Trade and Development Linkage 2. What Determines the Huge Income Differences in the World? B. Is WTO Membership Good for Trade at All? III. After Eight GATT Rounds – Markets Remain Still Subtly Protected: Market Access Issues A. Patterns of Protection B. Divergence of Applied Tariffs and their Bindings C. Developing Countries Face Higher Barriers to their Exports than Industrial Countries D. Tariff Peaks and Tariff Escalation 1. Tariff Peaks 2. Tariff Escalation E. Contingent Protection 1. Antidumping Measures 2. Standards and Non-tariff Barriers to Trade F. Trade Preferences G. Regionalism versus Multilateralism 1. Increasing Attractiveness of RTAs 2. Does the EU Need the WTO at All? H. Developing Countries’ Market Access in Agriculture 1. Costs of Agricultural Distortion 2. Removing Agricultural Subsidies alone is Negative for Developing Countries

232 235 235 235 240 244 246 246 248 249 252 253 260 264 264 265 266 270 270 274 277 278 279

232

Fritz Breuss

3. Agricultural Liberalization Benefits all Countries I. Developing Countries’ Market Access in Textile and Clothes IV. Gains from Trade Liberalization under a Successful Doha Round A. Formula Approaches to Market Access Negotiations 1. The General Procedure of Negotiations 2. The Danger of Preference Erosion for the Developing Countries B. How to Estimate Gains from Multilateral Trade Liberalization? 1. UNCTAD Simulations 2. CGE Model Simulations with Imperfect Competition V. Conclusions References Appendix: Tables and Figure on World Trade

280 282 286 286 286 290 292 294 297 303 304 313

I. Introduction After the failed attempts in Seattle in late 1999, the Ministerial Conference in its Fourth Session in Doha, on November 9-14, 2001 launched the agenda for a new comprehensive round of multilateral trade negotiations. At the behest of the European Union (EU), the ministerial declaration emphasised that the Doha Round should provide a major opportunity for developing countries. Consequently the agenda for the new WTO round has been coined the ‘Doha Development Agenda’ (DDA). On September 10-14, 2003 the fifth Ministerial Conference in Cancún ended without reaching a consensus. According to press reports and subsequent statements by those present at that meeting, the apparent and proximate cause of the Ministerial’s collapse was a failure to agree on launching formal negotiations on the so-called Singapore Issues. Others, however, have put forward alternative explanations for the meeting’s failure, including poor chairmanship of the Ministerial meeting by Mexicos’s Foreign Minister, Mr. Luis Ernesto Derbez; a failure to agree on the modalities for negotiations on agricultural trade barriers,

Does the ‘Development Round’ Foster Development?

233

export subsidies, and domestic support policies; the inability of many WTO members to negotiate or discuss many issues simultaneously during and before the Cancún Ministerial Conference; and a perception that some national representatives in Cancún were not prepared to go beyond pre-determined demands of others and showed little propensity to ‘negotiate seriously’ with other delegations.1 The Doha Round was after all aiming at opening markets in order to foster growth and alleviate poverty in the developing world. In this respect Cancún was a ‘disaster’ which could badly hit the developing countries, in particular the least developing countries (LDCs), notwithstanding the emergence of the G90 and G22.2 Although the previously-agreed commitments by WTO members are still binding, the Cancún failure may have political consequences. On the one hand liberalization (market access) is delayed; on the other hand representatives of the United States and the EU immediately afterwards expressed their sympathy with a switch in their trade policy preferences towards more bi- or unilateralism. Before Cancún, public opinion was focusing on the interpretation of the TRIPs agreement, concerning the enforcement of intellectual property rights for affordable medicines (compulsory licenses, production of generic drugs) to protect health in case of diseases such as AIDS. The WTO TRIPs agreement of August 30, 2003 together with a similar solution of the EU for this problem (Council Regulation 953/2003) has been a major achievement in legal and economic terms and from a moral point of view. It is, however the only result of Cancún. Some commentators argue that the present round was not only ‘overburdened’ by the so-called Singapore Issues (which are primarily in the interest of the developed world) but also by putting development considerations at the centre of the Doha Round. Evenett even questions the new development mandate of the WTO at all. This agenda is an intricate menu of objectives and means that could lead to deceptive or undesired outcomes.3 The DDA contains a series of other key issues on which progress has been delayed as a follow up of Cancún (even the follow1 2 3

Evenett (2003), 11. Fontagné (2003), 3. Evenett (2003), 16.

234

Fritz Breuss

up conference in Geneva on December 15-16, 2003 did not bring a break-through): 1. Implementation-related issues: The developing world considers that developed economies have not fulfilled their commitments of Marrakech (1994) concerning the pace of liberalization in labour intensive industries (implementation of the Agreement on Textiles and Clothing – ATC), whereas their own commitments (TRIPs) are disproportionate. 2. Agriculture: Market access as well as the distorting domestic policies (subsidies in the USA and EU) are key issues. 3. Services (GATS). 4. Market access for non-agricultural products. 5. The four Singapore issues: the relationship between trade and investment policy; the interaction between trade and competition policies; transparency in public procurement practices; and trade facilitation practices (more efficient customs procedures).4 6. LDCs: Considering their specific needs: duty-free, quota-free market access for their products; Special and Differential Treatment (SDT); technical co-operation and capacity building etc. The DDA raises a lot of concerns which are dealt with in this article. In the next chapter the relationship between market access (trade liberalization) and development and its complex connections with poverty reduction are analysed. Their relationships are far from clear. The third chapter addresses market access issues, ranging from the problems of tariff peaks and escalation to more subtle forms of protectionism still in place vis-à-vis the developing countries. The fourth chapter looks at gains from further liberalization or more generally, the potential loss of welfare and income in the developing countries due to a delayed Doha Round. Conclusions are drawn at the end.

4

In the Doha Work Programme adopted by the General Cuncil of the WTO on 1 August 2004, the so-called ‘Singapore issues’ wer put aside. No work towards negotiations will take place within the WTO during the Doha Round (WTO (2004), 3).

Does the ‘Development Round’ Foster Development?

235

II. Multilateral Trade Liberalization and Development A. Openness, Growth and Poverty Reduction In the discussion of whether trade liberalization and foreign direct investment have helped to spur growth in China and India, the answer by participants at a conference organized by the IMF and India’s National Council of Applied Economic Research in New Delhi, November 14-16, 2003 was a unanimous and resounding ‘yes’, while in the case of broader capital account liberalization, opinions were more divided.5 Although poverty reduction in both India and China has been strongly correlated with economic growth, the wide regional differences within the two countries suggest hat other policies are also relevant in enhancing the ‘povertyreduction efficiency of growth’. Participants in this conference sought to identify the factors behind the two countries’ impressive track record over the past two decades. Interestingly, both are nowadays fast growing developing countries but had a quite different attitude towards trade liberalization and membership in world trade organizations. India is a GATT signatory country since July 8, 1948 and WTO member since January 1, 1995. China became a WTO member only on December 11, 2001. Both countries witnessed an increase in openness. In 1980 the ratio to GDP of total trade in goods and services in both India and China stood at about 15 percent. By 2001, this ratio had more than tripled to about 50 percent in China, while it had risen to only around 25 percent in India. 1. The Trade and Development Linkage In general the question whether more openness is better for growth and development and whether it is even a remedy for poverty reduction is not always easy to answer. There is a huge amount of development literature on this topic. The nexus of openness and poverty reduction is ambiguous and complex. Reimer surveys and classifies thirty-five studies of the emerging literature which quantifies how international trade affects the poor in developing countries.6 A general discussion of the problems connected with trade and development linkages is also offered by the UNCTAD.7 The 5 6 7

IMF (2003). Reimer (2002). UNCTAD (2003).

236

Fritz Breuss

manifold relationships between trade and development may be demonstrated with the Figures 1 and 2. Let’s look first to the mainstream ‘Panglossian’ view in Figure 1. Representatives of researchers, seeing a positive link between openness and growth (arrow 1) are Sachs and Warner; Dollar and Kraay; Dowrick; Dowrick and de Long; and Greenaway, Morgan and Wright.8 For a sample of up to 73 countries Greenaway, Morgan and Wright9 estimate in a panel over the period 1975 to 1993 a ‘core’ new growth theory model. Growth of real GDP per capita is explained by the GDP per capita as in 1965 (catching-up variable), the level of secondary school enrolment as in 1965, a terms of trade index, population, the ratio of gross domestic investment to GDP (proxy for capital input) and liberalization dummies. They use three different definitions of liberalization, those of Sachs and Warner, Dean et al. and one provided by the World Bank.10 The first is constructed on the basis of measuring whether an economy is open or not. Their index of openness is based on five criteria relating to non-tariff barriers, average tariff levels, the black market exchange rate, whether state monopolies exist for major exports and whether the economy is socialist or not. By contrast, Dean et al. are more qualitatively based.11 They use information on average nominal tariffs, QR coverage and average black market premia to identify when reform has taken place. The liberalization indicator provided by the World Bank is indicated by Structural Adjustment Loans (SAL) as one of a number of possible determinants of growth, export and investment performance.12 Greenaway, Morgan and Wright equate the first year of a SAL with a trade component as the beginning of the liberalization period. The authors find the following empirical results: A low initial GDP and high initial level of schooling are associated with faster growth in GDP per capita as are a higher investment ratio and favourable terms of trade movement. 8

9 10 11 12

Sachs / Warner (1995); Dollar / Kraay (2001, 2002); Dowrick (1994); Dowrick / de Long (2001); Greenaway / Morgan / Wright (1998). Greenaway / Morgan / Wright (1998). Sachs / Warner (1995), Dean et al. (1994); World Bank (1993). Dean et al. (1994). World Bank (1993).

Does the ‘Development Round’ Foster Development?

237

Faster population growth is associated with slower GDP per capita growth and liberalization appears to have on average a favourable and substantial (2.7%) impact on growth in years following liberalization. Looking at the impact of the timing of reform on growth they estimate the initial impact of liberalization and the effects of two years later. They find evidence of a ‘J curve effect’ of liberalization on per capita GDP growth: in year 1 of liberalization the impact on growth is negative (but not significant), positive (but insignificant) in year 2 and positive, larger and significant in year 3.13 Whereas Dollar and Kraay find no effect from openness to inequality (arrow [2]), Kuznets sees no effect at all between growth an income inequality (arrow [3]). Dollar and Kraay find generally that ‘growth is good for the poor’. But when average incomes rise, the average incomes of the poorest fifth of society rise proportionately. This holds across regions, periods, income levels, and growth rates.14 Openness [policy variable] Positive (1)

No effect (2) No effect (3)

Growth

Inequality

Negative (4)

Poverty Figure 1: The Mainstream Panglossian View that Openness is Sufficient for Poverty Reduction. Source: Milanovic (2002)

In a less simplistic view it is found that openness is not a choice variable and its effects on poverty reductions are ambiguous and complex (see Figure 2). Many authors, such as Rodríguez and 13 14

Greenaway / Morgan / Wright (1998). Dollar / Kraay (2001), Kuznets (1995).

238

Fritz Breuss

Rodrik, O’Rourke, and Vamvakidis find that the link between openness and growth is not always positive (arrow [1] in Figure 2).15 Openness, however, may have a more indirect positive impact on growth via technology spill-overs. Coe and Helpman for the industrial countries and Coe, Helpman and Hoffmaister for the „NorthSouth trade’ in particular – in line with the endogenous growth theory – study the effects of R&D spillovers via imports on TFP growth in the importing countries.16 Hence, more openness may potentially lead to more growth. A positive link between growth and openness is postulated by Bairoch in a world historic perspective. He sets the record straight on twenty commonly held myths about economic history. Among these are that free trade and population growth have historically led to periods of economic growth; that a move away from free trade caused the Great Depression; and that colonial powers in the nineteenth and early twentieth centuries became rich through the exploitation of the Third World. Bairoch argues that these beliefs are based on insufficient knowledge and misguided interpretations of the economic history of the United States, Europe, and the Third World.17 This position is comparable to that of Chang. By historically analysing the reasons why the poor countries remain poor he reaches the conclusion that the nowadays-rich industrial countries (like the USA, England, Germany, and Japan) and the newly industrializing countries (NICs; like Korea, Taiwan) followed a different strategy than they suggest to the developing countries. They did not start their development by liberalizing trade. In contrast they started with protecting their markets and made industrial and export-led (subsidy) policy. When becoming successful they simply ‘kicked away the ladder’. A similar strategy was also followed by the rapid growing China.18 The relationship between openness and inequality may be positive, neutral or even negative (arrow [2]). Representatives of studies with such ambiguous results are Milanovic, Ravallion, and Barro.19 The link between growth and inequality is unclear. Many 15 16 17 18 19

Rodríguez / Rodrik (2001), O’Rourke (2000); Vamvakidis (1998). Coe / Helpman (1995); Coe / Helpman / Hoffmaister (1997). Bairoch (1993). Chang (2002). Milanovic (2003), Ravallion (2003), Barro (1999).

Does the ‘Development Round’ Foster Development?

239

authors find a negative relationship, whereby the causality runs from inequality to growth (arrow [3]). More (less) inequality is bad (good) for growth.20 For a recent survey of the literature on inequality and growth, see Leoni and Pollan.21 It is often claimed – not only by globalization critiques – that growth may have a negative impact on poverty, meaning more growth is connected with greater poorness (arrow [4]). Even if trade liberalization results in aggregate welfare gains over all households, it is possible that the poorest households could lose. Harrison, Rutherford and Tarr demonstrate with a computable general equilibrium (CGE) model two approaches to designing trade liberalization in Turkey which ensure that the poor will not lose. The first approach uses direct compensation to losers. The second approach uses limited policy reform, where exceptions to the across-the board reform are chosen to meet the equity goal. In each case, the authors map out some of the efficiency costs of attaining these equity goals so as to inform policy makers about the least costly way of attaining them.22 Anderson et al. evaluate the fear that China’s accession to WTO will impoverish its farmers via greater import competition in its agricultural markets. Results of simulations with the GTAP CGE world model suggest that farm/no-farm income inequality may well rise within China but rural-urban income inequality need not.23

20 21 22 23

See Alesina / Rodrik (1994); Persson / Tabellini (1994); Perotti (1993); Perotti (1996). Leoni / Pollan (2003). Harrison / Rutherford / Tarr (2003). Anderson et al. (2004).

240

Fritz Breuss

Openness [not fully a choice variable] Not always positive (1) Positive (5)

Negative, positive or neutral (2)

Unclear

Growth

Inequality Negative (3)

Negative (4)

Poverty

Figure 2: A Less Simplistic View: Openness is not a Choice Variable and its Effects on Poverty Reduction are Ambiguous and Complex Source: Milanovic (2002)

2. What Determines the Huge Income Differences in the World? In a comprehensive study, Rodrik, Subramanian and Trebbi as well as Rodrik estimate the respective contributions of institutions, geography, and trade in determining income levels around the world. Their results indicate that the quality of institutions ‘trumps’ everything else. Once institutions are controlled for, measures of geography have at best weak direct effects on incomes; also trade is almost always insignificant, and often enters the income equations with the ‘wrong’ (i.e., negative) sign.24 In the voluminous literature on the determinants of the huge income differences in the world (average income levels in the world’s richest and poorest nations differ by a factor of more than 100), three strands of thoughts stand out:25 1. Geography: Geography is a key determinant of climate, endowment of natural resources, disease burden, transport costs, and diffusion of knowledge and technology from more advanced areas. Geography has a direct effect on incomes, through its effect on agricultural productivity and morbidity. 24 25

Rodrik / Subramanian / Trebbi (2002); Rodrik (2003a); Rodrik (2003b). See Rodrik / Subramanian / Trebbi (2002), 1-5.

Does the ‘Development Round’ Foster Development?

2.

26 27

241

This is shown with arrow (1) in Figure 3. It may also have an indirect effect through its impact on distance from markets and the extent of integration (arrow [2]) or its impact on the quality of domestic institutions (arrow [3]). As geography is as exogenous a determinant as an economist can ever hope to get, it is easiest to identify the causality. Representatives of this school are Diamond; Gallup, Sachs and Mellinger, and Sachs.26 Integration view: Another camp emphasizes the role of international trade as a driver of productivity change. This view (also identified as trade fundamentalists) gives market integration, and impediments thereof, a starring role in fostering economic convergence between rich and poor regions of the world. This school and also the third school of institutionalists have it harder, since they have to demonstrate the causality of their preferred determinants, as well as identify the effective channels through which it works (see Figure 3). For the trade fundamentalists the task consists of showing that arrows (4) and (5) – capturing the direct impact of integration on income and the indirect impact through institutions, respectively – are the relevant ones, while arrows (6) and (7) – reverse feedbacks from incomes and institutions, respectively – are of minor importance. Research in this area includes Frankel and Romer, and Sachs and Warner.27

Diamond (1997); Gallup / Sachs / Mellinger (1998); Sachs (2001). Frankel / Romer (1999); Sachs / Warner (1995).

242

Fritz Breuss Income level growth

endogenous

(4)

(7)

(8) (5)

Integration openness

(1)

(9)

Institutions rule of law

(6)

(2) (3) exogenous

Geography climate, resources

Figure 3: What accounts for the income difference between the richest and the poorest nations? Source: Rodrik / Subramanian / Trebbi (2002), 24.

Institutionalists: This group of explanations centres on institutions, and in particular the role of property rights and the rule of law. According to them, what matters are the rules of the game in a society and their conductiveness to desirable economic behaviour. They have to worry about different kinds of reverse causality. They have to show that improvements in property rights, the rule of law and other aspects of the institutional environment are an independent determinant of incomes (arrow [8] in Figure 3), and are not simply the consequences of higher incomes (arrow [9]) or of greater integration (arrow [5]). This view is associated primarily with North. It has received careful econometric treatment by Hall and Jones, who focus on what they call ‘social infrastructure’, and by Acemoglu, Johnson and Robinson, who focus on the expropriation risk that current and potential investors face.28 Growth theory has traditionally focused on physical and human capital accumulation, and, in its endogenous growth variant, on technological change. But long-term economic development is a very complex phenomenon, which cannot easily be explained by any one of the above determinants. Therefore Rodrik, in line with

3.

28

North (1990); Hall / Jones (1999); Acemoglu / Johnson / Robinson (2001).

Does the ‘Development Round’ Foster Development?

243

historians and many social scientists prefers nuanced, layered explanations of centuries of economic history, where the factors of the three schools interact with human choices and many other not-sosimple twists and turns of fate.29 As is indicated in Figure 3, the extent to which an economy is integrated with the rest of the world and the quality of its institutions are both endogenous, shaped potentially not just by each other and by geography, but also by income levels. Problems of endogeneity and reverse causality plague any empirical researcher trying to make sense of the relationships among these factors. Rodrik, Subramanian and Trebbi estimate the following equa30 tion: log yi P  DINS i  EINTi  JGEOi  H i (1) where yi is income per capita (in Purchasing-Power-Parity in US$, PPP GDP, in 1995) in country i, INSi, INTi, and GEOi are respectively measures for institutions,31 integration (measured by openness = ratio of nominal imports plus exports to GDP), and geography (distance from equator of capital city), and İi is the random error term. The authors are interested in the size, sign, and significance of the three coefficients D, E, and J. Three country samples are used, one with 64, one with 80, and one with 140 countries. First, there is a clear and unambiguously positive relationship between income and its possible three determinants. Also the OLS (one stage least-square) estimation of equation (1) results in the correct (positive) sings for all three indicators – institutions, openness, and geography. This suggests that countries with stronger institutions, more open economies, and more distant from the equator (measure of geography) are likely to have higher levels of income. However, if one takes into account the possibility of reverse causality, omitted variables bias, and measurement errors, equation (1) cannot be interpreted as causal or accurate. To address these problems, the authors employ a two-stage least squares estimation procedure, using instruments to determine the variables INSi and INTi, respectively. As a result, once the institutional variable is added, geography and openness do not have any additional power 29 30 31

Rodrik (2003b). Rodrik / Subramanian / Trebbi (2002), 6. Institutional quality measure by Kaufman / Kraay / Zoido-Lobaton (2002).

244

Fritz Breuss

in explaining growth and hence development. Institutions trump geography and openness. The importance of institution building for development is also stressed by the World Bank.32 B. Is WTO Membership Good for Trade at All? In provocative papers, Rose asserts that WTO membership does not stimulate trade. He estimates the effect on international trade of multilateral trade agreements: the World Trade Organization (WTO), its predecessor the General Agreement on Tariffs and Trade (GATT), and the Generalized System of Preferences (GSP) extended from rich countries to developing countries. He uses a standard ‘gravity’ model of bilateral merchandise trade and a large panel data set covering over fifty years and 175 countries. An extensive search reveals little evidence that countries joining or belonging to the GATT/WTO have very different trade patterns than outsiders. The GSP does seem to have a strong effect, and is associated with an approximate doubling of trade.33 To estimate the effects of international institutions on trade, Rose uses the following specification of the gravity model:34 ln( X ij ) t E 0  E1 ln Dij  E 2 ln(YiY j ) t   E 3 ln(YiY j / Popi Pop j ) t  E 4 Lang ij  E 5 ( FTAij ) t 

 E 6 (CU ij ) t  E 7 (GSPij ) t  E 8 (WTOij ) t   E 9 ( IMFij ) t  E10 (OECDij ) t  Dummies  H ijt

(2)

where i and j and denote trading partners, t denotes time, and the variables are defined as follows: Xij denotes the average value of real bilateral trade between countries i and j; Y is real GDP; Pop is population; D is the distance between i and j; Lang is a binary ‘dummy’ variable which is unity if i and j have a common language and zero otherwise; FTA is a binary variable which is unity if i and j both belong to the same regional trade agreement (e.g. EU countries); CU is a binary variable which is unity if i and j use the same currency (‘currency union’) at t; GSP is a binary variable which is unity if i extended a GSP concession to j at t or vice versa; WTO is a binary variable which is unity if i and j are GATT/WTO members at t; IMF is a binary variable which is unity if i and j are IMF members at t; OECD is a binary variable which is unity if i and j are 32 33 34

World Bank (2002), chapter 3. Rose (2003); Rose (2004). Rose (2003), 5-6.

Does the ‘Development Round’ Foster Development?

245

OECD members at t; Dummies represent all the additional variables used in the gravity model, such as sharing a land border, landlocked countries, islands, area of countries in square kilometres, colony countries, colonizer countries, fixed effects etc.; İij represents the omitted other influences on bilateral trade. Rose finds the following benchmark results (OLS estimation; using other estimation techniques the key results remain quite robust): Distance (in the geographic, linguistic, monetary, and historical senses) reduces trade, while greater economic ‘mass’ (real GDP and/or GDP per capita) expands it. The effects are economically and statistically significant.35 The coefficients of interest concern the effects of membership in international organizations; what do they reveal? There are two surprises; one negative and one positive. The negative surprise is that membership in neither the GATT/WTO nor the IMF is associated with deeper trade. Indeed, the point estimates for all four coefficients (both or one of the countries being in the GATT/WTO or IMF) are negative. The other surprise is the effect of OECD membership on trade that appears to be strong and positive. The estimations suggest that trade between one OECD member and another is 55% higher (between one OECD member and a non-member is 49% higher). Belonging to a regional free trade arrangement (NAFTA, EFTA, EU etc.) leads to trade creation by a considerable amount. According to Rose’s estimates bilateral trade between FTA members may rise by around 200%.36 In contrast to GATT/WTO membership, the extension of the GSP from one country to another (which primarily concerns North-South trade) seems to have a large positive effect on trade. Trade may be raised to around hundred percent.37 Belonging to a currency union raises trade by around 200%.38 This is also good news for the EMU of the European Union. In concluding, Rose finds his results puzzling in many respects. If WTO membership is irrelevant, why is it so attractive for many (also developing) countries to become members of WTO? Why should one care whether China is in the WTO? Anyhow, membership seems to be a big deal. Perhaps the GATT and WTO 35 36 37 38

Rose (2004), 104-105. 222% in Rose (2003), and 232% in Rose (2004). 93% in Rose (2003), and 136% in Rose (2004). 194% in Rose (2003) and 206% in Rose (2004).

246

Fritz Breuss

have large effects on income or welfare but only through mechanisms other than trade. Perhaps, he conjectures, the GATT and WTO have acted as an international public good, freeing trade for all countries independent of whether they are members or not.39 However, one cannot test this hypothesis, since there is no data for the counter-factual GATT-free world. It is an open question whether the liberalization of world trade after World-War II would have happened without GATT. Anyway, estimates by Baier and Bergstrand indicate that post-war growth of world trade, although primarily stimulated by income growth, was nevertheless propelled by tariff reductions.40 Similar results were found by Badinger and Breuss in a dynamic panel data approach to estimate the relative contributions of income growth, income convergence, and the reductions in tariffs and trade costs to the growth of intra-EU trade over the period 1960 to 2000.41 The results suggest that income growth was the major force, accounting for approximately two third of total growth. Trade liberalization still had a sizeable effect, accounting de facto for the rest of growth, while income convergence played only a minor role. Reductions in trade costs had no significant effect on the growth of intra-EU trade. Of course this leaves open the question whether this was due to the institution of GATT. Rose concludes by speculating that perhaps the GATT has not had much of an effect on trade, but the WTO will. Members of WTO use a more wide-reaching permanent framework to resolve disputes about trade in goods, services, and intellectual property.42 III. After Eight GATT Rounds – Markets Remain Still Subtly Protected: Market Access Issues

A. Patterns of Protection Even after eight GATT rounds of trade liberalization improving market access is still an ‘unfinished business’.43 Despite low average level of protection, agriculture and labour intensive industries carry a much higher level of protection than the average. 39 40 41 42 43

Rose (2004), 112. Baier / Bergstrand (2001). Badinger / Breuss (2004). Rose (2004), 112. See WTO (2002).

Does the ‘Development Round’ Foster Development?

247

The IMF and the World Bank notice in this context that averages of most-favoured-nation (MFN) applied tariffs by importing country or region provide an incomplete picture of protection.44 First, a number of barriers are not covered by the standard MFN databases, including specific tariffs (that is an absolute monetary value per unit of imports), tariff rate quotas (TRQs), prohibitions, contingent protection (refers to import barriers which, rather than being permanent, are introduced on a temporary and often selective basis in response to certain events – import surges, alleged unfair trading practices), the costs of rules of origin, and environmental and technical standards. Second, the averages do not capture the impact of tariff dispersion, in particular tariff peaks and escalation (international tariff peaks are defined as tariffs of 15 percent or higher; escalation refers to tariffs rising with the degree of processing of imports, and the resultant high levels of effective protection). Third, because of preference schemes (GSP) and differing export structures, the barriers faced by exporters to the same market can vary widely. Finally, uncertainty about market access, related to contingent protection, interpretation of norms and procedures, and the discretionary nature of many preference schemes, may represent a further disincentive to exporters. In the context of the DDA, WTO members are committed to negotiations aimed at substantially improving market access for agricultural and industrial products,45 in particular for developing countries in general and for the LDCs.46 There are a lot of subtle barriers to market access for developing countries. There remain ‘pockets of protection’ in products of particular interest to developing countries.47 The patterns of protection the development countries are confronted with are characterized by the following features:48 44 45

46

47 48

IMF / The World Bank (2002), 11. IMF / The World Bank (2002) studies in detail the problems of market access for developing countries in the sectors agriculture and textiles and clothing. See Doha Ministerial Declaration, paras. 13 and 16; WTO (2001). ‘Least-Developed Countries’ means 40 countries designated as such by the United Nations; 30 of which are WTO members (see http:// www.wto.org/english/thewto_e/whatis_e/tif_e/org7_e.htm). IMF / The World Bank (2002), 5. Ibidem, 10 et seqq.

248

Fritz Breuss

B. Divergence of Applied Tariffs and their Bindings Industrial countries have generally set applied tariff rates close to their tariff bindings (legally committed maximum tariff rates), enhancing the predictability and transparency of market access regimes (see Table 1). In contrast, most developing countries bind their tariffs at levels well above their applied rates so that they could in principle substantially increase their applied tariffs without infringing their WTO commitments. Applied tariff rates in 2001 varied considerably across country groupings.49 Sub-Saharan African countries continue to have the highest simple average tariff protection (17.2 percent), followed by the Middle East and North Africa (16.8 percent). Among broad country groupings, it is notable that the average tariff of the LDCs (17.9 percent) is higher than that of other developing countries (14.0 percent) and well above that of industrial countries (5.2 percent). Table 1: Bound and Applied Tariffs on Industrial Products (Simple Averages) Import markets

End of Share of Simple Simple imple- bound MFN average mentation tariffs average applied period bound

NORTH AMERICA Canada 2000 United States 2000 LATIN AMERICA Argentina 2005 Chile 2005 Colombia 2005 Costa Rica 2005 Mexico 2005 Peru 2005 WESTERN EUROPE EC 2000 Norway 2000 Turkey 2000 EASTERN EUROPE Czech Republic 2000 Hungary 2000 Romania 2000

49

99.6 100.0

5.2 3.9

100.0 100.0 100.0 100.0 100.0 100.0

31.0 25.0 35.5 44.6 34.8 30.0

100.0 100.0 36.3 100.0 95.4 100.0

Year ǻ bound and applied tariffs

4.8 1998 4.3 1999 13.7 10.9 11.2 6.4 12.6 13.0

0.4 -0.3

1998 1997 1998 1998 1998 1998

17.3 14.1 24.3 38.2 22.2 17.0

4.1 3.4 42.6

5.0 1998 3.3 1998 7.5 1996

-0.9 0.1 35.1

4.3 7.4 30.1

4.8 1998 9.0

-0.5 -1.6

See Table 1 and IMF / The World Bank (2002), 11.

Does the ‘Development Round’ Foster Development? Import markets

End of Share of Simple Simple imple- bound MFN average mentation tariffs average applied period bound

249

Year ǻ bound and applied tariffs

Slovakia 2000 100.0 4.3 4.9 1998 ASIA Australia 2000 95.9 14.2 5.8 1998 Hong Kong (China) 2005 23.5 0.0 0.0 1998 India 2005 61.6 58.7 Japan 2000 99.2 3.5 4.2 1998 Republic of Korea 2005 90.4 11.7 7.9 1998 Macao (China) 2005 9.9 0.0 0.0 Philippines 2005 58.6 26.1 9.5 1998 Singapore 2005 65.5 4.6 0.0 AFRICA Cameroon 2005 0.1 17.6 17.6 1999 Chad 2005 0.4 17.6 17.6 1999 Gabon 2005 100.0 15.5 17.6 1999 Senegal 2005 32.3 13.8 South Africa 2005 98.1 17.7 Tunisia 2005 46.3 34.0 Zimbabwe 2005 8.8 11.3 Source: UNCTAD (2003), 11 based on Bacchetta / Bora (2001).

-0.6 8.4 0.0 -0.7 3.8 0.0 16.6 4.6 0 0 -2.1

C. Developing Countries Face Higher Barriers to their Exports than Industrial Countries There are large variations in market access conditions depending on the type of product and the particular exporter-importer combination.50 Table 2 presents combined ad valorem tariff equivalents (AVEs) of a range of protective measures, while taking into account preferences and export structures.51 The results suggest that the EU 50 51

Ibidem, 13. The Market Access Maps database has been developed by the International Trade Centre UNCTAD/WTO (ITC), Geneva (http://www.intracen.org/home.htm), and offers broader coverage of restrictions and preferences schemes than other sources. It incorporates the market access regimes of 137 countries, including preferential regimes, antidumping measures, and ad valorem equivalents of specific duties and tariff rate quotas (the current release does not yet incorporate recent preferential agreements, such as the EU’s ‘Evertything-but-Arms’ (EBA) initiative and the United States’ African Growth and Opportunity Act (AGOA), which would further reduce

250

Fritz Breuss

protection is heavily skewed against imports from middle-income developing countries, as the US protection is against imports from LDCs. The geographical patterns of Canadian and Japanese protection are less marked, although the former’s protection pattern appears tilted against LDCs and the latter against other low-income countries. Levels of protection in other OECD markets, and in middle-income developing countries as a group tend to be well above those in the Quad (Canada, the EU, Japan and the United States). Given the potential for trade among the developing countries, now at 40 percent of their total exports, barriers to this trade are increasingly significant. The AVEs in the North-North (intra-OECD) trade are around 1.5 to 2.5 percent for manufactures and 14.5 to 41.5 percent for agriculture (see Table 2). The AVEs in the SouthSouth (intra developing countries) trade are 6.5 percent for manufactures and 17 percent for agriculture. Generally therefore, the impediments to trade in agricultural products remain far greater than in manufacturing trade. In the context of the Uruguay Round, quantitative restrictions and other nontariff measures (NTMs) were converted into tariffs. While improving transparency, the modalities of conversion have in many cases allowed an increase in effective protection. Specific tariffs and tariff-rate quotas, which are most frequent in agricultural trade, account for a significant share of the AVEs.52 The average rate of duty on agricultural imports into Quad markets from LDCs, excluding specific tariffs and effect of tariffrate quotas, is 1.7 percent.53 Ad valorem tariff equivalents (not covering domestic measures of support or the effect of export subsidies) of middle-income developing countries are broadly comparable with those of the Quad (see Table 2).

52 53

applied tariffs on imports from LDCs (see footnote i to Table 2). These data are combined with bilateral product-specific trade flows form the United Nations’ COMTRADE database. Information on tariff and other barriers refers to 2000, on trade flows of the most recent available year. For more information about this database and the methodology for calculating AVEs, see Bouët et al. (2001). IMF / The World Bank (2002), 13. See Bacchetta / Bora (2002).

Does the ‘Development Round’ Foster Development?

251

Table 2: Effective Ad-Valorem Tariff Equivalents on Bilateral Trade Flowsi)ii) LDCs Importers Canada EU Japan USA Other OECD Countries Developing Countries Middle Income Countries Canada EU Japan USA Other OECD Countries Developing Countries Middle Income Countries

6.7 2.8 4.9 13.6 8.7 8.1

3.4 7.6 29.1 28.1 19.6

Exporters Other Low- Middle- All Deve- OECD Income Income loping Countries Countries Countries Total Trade 5.4 4.4 4.4 7.0 10.3 7.2 6.4 4.5 4.7 6.2 3.6 4.5 13.1 10.4 10.2 -

-

7.5

-

11.9

12.7

-

-

18.7 13.4 16.3 9.5 28.0

18.2

Trade in Agriculture 16.3 17.5 24.8 20.0 21.2 21.9 13.0 12.7 35.4 32.5 -

18.4

23.1

17.0 -

Trade in Manufactures Canada 7.7 4.2 2.0 2.9 EU 0.0 5.7 5.5 4.5 Japan 0.1 5.0 1.4 2.5 USA 8.0 5.9 2.1 3.6 Other OECD 5.0 10.8 5.7 7.4 Countries Developing 6.4 Countries Middle Income 6.0 11.1 10.9 Countries Source: IMF / The World Bank (2002), 12, based on ITC.iii) i)

33.7 41.6 28.3 14.5 42.1 14.5 -

2.0 2.5 1.2 1.6 7.4 6.9 -

The information in the table does not yet reflect the EU’s EBA initiative and the United States’ AGOA. Taking account of the former

252

ii)

iii)

Fritz Breuss would reduce AVEs on EU agricultural imports from LDCs significantly, though not to zero (restrictions remain on sugar, rice and bananas for a transition period). AGOA would lower AVEs on both agricultural and manufactured imports into the United States for some African LDCs and low-income countries, but the extent of the reduction is hard to predict. The protection levels of importing countries in this table are weighted by the imports of the reference group this country belongs to, with the grouping criteria being GDP per capita. This is done to minimize the potential endogeneity bias of using national import weights (a high tariff can limit imports, and in the extreme could carry zero weight if its level is prohibitive). For the methodology in calculating AVEs, see Bouët et al. (2001).

D. Tariff Peaks and Tariff Escalation As far as the current pattern of protection is concerned, tariff barriers to exports from developing countries appear to be heavily concentrated in agriculture, textiles and clothing, and other sectors of export interest to developing countries. The post-Uruguay Round protection pattern is characterized by a high dispersion in tariff rates, with a large number of tariff peaks concerning products of interest to developing countries in agriculture, food, textiles, apparel and some mid-technology products. Tariff escalation also affects trade flows in a number of products of interest to developing countries. It is a pervasive feature in both developed and developing countries and concerns both agricultural and industrial goods.54 Even tariffs are sometimes applied as specific or mixed rates or tariff rate quotas, whose ad valorem or percentage equivalents can be difficult to estimate. Tariff duties are sometimes waived under a variety of national schemes. Non-tariff barriers (NTBs) are inherently complex, they have multiple effects and their incidence varies across time and trade partners. The UNCTAD reports that reductions of tariff rates have gone together with the reduced use of NTBs.55 Participation in the WTO has been a mixed experience for the developing countries, providing benefits and also challenges. On the one hand it means an improved and more secure access to third country markets. On the other hand it entails taking on an increas-

54 55

UNCTAD (2003), VIII. Ibidem, 13.

Does the ‘Development Round’ Foster Development?

253

ing level of obligations, including market opening and the application of WTO rules. The future challenge – in particular vis-à-vis the developing countries – is the mitigation of the problem of tariff peaks and tariff escalation. All important world-trade related international institutions identify the reduction (phasing out) of tariff peaks and tariff escalation as the most important issue to increase market access for the developing world.56 1. Tariff Peaks While negotiations on reducing trade barriers and support measures in agriculture are part of the ‘built-in-agenda’ established during the Uruguay Round, market access in industrial products was added to the negotiating agenda in Doha. WTO members, acknowledging the importance of enhanced market access for developing countries, started to negotiate on the reduction or elimination of tariff peaks, high tariffs and tariff escalation. It was pointed out that ‘tariff peaks’ and ‘high tariffs’ are not defined in the WTO.57 Following the practice of the OECD, tariff peaks may be defined as rates that are more than three times the national average.58 IMF and World Bank define international tariff peaks as tariffs of 15 percent or higher.59 It is widely agreed among trade economists that a relatively uniform or flat tariff structure is preferable to one exhibiting considerable dispersion. At least two reasons are advanced in favour of a flat tariff structure. First, the costs in terms of welfare and economic efficiency of a tariff regime increase as the degree of dispersion increases. Tariff peaks increase the economic inefficiency stemming from protection, as it hampers the exploitation of increasing returns to scale across different markets, while reducing competition and specialization according to comparative advantage. Second, political economy arguments support a flat tariff structure. Uniform tariff rates are more transparent and easier to administer than non-uniform tariffs, and are less likely to be determined

56

57 58 59

See European Commission (2002), 83; IMF / The World Bank, (2002), 5, 20; UNCTAD (2003), 16 et seqq.; WTO (2003), XIX, 45, 61. UNCTAD (2003), 18. OECD (1997). IMF / The World Bank (2002), 11.

254

Fritz Breuss

by the relative political power of domestic industries. Finding a formula to reduce tariff peaks is therefore highly desirable.60 The UNCTAD gives a comprehensive picture on the problems of tariff peaks in agriculture and manufactures. The incidence of international tariff peaks is calculated by comparing each tariff line with a 15 percent benchmark. This gives an indicator for international peaks. They are more frequent in developing (22.5%) than in developed countries (OECD, 7.3%). The international tariff peaks for manufactured exports from developing countries are highest in South Asia (55.12%) and Sub-Saharan Africa (31%) and Latin America (28.4%), followed by Asian NICs (19.7%) and North Africa and Middle East (10.8%).61 18

Canada

16

EU

14

Japan United States

12 10 8 6 4 2

Other

High technology, electronic/electgrical products

Medium technology, process industries

Low technology, textile/fashion cluster

Other resource-based manufactures

Primary products

0

Figure 4: Weighted MFN Tariffs Applied by Quad Countries on Exports from Developing Countries, 2000. Source: UNCTAD (2003), 26. 60 61

See UNCTAD (2003), 19. Ibidem, 19-25.

Does the ‘Development Round’ Foster Development?

255

To understand the extent by which the structure of world protection may hamper the possibility for developing countries to follow an export-driven shift from traditional commodities to highvalue added products one may look at market access opportunities offered by developed countries to developing countries in different technology-differentiated products. Figure 4 shows that, overall, protection in Quad markets is quite clearly concentrated in typical export categories of interest to low- and middle-income developing countries, such as textiles and agriculture. Developing countries which are mainly specialised in raw materials and primary agricultural products are facing higher trade barriers when trying to move into the subsequent production stages (low technology sectors such as processed agricultural products and textiles, or medium technologies such as automotives). According the World Trade report of the WTO the average applied tariff across 23 categories used during the Uruguay Round shows that the average tariff in the agricultural categories is higher than that in most of the industrial categories (see Figure 5). The highest rates are applied to animals, beverages and spirits, dairy products and tobacco. In general the pattern of protection is lower on lower value-added products such as cut flowers, fruits and vegetables, coffee and tea.62 According to the report the incidence of high tariffs in agricultural products poses a particular challenge to negotiations. Furthermore, some developed countries have insulated sensitive sectors from international trade reform: the United States (peanuts), Canada (dairy and poultry), Japan (rice) and the Republic of Korea (rice).63

62 63

WTO (2003). Ibidem, 128.

256

Fritz Breuss

Wo o d , p ulp , p ap er Text iles and clo thing Leather and fo o twear

Non-agricultural products

M et als Chemicals Trans p o rt No n-elect rical machinery Elect rical machinery M inerals (includ ing o il) Other no n-ag ricultural p ro d . Fis h and fis h p ro d uct s Fruit and veg etab les Co ffee and t ea Sug ar

Agricultural products

Sp ices , cereal and o ther fo o d p rep aratio ns Grains Animals and p ro d uct s thereo f Oil s eed s , fat s and o ils Cut flo wers , p lants , et c. Beverag es and s p irit s Dairy p ro d uct s To b acco Other ag ricultrual p ro d .

0

5

10

15

20

25

30

Figure 5: Average MFN Applied Rates by Product Category Source: WTO (2003), 128.

According to the IMF and the World Bank between 6 and 14 percent of Quad tariff lines are subject to tariff peaks, in some cases at rates well over 100 percent.64 Tariff peaks are also a prominent feature of tariff regimes in developing countries. Most preference schemes, moreover, offer little relief from tariff peaks.65 In Canada and the United States, tariff peaks are concentrated in textiles and clothing, and in the cases of the EU and Japan in agriculture, food products and footwear. Notably, estimates suggest that the capping 64 65

IMF / The World Bank (2002), 13 et seq. See Hoekman et al. (2001).

Does the ‘Development Round’ Foster Development?

257

of all peaks at the threshold of 15 percent would reduce AVEs in textiles and clothing by around 20 percent for imports from most source countries into the United States, and by 59 percent for imports from China. In agriculture and food products, they would decline by 40-60 percent on imports into the EU.66 According to simulations by Hoekman et al. full duty and quota free access for LDCs in the Quad for tariff peak products would result in a 11 percent increase in their total exports – on the order of $ 2.5 billion. Exports to Quad countries of tariff peak products would expand by 30 to 60 percent.67 Given that LDC exports on tariff peak items account for only a small share of total developing country exports, granting LDCs duty free access would have a negligible impact on other developing countries. For the same reason, Quad imports increase only marginally, suggesting that this should not be a factor constraining implementation of duty free access for the poorest countries. Most developing countries enjoy preferential access to Quad markets, either through unilateral schemes such as the GSP,68 or through free trade agreements such as NAFTA or EU Association Agreements. In the case of Canada, Japan and the EU, around 170 developing countries benefit from GSP (or better) preferences. In the case of the US, 29 developing countries are excluded from GSP, so that only 140 developing countries benefit from some sort of preferential access. Preferences granted by the Quad are of a cascading nature.69 Countries with FTAs generally get the best treatment, followed by LDCs and other developing countries (see Table 3). The US grants preferences to the members of the Andean Pact (ATP) and the Caribbean countries, and to Mexico under NAFTA. For the EU, in Table 3 both Lomé preferences (ACP), and the FTA preferences granted to Eastern Europe (Europe Agreements) and Mediterranean countries are reported. In the case of the EU three different groups of countries are constructed: Non-ACP LDCs; ACP countries (broken down into LDCs and non-LDCs); and nonACP, non-LDC developing countries that benefit from GSP treat66 67 68

69

See Bouët et al. (2001). Hoekman et al. (2001). The EU was the first customs territory to grant GSP preferences to developing countries in 1971. See Kennan / Stevens (1997) for a detailed description of the European GSP. See Hoekman et al. (2001), 13.

258

Fritz Breuss

ment. Finally, in the case of Canada, developing countries are grouped into those benefiting from LDC, GSP, or Caribbean preferences, and Mexico and Chile that benefit from FTAs. On average these preferential schemes are quite generous. In the EU, the average tariff (for all goods) faced by LDCs or ACP members is below 1 percent, compared to the 7.4 percent average MNF tariff. GSP preferences in the EU are close to 50 percent (see Table 3; last column). In the United States LDC and GSP preferences offer more than a 50 percent average margin – LDC preferences being more generous around 65 percent. Japan offers a 48 percent preference margin under their GSP regime, and an average 60 percent preference for LDCs. Canada gives a 25 percent preference to GSP countries and 45 percent to LDCs (see Table 3). Preferences are much less generous for tariff peak products (see Table 3, third column). Except for the EU, the preference margins are significantly below the average across all products.70 Table 3: Tariff Peaks and Preferential Duty Rates in the Quad, 1999 Preferential Trade Agreements/GSP

US Canada Mexico Israel ANDEAN /a Caribbean Community /b GSP-only beneficiaries /c LDCs /d Others (MFN Rate) EU Eastern Europe and Middle East /e GSP-only beneficiaries /f LDCs (ACP) /g Other ACP Countries /h Other LDCs /i Others (MFN Rate) /j

70

Number of Countries

1 1 1 4 22 80 38 15 30 42 37 32 11

Average Preference Rate (unweighted in %) Tariff Peak All Goods at Products HS-6 0.6 1.6 0.6 14.0 13.5 16.0 14.4 (20.8)

0.1 0.3 0.1 1.7 1.6 2.4 1.8 (5.0)

20.1

1.8

19.8 11.9 12.4 12.6 (40.3)

3.6 0.8 0.9 0.9 (7.4)

For a very detailed analysis by products, see Hoekman et al. (2001).

Does the ‘Development Round’ Foster Development? Preferential Trade Agreements/GSP

Number of Countries

259

Average Preference Rate (unweighted in %) Tariff Peak All Goods at Products HS-6

Japan GSP-only beneficiaries /k 127 22.7 2.3 LDCs /l 42 19.0 1.7 Others (MFN Rate) (27.8) (4.3) Canada United States 1 7.1 1.6 Australia 1 28.2 7.8 New Zealand 1 28.2 7.8 Mexico 1 15.9 3.1 Chile 1 12.2 2.4 Israel 1 11.8 2.5 Caribbean countries /m 18 23.3 4.3 GSP-only beneficiaries /n 108 28.2 6.2 LDCs /o 47 22.8 4.4 Others (MFN Rate) (30.5) (8.3) Notes: /a Bolivia, Colombia, Ecuador, Peru under Andean Trade Preference Act. /b Based on 20 Caribbean countries under Caribbean Basin Economic Recovery Act and Bahamas, Nicaragua. /c Included 80 developing countries or territories under GSP scheme but excluding 29 other developing economies. /d Based on UN 48 LDCs but excluding 10 countries. /e Including countries with reciprocal and non reciprocal trade agreements with the EU. /f Most developing countries in Latin America and Asia; excludes Hong Kong, Korea and Singapore (non-GSP nations). /g Included 37 ACP and LDCS under Lomé Convention. /h Included ACP 32 countries not under the group of LDCs. /I Included 11 LDCs but not under ACP countries. /j Included all industrial countries, Hong Kong, Korea, Singapore and 14 transition countries. /k 127 countries; excludes Albania, Bosnia, Estonia, Latvia, Lebanon, Lithuania, Macedonia, Moldova, Vietnam, Yugoslavia. /l Excludes 3 LDCs: Comoros, Djibouti and Tuvalu. 3 others (Congo DR, Kiribati and Zambia) are included in the GSP group. /m Included 18 Caribbean countries or territories under CCCT. /n Excluded 8 developing countries: Albania, Aruba, Bosnia & Her, Macedonia, Mongolia, Oman, Saudi Arabia, Yugoslavia. /o Excluded Myanmar. Source: Hoekman / Ng / Olarreaga (2001), 14.

260

Fritz Breuss

2. Tariff Escalation The practice of tariff escalation biases exports towards unprocessed resource-based commodities, characterized by low value-added. This may cause difficulties to commodity-dependent developing countries in their attempt to diversify their export base. The pattern of protection creates particular hurdles for countries taking the first steps up the technology ladder. According to the IMF and the World Bank protection is relatively low for primary products, but increases sharply for low-technology, labor-intensive food processing and light industries, declines somewhat in the medium-technology range – such as automotive products – and is lowest at the upper end of the technology spectrum (see also Figure 4).71 In a snapshot of the post-Uruguay Round tariff levels by product and by processing stage in the Quad markets the UNCTAD study shows the following picture: First, with few exceptions, post-Uruguay Round tariffs escalate (that means increase between these categories) not only between raw and semi-finished but also between semi-finished and finished goods.72 On average, the escalation in Canada and Japan and the EU is higher between raw and finished, while in the United States the highest average escalation is found between semi-finished and finished goods. Tariffs tend to escalate not only in agriculture but also in manufacturing. The average post-Uruguay Round tariff for all industrial products ranges from 0.8 percent on raw materials to 4.8 percent on the finished product, resulting in an average tariff level of around 3 percent. A more detailed analysis of tariff escalation, distinguishing between markets of developing countries shows that tariff escalation is not just a feature of developed markets but is present in fact (sometimes even more prominently) in developing countries as well.73 As in the case of Quad countries, in most cases escalation in developing countries is greatest between raw and finished products. However, as in the case of the United States, in Asian NICs, there is de-escalation between raw and semi-finished products, and the highest escalation is found between semi-finished and finished products. Of course the pattern of escalation is different in the 71 72 73

IMF / The World Bank (2002), 14. UNCTAD (2003), 27. Ibidem, 28.

Does the ‘Development Round’ Foster Development?

261

world regions analysed by the UNCTAD: Asian NICs, South Asia, North America, Sub-Saharan Africa, Oceania, North Africa and Middle East, Latin America.74 In summary, the evidence shows that tariff escalation is a quite widespread phenomenon that affect both agricultural and industrial products, and is present in markets of both developed and developing countries (for a compact representation, see Figure 6). 30

Agricultural Products

Industrial Products

25

Tariff rate (%)

20

15

10

5

0 Developing

High income First stage

Semi processed

Developing

High income

Fully processed

Figure 6: Tariff Escalation Source: European Commission (2002), 83

At the May 2001 3rd UN Conference on LDCs, all industrialised countries for the first time committed to the objective of duty and quota free access for all exports originating in LDCs. Such an emulation of the unilateral EU ‘Everything But Arms’ (EBA) initiative (opening markets to all LDC exports, by allowing duty and quota free access) launched in early 2001 by other major industrialised nations, would contribute significantly to LDCs’ opportunities for trade based growth. The EU was the major destination for

74

Ibidem.

262

Fritz Breuss

LDCs’ exports even before the EBA initiative was adopted.75 In 1998, the EU accounted for 56 percent of their total exports. Moreover, the EU already had very low tariffs for LDC imports. Exports of the main products liberalised by EBA were, however, very low. Eliminating protection will certainly enhance trade in the products concerned. Studies by the UNCTAD and Ianchovichina et al. on the impact of EBA have forecasted large increases in welfare as a result – between US$ 400 and US$ 317 million depending on the study.76 However, even prior to the adoption of EBA, levels of protection against LDC exports were far higher in other Quad countries than in the EU (see Figure 7). Thus if all Quad members were to adopt similar measures, the welfare impact would be much higher (US$ 1.8 bn to US$ 2.5 bn).77 60

50

% of LDC exports

40

30

20

10

0 Canada

European Union Exports facing tariffs > 0%

Japan

United States

Exports facing tariffs > 5%

Figure 7: Pattern of Protection Facing LDC Exports (pre-EBA) Source: European Commission (2002), 75.

75 76 77

Council Regulation (EC) No 416/2001 of February 2001. UNCTAD (2001); Ianchovichina et al. (2001). See European Commission (2002), 84.

Does the ‘Development Round’ Foster Development?

263

Table 4: Liberalization in Agriculture: The Role of Tariff Escalation Results of Model Simulations of a 50 % Worldwide Cut in Tariffs on Processed Agricultural Products Welfare effects Aggregate trade data Percentage Total value (‘97 Exports Terms of trade change US$ mill.) Percentage change 0.101 994.9 0.578 0.037 0.040 475.4 0.697 -0.059 0.047 230.7 1.215 -0.243 0.022 1,613.2 0.340 0.038 0.018 1,415.7 0.403 0.080 0.098 750.0 1.150 -0.039

Regions Asian NICs China South Asia Western Europe North America Transition economies Sub-Saharan 0.049 153.0 Africa Oceania 0.232 951.4 North Africa 0.260 2,036.4 and Middle East Latin America 0.057 1,013.8 Japan 0.058 2,127.0 Rest of the World 0.096 242.1 Total 12,003.4 Source: UNCTAD (2003), 45 et seq.

1.324

-0.220

1.425 1.706

1.003 -0.408

1.042 1.196 1.843

0.042 -0.255 0.183

The effects of the elimination of tariff escalation in agriculture were evaluated by the UNCTAD with the help of CGE model simulations.78 The model used in the simulations was the standard static GTAP5 model, with perfect competition in all sectors and constant returns to scale.79 The database was GTAP5 (1997 data), modified by the UNCTAD to account for tariff preferences (related to GSP, no-reciprocal agreements as the Lomé-Cotonou agreement, and regional trade agreements) available from the UNCTAD TRAINS database.80 The model was aggregated to cover 12 world

78 79 80

UNCTAD (2003), 45 et seq. See Dimarana / McDougall (2002). UNCTAD-TRAINS (Trade Analysis and Information System) is a comprehensive computerized information system at the HS-based tariff line level covering tariff, para-tariff measures as well as import flows by origin for more than 140 countries. See the UNCTADTRAINS website at http://r0.unctad.org/trains/. In country notes this homepage also informs on trade control measures adopted by se-

264

Fritz Breuss

regions (3 of which comprise developing countries: South Asia, Sub-Saharan Africa, North Africa and Middle East) and 6 sectors (natural resources, manufactures, primary agriculture, processed agricultural products, textiles and apparel, services).81 A 50 percent worldwide cut in tariffs on processed agricultural products would result in increase world welfare by about 12 US$ billion, roughly half those obtained from the liberalization of all agricultural sectors. The largest welfare gains are received in Oceania, North Africa and Middle East and in the Asian NICs (see Table 4). The elimination of tariff escalation in the agricultural sector improves export chances for developing regions such as South Asia, Sub-Saharan Africa as well as North Africa and Middle East (see Table 4). E. Contingent Protection 1. Antidumping Measures Among the trade remedies permitted under WTO rules, antidumping has become by far the most widely used, in both industrial and developing countries. Since 1995 over 1,800 antidumping investigations have been initiated.82 While industrial countries (511 cases) have traditionally been the main users of such measures, developing countries (1086 cases) have been more active in recent years, led by India, Argentina, Brazil, and South Africa. The transition countries initiated 248 cases since 1995. In the seven years to 2001, developing countries initiated almost two thirds of all investigations, well in excess of their share in world trade. However, developing countries have also been the target of nearly 60 percent of investigations, mostly initiated by other developing countries. According to the IMF and the World Bank the recent steep rise in antidumping investigations may put at risk the predictability and non-discriminatory application of trade policies.83 Recent enforcement practices have raised serious concerns about the influence of special interests on public policy, and may impose large costs on consumers and downstream industries in importing countries.

81 82 83

lected developing countries. HS is the harmonized system of tariff classification. See UNCTAD (2003), 50 et seq. See IMF / The World Bank (2002) 15. Ibidem, 15 et seq.

Does the ‘Development Round’ Foster Development?

265

Moreover, the deterrent effect of an investigation typically reaches well beyond the targeted exporter, and impedes incentives to pass on efficiency gains.84 Additionally the frequency of antidumping measures increases during, and may thus reinforce, economic downturns.85 Small firms and countries face greater uncertainty as they often lack the resources to challenge antidumping. The introduction of competition law principles and of public interest clauses, giving affected importers and users legal standing to argue against protection, could reduce the protectionist bias of antidumping.86 2. Standards and Non-tariff Barriers to Trade Many developing countries are concerned that they are ill-prepared to meet increasingly complex and burdensome standards and regulations. Such regulations play an important role in facilitating trade by ensuring quality, safety and technical compatibility. However, there is often a risk that such regulations may be captured by special interests, particularly when regulatory processes are not transparent. And because the industrial countries are leaders in such standards, conditions might be imposed that are tighter than needed to achieve their objectives and hence just serve as a new kind of protectionist measure. Technical barriers have become a key concern regarding market access. Annual notifications of new technical barriers (including health and safety standards, and product standards) to GATT/WTO increased steadily from a dozen or two in the early 1980s to over 400 in 1999. Low- and middle-income countries reported that over the period from 1996-1999 more than 50 percent of their potential exports of fresh and processed fish, meat, fruit and vegetables into the EU were ‘prevented’ by their inability to comply with SPS (Sanitary and Phytosanitary Measures) requirements.87 SPS and other technical requirements have been viewed by developing country trade officials as a greater constraint on their ability to exploit their comparative advantages and hence to export than tariffs and quantitative restrictions. According to Henson et al. surveying government officials in 65 low- and middle-income countries SPS requirements were con84 85 86 87

Finger (1993). Knetter / Prusa (2000). Hoekman / Mavroïdis (1996). See IMF / The World Bank (2002), 16.

266

Fritz Breuss

sidered the most significant impediment to exports to the EU.88 Other technical requirements (e.g. labeling regulations or compositional standards) were also considered significant impediments to trade. Other factors of impediments are transport and other direct export costs, tariffs and only at the last palce quantitative restrictions.89 Overall, the developing countries have found it difficult to participate in designing standards in ways that better reflect their concerns and capabilities, and to challenge them where they were imposed in a discriminatory manner. A number of agreements in the Uruguay Round have addressed these concerns by strengthening international rules governing product standards in order to minimize their abuse for protectionist purposes. Among others, these are the Agreement on Technical Barriers to Trade (TBT, for trade in manufactured products), and the Agreement on the Application of Sanitary and Phytosanitary Measures (SPS, relating to health and safety for humans and animals). Nevertheless, SPS and TBT adversely affect LDCs exports.90 F. Trade Preferences Most developing countries have preferential access to industrial country markets for a wide range of products. This departure from the traditional non-discrimination principle of the GATT has been sanctioned under the GSP. In 2001, some 15 such schemes were in effect, though country coverage and preference margins over applied MFN tariffs varied widely. In addition to GSP an important recent development has been the proliferation of bilateral and regional free trade agreements between industrial and developing countries. Such agreements have to cover substantially all trade, unlike GSP schemes. However, the drawbacks related to rules of origin apply to both measures. According to the IMF and the World Bank the benefits of many GSP schemes for their beneficiaries have been limited.91 The reason is that preference margins are smaller for products that the importing country deems to be sensitive – which are also among the most protected.92 Ozden and Reinhard found evidence that the 88 89 90 91 92

Henson et al. (2000). See also IMF / The World Bank (2002), 17. See Fontagné (2003), 5. IMF / The World Bank (2002), 18. Ibidem.

Does the ‘Development Round’ Foster Development?

267

availability of unreciprocated market access preferences has undermined the incentives of benefiting countries to engage in trade liberalization, thus at times perpetuating anti-export biases in their trade regimes.93 The cost of monitoring of rules of origin to avoid transshipment may have reduced the benefits expected from such schemes. Rules of origin are akin to local content requirements (value added thresholds). Costs arise both from exporters seeking to benefit from preferences by producing inputs from less efficient sources (trade diversion), and from the administration of, and accounting for ‘origin’.94 Brenton and Manchin have shown that as a result of unattractive rules of origin, only one-third of imports that were eligible for preferential treatment did in fact enter the EU market with reduces duties.95 This problem is particularly acute for textiles and clothing. Recently market access under GSP schemes has been enhanced on a regional basis, in particular for African countries. To date 36 Sub-Saharan African countries have qualified in principle for preferential access under the United States’ AGOA (African Growth and Opportunity Act), adopted in 2000 (signed into law on May 18, 2000).96 Margins of preference are substantial for textile and apparel products as well as for a range of other light manufactures and food products. In order to benefit from this scheme, countries have to meet, in addition to relatively tight rules of origin and standard GSP criteria, requirements relating to child labour and the protection of internationally recognized workers’ rights. The administrative requirement involved in documenting eligibility may explain why only 15 countries had availed themselves of benefits under this scheme in the year 2002, with most of the benefits accruing to four countries – Gabon, Lesotho, Nigeria, and South Africa and with fuel accounting for 85 percent of AGOA imports.97 Mattoo et al. estimate that by 2008 the volume of African exports to the US market may rise by an additional 6-7 percent.98 Presently Sub-Saharan 93 94 95 96 97 98

Ozden / Reinhardt (2002) IMF / The World Bank (2002), 18 et seq. give examples in case of NAFTA and the EU. Brenton / Manchin (2002) IMF and the World Bank (2002), 19. Ibidem; and USITC website at http://www.usitc.gov/. Mattoo et al. (2002).

268

Fritz Breuss

African exporters hold their US market share in textiles and clothing of 1.6 percent. In 2002 the AGOA imports increased by 10 percent to US$ 9 billion. Excluding petroleum imports, the AGOA imports were less than US$ 2.2 billion.99 However, effective preference margins will decline as quotas under the WTO Agreement on Textiles and Clothing are phased out. A number of industrialised countries have recently granted comprehensive tariff and quota free access to LDCs. The EU’s EBA initiative has extended such preferential access since coming into effect in March 2001. It covers all products, except for sugar, bananas, and rice, which are to be liberalised more gradually.100 Unlike the EU’s GSP scheme, benefits under the EBA are extended on an indefinite basis, subject however, to broad safeguards. Similar schemes providing for virtually unqualified duty- and quota-free access for LDCs have also been adopted by New Zealand, Norway, and Switzerland. Such broad-based tariff-free market access for LDCs can assist in diversifying their export structures.101 If such schemes are adopted by all Quad markets, LDC exports to the Quad might increase by US$ 2.5 billion, or about 11 percent, with relatively limited cost in terms of trade diversion.102 Table 5: Market Access for LDCs Duty-Free imports into developed countries from developing countries and LDCs. 1996-2001 (percent) 1996 1997 1998 1999 2000 2001 Excluding arms Developing 54.8 50.5 49.9 57.2 62.8 65.7 countries LDCs 71.5 67.2 77.7 77.1 75.4 75.3 Excluding arms and oil Developing 56.8 51.5 49.9 58.1 65.1 66.0 countries LDCs 81.1 75.5 75.0 73.6 70.5 69.1 Source: WTO (2003), 126.

In spite of all these activities the performance of the LDCs is mixed. The share of the value of LDC exports, excluding arms, that 99 100 101 102

See http://www.agoa.gov/resources/TRDPROFL03.pdf. See Breuss / Griller / Vranes (2003). Bachetta / Bora (2002). IMF / The World Bank (2002), 20.

Does the ‘Development Round’ Foster Development?

269

enters developed country markets duty-free has increase since 1996 (from 71.5 to 75.3 percent). However, when the figure is further adjusted for oil there is a clear downward trend by around 12 percentage points (see Table 5). This downward trend reflects the shift in LDC exports to products and export markets that are not dutyfree. In fact, the trade values show that there is basically no increase in the value of duty-free imports from LDCs while at the same time there is a significant increase in the dutiable imports from LDCs.103 In 2001, the average trade weighted tariff facing LDC agricultural exports into developed country markets was 3.2 percent. The equivalent figures for textiles and clothing are 4.5 and 8.5, respectively.104 LDCs account for less than one half of one percent of world trade. In the Doha Ministerial Declaration, Ministers committed themselves to considering additional measures to progressively improve market access for LDCs and to the objective of duty-free and quota-free access for products originating in LCDs. Similar goals were also proclaimed in the context of the eighth Millennium Development Goal (MDG). The WTO noticed that in 2000, the distribution of markets for LDC products remains heavily concentrated.105 Sixty-three percent of all exports go to the EU and the United States. In addition to the EU and US, the major developed country markets are Australia, Canada, Japan, Norway and Switzerland. Together the developed countries import 69 percent of total LDC exports. Three of the top five markets are developing countries in East Asia, China, Republic of Korea and Thailand. These countries account for 20 percent of total LDC exports. The remaining top 10 markets are: Canada, India, Japan, Singapore and Chinese Taipei. The market penetration of LDC exports is greatest in India and Thailand at 2.1 percent, followed by the EU at 1.4 percent. This in a way underlines the general trend of the last decade of a steady increase of the South-South trade noticed by the WTO.106 Accordingly, over the eleven-year period (1990-2001), South-South trade expanded twice as fast as world trade. The share of developing country exports to developing countries rose from 28 percent to 103 104 105 106

WTO (2003), 126. Ibidem. Ibidem. Ibidem, 24 et seqq.

270

Fritz Breuss

37 percent of their total exports. In the same period, the import share rose by 10 percentage points to over 41 percent. More than two-thirds of intra-developing country trade originates from and is destined to developing Asia.107 G. Regionalism versus Multilateralism 1. Increasing Attractiveness of RTAs The global trading system has seen a sharp increase in regional trade agreements (RTAs) over the past decade. A total of 259 RTAs had been notified to the GATT/WTO by the end of December 2002,108 although only 176 RTAs are currently in force. An additional 70 RTAs are estimated to be operational although not yet notified and about 70 are under negotiation. As of March 2003, only four WTO members – Hong Kong, China; Macao, China; Mongolia and Chinese Taipei – were not party to a regional trade agreement. With the sole exception of Mongolia, these WTO members are all engaged in negotiations on preferential agreements. While the recent rapid growth of RTAs began in the 1990s, the seeds of this development were sown in the early 1980s. Part of the impulse towards regionalism was driven then by the bleak prospects for progress on the multilateral agenda and the decision of the United States to explore the preferential approach to trade. The United States signed its first free trade agreement (FTA) with Israel in the mid-1980s, followed by an FTA with Canada in 1988 and the North American Free Trade Agreement (NAFTA) in 1994. The current negotiations on free trade for the Americas (FTAA) span two continents and involve over 30 countries. More recently, also Japan and other Asian countries have departed from exclusive reliance on MFN-based trade. The WTO sees the major explanation for the expansion in the number of RTAs in the 1990s in the collapse of the COMECON (the preferential arrangement involving the old Soviet Union and Eastern European countries) and the alignment of the Central and Eastern European countries (CEEC) to the EU.109 Of the 123 new RTAs in force since 1990 (see Table 6), covering trade in goods, about a third were signed among transition economies (e.g. 107 Ibidem, 25. 108 Ibidem, 46 et seq. 109 Ibidem, 46.

Does the ‘Development Round’ Foster Development?

271

CEFTA). Another third were agreements concluded as part of the effort of the transition economies to integrate with the EU (e.g. Europe Agreements). Regional agreements among developing countries (SouthSouth RTAs) account for about 30-40 percent of all RTAs currently in force, including those not notified to the WTO. In Africa alone, there are about eighteen trading agreements. Partly, they establish customs unions or common markets. They tend to encompass a large number of countries and have extended transition periods, often 20 or 30 years, which makes some recent RTAs more a declaration of intent than agreements affecting actual trade flows. The number of RTAs signed between developed and developing countries (North-South RTAs) has increased over the years. The EU has played a major role in this respect through a series of agreements with a number of countries including Turkey (customs union since 1996), Mexico, South Africa and Chile. Euro-Mediterranean Association Agreements have also been concluded or are being negotiated between the EU and the countries of North Africa and the Middle East. These replace the earlier non-reciprocal RTAs signed in the 1970s. Moreover, the post-Lomé Cotonou agreements will be negotiated between the EU and the ACP countries on the basis of reciprocal preferential trade. Table 6: Notified RTAs in Goods by the Date of Entry into Force and Type of Partners (as of January 2003) N-N N-S N-E S-S S-E E-E Total 1958-1964 2 0 0 1 0 0 3 1965-1969 0 0 0 0 1 0 1 1970-1974 5 3 0 2 0 0 10 1975-1979 0 5 0 1 0 0 6 1980-1984 2 1 0 1 0 0 4 1985-1989 1 1 0 2 0 0 4 1990-1994 3 3 12 5 0 6 29 1995-1999 3 7 10 4 12 28 64 2000-2002 0 11 4 5 4 6 30 Total 16 31 26 21 17 40 151 N (North) = developed countries, including Canada, the United States, EU (15), EFTA, Japan, Australia and New Zealand; E (East) = transition countries, including the former Soviet Union, Eastern and Central Europe, the Baltic States and the Balkans; S (South) = developing countries, including the remaining countries. Source: WTO (2003), 47.

272

Fritz Breuss

There are several economic and political considerations why countries are interested in regionalism.110 Anyhow, the international trading system is increasingly characterized by a complex network of preferential trade regimes, sitting side-by-side with the WTO multilateral trading system.111 There is a rise in cross-regional bilateral agreements, the growing involvement of countries that have traditionally remained outside regional arrangements, the development of inter-linked (overlapping) agreements, and considerable variations in the design and content of RTAs. One third of the FTAs currently under negotiations are among countries that belong to different geographical areas. All major countries are involved in cross-regional FTAs. The EU has concluded FTAs with Mexico, Chile, South Africa and numerous other African and Middle Eastern countries and is in the process of negotiating regional agreements with ACP countries under the framework of the Cotonou Agreement. The EU is also negotiating an agreement with MERCOSUR. EFTA has signed a FTA with Mexico and various African countries, and is negotiating FTAs with Canada, Chile and South Africa. The United States signed a FTA with Jordan, and is negotiating with Australia, Chile, Egypt and Singapore. Countries that have traditionally remained outside regional agreements are now negotiating and joining RTAs. The last major country to join the trend is Japan, which singed a FTA with Singapore in January 2002. Table 7: Preferential Trade Share of Intra RTAs Trade in Merchandise Imports, 2002 and 2005 (as of January 2003) Western Europe Transition economies North America (incl. Mexico) Africa Middle East Latin America (excl. Mexico) Asia World Source: WTO (2003), 48.

2000 64.7 61.6 41.4 37.2 19.2 18.3 5.6 43.2

2005 67.0 61.6 51.6 43.6 38.1 63.6 16.2 51.2

110 See WTO (2003), 49 et seqq. for a discussion. 111 See WTO (2003), 53 for an impressive world map of such a network.

Does the ‘Development Round’ Foster Development?

273

The significance of RTAs is demonstrated in the Tables 7 and 8. Table 7 shows that 43 percent of world merchandise trade now occurs under the umbrella of preferential trade arrangements. This share will increase as more RTAs are negotiated in the future. If all RTAs under negotiation at present are successfully concluded within the next three years, over 50 percent of world merchandise trade will then occur among countries linked by preferential agreements. Table 8: Intra-regional Export Shares, 1970-2001 (Ratio of trade among members over total trade with members and non-members) 1970 1980 1985 1990 1995 2000 2001 in force since Europe and North America CEFTA ... … … … 14.6 11.5 12.4 1993 EU 59.5 60.8 59.2 65.9 62.4 62.1 61.2 1957 NAFTA 36.0 33.6 43.9 41.4 46.2 55.7 54.8 1994 Latin America and the Caribbean CACM 26.0 24.4 14.4 15.4 21.7 13.7 15.0 1961 Andean Group 1.8 3.8 3.2 4.2 12.2 8.8 11.2 1988 CARICOM 4.2 5.3 6.3 8.1 12.1 14.6 13.4 1973 MERCOSUR 9.4 11.6 5.5 8.9 20.3 20.7 20.8 1991 Africa CEMAC 4.8 1.6 1.9 2.3 2.2 1.2 1.3 1999 (UDEAC) COMESAi) 7.4 5.7 4.4 6.3 6.0 4.8 5.2 1994 ECCAS 9.8 1.4 1.7 1.4 1.5 0.9 1.1 1983iii) ECOWAS 2.9 9.6 5.1 8.0 9.0 9.6 9.8 1975iii) ii) SADC 4.2 0.4 1.4 3.1 10.6 11.9 10.9 1992iii) UEMOA 6.2 9.9 8.7 12.1 10.3 13.0 13.5 2000 Middle East and Asia ASEAN/ 22.4 17.4 18.6 19.0 24.6 23.0 22.4 1992 AFTA GCC 4.6 3.0 4.9 8.0 6.8 5.0 5.1 1981iii) SAARC 3.2 4.8 4.5 3.2 4.4 4.3 4.9 1985c i) Prior to 2000, data unavailable for Namibia and Swaziland. ii) Prior to 2000, data unavailable for Botswana, Lesotho and Swaziland. iii) Year of foundation. Source: WTO (2003), 56.

However, not all trade among preferential trading takes place at preferential rates. Most agreements exclude certain sensitive sectors or even agricultural trade altogether. Traders may choose to

274

Fritz Breuss

forgo preferential treatment because the costs of satisfying the requisite rules of origin might be higher than the advantage offered by the preferential margin.112 Moreover, many applied MFN tariffs in developed countries are already zero. For these reasons, the estimates reported in Table 7 will overstate the impact of preferential trading arrangement as far as tariffs are concerned. The WTO in defending multilateralism sees no strong evidence of trade creation due to RTAs.113 Its own data can, however, also be interpreted differently (see Table 8). The share of intra-regional exports as a percentage of regional bloc exports has been increasing since 1970 in most of the major regional trade blocs. Over 60 percent of EU exports are to other EU-15 partners, a share which may increase by an additional 10 percentage points after enlargement by 10 new Member States on May 1, 2004. Taking into account that the EU and EFTA have free trade arrangements via the Free Trade Agreements of 1973 one must also add EU-EFTA intra-trade which accounts for another five percentage points. That amounts to a share of intra-EU-EFTA trade of around three quarters of their total trade. Over half of NAFTA exports are to other NAFTA partners. Whereas intra-regional export shares within the EU have remained nearly constant, the intra-NAFTA trade share has shown an upward trend well before NAFTA entered into force in 1994. Similar patterns can be identified for other major RTAs. MERCOSUR is an exception, where data show a sharp increase in intra-regional export shares after the agreement entered into force in 1991. 2. Does the EU Need the WTO at All? The significance of RTAs depends on its degree of integration. Most of the RTAs are simple free trade agreements or customs unions. Only the NAFTA goes a little bit beyond this status implying besides free movements of goods also partial free movement of services and of capital and labour. The highest form of integration however, is reached in the EU. It not only consists of a customs union and the four freedoms (Single Market concept) but is also an Economic and Monetary Union (EMU) with a single currency (Euro). EU-25 together with the EU-EFTA free trade realtions will comprise a potential of nearly three quarters of intra-EU-EFTA regional trade relative to total trade. Other RTAs (with the ACP, 112 WTO (2000), 48. 113 Ibidem, 55.

Does the ‘Development Round’ Foster Development?

275

with Mexico, South Africa, the Andean pact) make the EU a trading superpower of its own. In addition, in contrast to most RTAs the EU does go much beyond the WTO commitments.114 This provokes the question why the EU needs the WTO at all? Nearly three quarter of all merchandise trade is conducted with free trade partners. Measured by this huge intra-trade share the EU could easily forgo the multilateral commitments of the WTO. The EU is, however, also a leading world trader. The EU holds a world market share of around 19 percent (excluding intra-EU trade). Adding EFTA’s world market share of around three percentage points and that of the 10 new EU members of Central and Eastern Europe of also around three percentage points this increases the world market share of the enlarged EU-EFTA to 25 percent. In comparison the United States hold a market share of 14 percent, Japan one of only nine percent and China nearly seven percent, followed by Canada with five percent. All other single countries are far below a five percent world market share (measured with exports figures of the year 2002).115 If the EU would just rely on its bilateral and unilateral agreements with developing countries via GSP and the Lomé-Cotonou agreements with the ACP countries it would be self sustainable. Intra-EU trade would cover industrial trade, EUACP trade would fulfil the function of supplying the necessary raw material inputs. The only important trade partners in the industrial world are the United States and to a lesser degree Japan. From an optimal tariff theory point of view a dominant player in world trade (such as the EU) can improve welfare by imposing an optimal tariff rate which is usually very high (more than 100 percent), whereas the optimal tariff for small countries is zero (i.e. free trade).116 This implies that the starting position for trade liberalization is quite different for large and small (and developing) countries. Larger countries and regions tend to lose from unilateral reductions in protection, which move them further away from their optimal level of protection. Only mutual or co-operative tariff reductions can benefit all countries. This highlights the importance of the WTO multilateral negotiations for a co-operative solution to bring together the different interests of large and small, rich and poor countries, resulting in world-wide trade liberalization. 114 See OECD (2002). 115 See WTO (2003), 69. 116 See Breuss (2003), 140 et seqq.

276

Fritz Breuss

Bagwell and Staiger in formulating an economic theory of GATT/WTO reach the conclusion that GATT’s major principles, reciprocity and non-discrimination are thus simple rules that, when used together can deliver an efficient outcome.117 Non-discrimination (the MFN clause) ensures that all international externalities are channelled through world-price movements, and the principle of reciprocity serves effectively to neutralize externalities of exactly this nature (terms-of-trade effects). More interesting in answering the question whether the EU needs the WTO, however, is the treatment of cases of bilateral imbalances. Maggi analyses such cases in the context of a three-country CGE model of trade with the presence of bilateral imbalances of power, where the ‘more powerful’ country (e.g. the EU) in a given pair is the one that stands to lose less (or to gain more) from a trade war (increasing tariffs). Hence, in case of trade liberalization a large and powerful country would lose more (or gain less) than smaller countries.118 Seen from a world perspective of welfare maximization and the target of equality Maggi comes to the following conclusions: a) In the presence of bilateral imbalances of power (e.g. the EU versus small and developing countries), countries can sustain a higher symmetric welfare with multilateral enforcement (or punishment strategy whereby any defection is followed by a permanent Nash reversion in multilateral relationships) than with bilateral enforcement. b) Under bilateral enforcement (punishment strategy whereby any defection is followed by a permanent Nash reversion in bilateral relationships) the weaker partner makes a larger ‘concession’ than the stronger partner.119 The reverse is true under multilateral enforcement. To overcome this inequality, international transfers (e.g. development aid) could be used. According to Maggi such a strategy is, however, an imperfect substitute for multilateral enforcement.120 c) Under absent power imbalances (if all countries would be of equal size and power), bilateral and multilateral enforcement are equally efficient. An important conclusion of Maggi’s game-theoretic analysis is that international trade institutions like WTO are important in order to neutralise the imbalances of power in world trade via rule-making procedures. 117 118 119 120

Bagwell / Staiger (1999); see also Breuss (2003), 140 et seqq. Maggi (1999); see also Breuss (2003), 155 et seqq. Maggi (1999), 198. Ibidem, 200.

Does the ‘Development Round’ Foster Development?

277

The more power the EU accrues in world trade the more is its attitude towards multilateral solutions of WTO no more than a fig leaf to calm down its own sole or the critiques of globalization or a demonstration of its good governance. H. Developing Countries’ Market Access in Agriculture A key achievement of the Uruguay Round has been to extend multilateral discipline to domestic support in the farming sector, as well as to export subsidies. Domestic support related measures have been classified according to the associated level of market distortions. The so-called ‘boxes’ characterise what is prohibited, allowed, or to be phased out. A slight reduction in the market distortion (domestic support granted to farmers) can be observed in the 1990s. The ratio of producer support for the OECD was 31 percent in 2000-2002, compared to 36 percent in 1986-1988. The coefficient of nominal protection was 31 percent in 2002, compared to 57 percent in 1986-1988.121 Agriculture belongs to the built-in Agenda of the Uruguay Round and hence also in the Doha Round. Rich countries can afford farm support. Therefore the US$ 300 billion spent by the industrialised countries on farm support are often compared to the amount of their aid to development, which represents only a sixth of this sum.122 The Official Development Assistance (ODA) defined by the OECD123 as the sum of grants and concessional loans (i.e. with grant elements of at least 25 percent), undertaken by the official sector and with the primary objective of promoting the economic development stayed at US$ 53 billion in the last decade (2002, US$ 58.3 billion).124 In contrast total long-term capital flows to developing countries increased from US$ 98 billion in 1990 to over US$ 295 billion in 2000.125 Nevertheless the share of ODA decreased 121 See OECD (2003). 122 See Fontagné (2003), 6. 123 Chapter 3 of the OECD (2004) report describes the „Millennium Development Goals’ (MDG) defined at the UN Millennium Summit. The Millennium Declaration signed by 189 countries, including 147 Heads of State, in September 2000 can be found on the UN website at http://ods-dds-ny.un.org/doc/UNDOC/GEN/N00/559/51/PDF/N0 055951.pdf?OpenElement. 124 See OECD (2004), 22. 125 See European Commission (2002), 65.

278

Fritz Breuss

from 0.33 percent in 1990 to 0.23 percent of GDP in 2002126 and hence further away from the recommended aid target of 0.7 percent. In 2002 only five OECD members (Denmark, Luxembourg, Netherlands, Norway and Sweden) have reached the level of 0.7 percent GDP to be spent on ODA. 1. Costs of Agricultural Distortion According to the comprehensive analyses of international organizations increased market access for agricultural products would work to directly address poverty reduction in developing countries.127 While the rapid expansion of demand for unskilled labour in manufacturing and urban services in many developing countries has sharply reduced rural poverty, about three quarters of the world’s poor still live in rural areas, where agriculture is often the dominant economic activity. Agriculture accounts for about 27 percent of GDP in developing countries, a similar share of exports and 50 percent of employment. This dependency on agriculture is most pronounced in LDCs and in Sub-Saharan Africa, where, in addition, production tends to be concentrated in only a small number of commodities. Agricultural distortions inflict large costs on the global economy, by some estimates exceeding those of protection in the industrial sector. Based on CGE model simulations with the GTAP5 model the IMF and World Bank study finds that the global income loss from agricultural distortions worldwide may be well over US$ 128 billion (US$ 98 billion in industrial and US$ 30 billion in developing countries).128 Most of the cost results from market price support measures, of which tariffs are the dominant form. The losses of export revenues (US$ 378 billion worldwide) are much larger, by a factor of three to four in the case of developing countries (US$ 256 billion in industrial and US$ 122 in developing countries). Interestingly, both groups of countries suffer the most from their own restrictive polices. For developing countries, these policies are responsible for about 71 percent of the total income loss, while for developed (industrial) countries, the share is as high as 95 126 See OECD (2004), 62. 127 IMF / The World Bank (2002), chapter III; UNCTAD (2003), chapter III; WTO (2003), chapter II. 128 IMF / The World Bank (2002), 32.

Does the ‘Development Round’ Foster Development?

279

percent. Within the group of the industrial countries the major players in protecting its agricultural markets are the United States and the European Union.129 For example, the United States granted US$ 3.6 billion subsidy to US cotton producers in 2001.130 2. Removing Agricultural Subsidies alone is Negative for Developing Countries While global liberalization of both tariffs and subsidies would benefit every region, the static effect of removing subsidies alone is likely to be negative for developing countries as a group and many individual countries. It is less clear whether the various objectives contemplated in the DDA are mutually compatible.131 If market access is favourable to growth in the LDCs, then liberalizing imports in the North on a multilateral basis will erode the margin of preference conceded to LDCs and will reduce their access to these markets. If less distorting farm support in the North increases world prices of food products, LDCs that are net importers of food will be adversely affected through negative terms-of-trade effects.132 This can be demonstrated by the simulation outcomes of the UNCTAD study.133 The elimination of export subsidies in agriculture, without parallel changes in tariffs leads to modest worldwide welfare losses. The distributional implications are the following (see Table 9): After the elimination of subsidies, all regions except Europe start increasing their agricultural value-added. However, since many countries still face high protection against their agriculture export, this shift might be counterproductive. Most regions actually stand to lose from the elimination of subsidies, while the gains appear to be very concentrated in Western Europe – which (in particular due to the CAP of the EU) is the area characterised by the highest value of initial subsidies – and in regions that are net agri129 For a short description of the main features of the US Farm and Security and Rural Investment Act of 2002 and of the EU’s Common Agricultural Policy (CAP) and its reform proposals, the 2002 midterm review of the CAP, see IMF / The World Bank (2002), 29 et seq. 130 See Fontagné (2003), 6. 131 Ibidem, 3. 132 For similar results, see also Hoekman et al. (2002). 133 UNCTAD (2003), 44 et seq.

280

Fritz Breuss

cultural exporters, such as Oceania and Latin America. Western Europe (the EU) gains both from better resource allocation (the elimination of subsidies brings the specialization pattern of this region more into line with its natural comparative advantages) and from improved terms of trade. On the other hand the removal of export subsidies directly reduces the agricultural exports of Western Europe, thus leading to a lower world supply for these goods and to improved terms of trade for Europe, whose exports are sold now at higher prices on international markets. As for the terms-of-trade effects on the other regions, they depend on their agricultural export pattern. Countries that are net agriculture and food exporters (e.g. North America, Oceania and Latin America) are likely to gain, while those that are not may lose (e.g. Asian NICs and North Africa and Sub-Saharan Africa) – which hits again primarily the developing countries. Table 9: Liberalization in Agriculture: Export Subsidy Removal. Results of Model Simulations of the Elimination of Export Subsidies in Agriculture, without Parallel Changes in Tariffs Welfare effects Aggregate trade data Percentage Total value (‘97 Exports Terms of trade change US$ mill.) Percentage change -0.008 -73.9 0.008 -0.007 -0.015 -178.8 0.006 -0.013 -0.000 -1.9 0.125 0.082 0.033 2.410.0 -0.124 0.065 -0.001 -88.0 -0.013 0.013 -0.117 -891.5 -0.056 -0.172 -0.113 -354.9 -0.234 -0.161

Regions Asian NICs China South Asia Western Europe North America Transition econ. Sub-Saharan Africa Oceania 0.024 100.1 North Africa and -0.283 -2.209.7 Middle East Latin America 0.004 80.3 Japan -0.013 -484.9 Rest of the World -0.063 -158.7 Total -1.851.7 Source: UNCTAD (2003), 44 et seq.

0.107 -0.148

0.119 -0.296

0.056 -0.047 -0.225

0.035 -0.061 -0.189

3. Agricultural Liberalization Benefits all Countries A comprehensive agricultural liberalization, which is both part of the built-in WTO agenda and one of the major pillars to the Doha

Does the ‘Development Round’ Foster Development?

281

agreement could be beneficial for both the developing and the developed world. Whereas an isolated elimination of export subsidies may have negative effects for the developing countries, tariff liberalization is beneficial for all countries, depending, however, on the existing rates of protection in the particular commodities. Table 10: Agricultural Tariff Liberalization: Results of Model Simulations of a Worldwide 50 % Cut in all Agricultural Tariffs Welfare effects Aggregate trade data Percentage Total value (‘97 Exports Terms of trade change US$ mill.) Percentage change 0.342 3,363.6 0.888 -0.072 0.082 964.0 1.199 -0.083 0.074 361.2 1.954 -0.302 0.021 1,562.1 0.476 0.006 0.046 3,613.3 0.914 0.266 0.118 900.8 1.474 -0.045

Regions Asian NICs China South Asia Western Europe North America Transition economies Sub-Saharan 0.072 Africa Oceania 0.419 North Africa 0.387 and Middle East Latin America 0.073 Japan 0.116 Rest of the World 0.110 Total Source: UNCTAD (2003), 43.

226.2

1.810

-0.210

1,719.8 3,033.8

2.299 2.829

1.833 -0.595

1,304.7 4,221.2 277.1 21,547.9

1.708 1.763 2.248

0.056 -0.392 0.223

According to UNCTAD simulations, a worldwide reduction of 50 percent in all agricultural tariffs brings about an aggregate welfare gain of US$ 21.5 billion (see Table 6).134 All the world regions appear to gain, but gains differ widely both in absolute and in relative terms. The largest absolute gains are captured by Japan, North America, the Asian NICs, North Africa and the Middle East and Oceania. In percentage terms, those regions that appear to gain most are Oceania, the Asian NICs and North Africa. The estimated percentage gain for Sub-Saharan Africa and Latin America are relatively low. The aggregate trade data indicate that the value of exports increases in all regions after liberalization. Lower worldwide protection in agriculture translates into increased worldwide 134 Ibidem, 43.

282

Fritz Breuss

import demand and improved trade opportunities in all areas. Interestingly, the strongest increases are in the developing regions North Africa and Middle East, Oceania and Sub-Saharan Africa. In Western Europe the export gains are modest (see Table 10). In addition to this static welfare gains a worldwide liberalization in agricultural trade would cause additional gains. On the one hand it would stabilize agricultural prices and lead to a downward pressure on world prices for key commodities. Developing countries suffer the most from price instability as they have fewer resources available to smooth consumption and income flows. On the other hand agricultural liberalization is likely to have long-term, dynamic effects on world production and trade. These effects could include increased farm investment and enhanced technologies and productivity in response to better market opportunities. According to the IMF and the World Bank for the developing countries to reap the full benefits of liberalization, a framework of supportive policies is required – including the elimination of anti-agriculture biases in pricing policies so that (higher) world prices are passed through to the farm-gate – and essential infrastructure (transport, logistic, credit, extension services).135 Hoekman and his collaborators in recognizing that a number of developing countries could lose from agricultural liberalization (in particular from reducing domestic support measures) suggest that agricultural reforms should be accompanied by compensation mechanism, which could include additional ‘aid for trade’.136 I. Developing Countries’ Market Access in Textile and Clothes The production of textiles and clothing is intensive in unskilled labour and uses only simple technology. Therefore, a regional shift in specialization (from the developed to the developing countries) took place since the sixties. In the mid-1960s, developing countries accounted for 15 percent of world textile exports and less than 25 percent of world clothing exports. By the end of the 1990s, these shares had reached 50 percent and 70 percent respectively.137 Although exports of textiles and clothing (T&C) account only for 5.5 percentage points of total world exports (see Appendix, Table A1) their production is highly concentrated in the developing countries. 135 IMF / The World Bank (2002), 33. 136 Hoekman and Hoekman et al. (2002), 18. 137 See IMF / The World Bank (2002), 36 et seq.

Does the ‘Development Round’ Foster Development?

283

This development has created a high dependency on these products for export earnings. In some developing countries textiles and clothing account for more than 80 percent of total merchandise exports.138 In Pakistan (73 percent), Mauritius (64 percent), Sri Lanka (54 percent), Tunisia (43 percent), Turkey (38 percent) and Morocco (34 percent) these products account for more than one third of their total exports. Then there is a group of countries with export shares between 20 and 30 percent (India, 28; Romania, 24; China, 21 percent). Despite extensive quantitative restrictions discriminating against developing countries and high tariffs in developed countries (the main export markets for most developing countries) there was a remarkable growth in T&C exports. For nearly half a century, world trade in T&C has been subject to quantitative restrictions under derogation from GATT rules, beginning with Japan’s 1955 voluntary export restraints on its export of cotton fabrics and clothing to the United States. These led to the multilateral ShortTerm Arrangement regarding International Trade in Cotton Textiles in 1961, the Long-Term Arrangement in 1962, and eventually the Multifibre Arrangement (MFA) in 1974.139 The MFA expanded quantitative restrictions beyond cotton products to wool and manmade fibre products and was extended several times until the Uruguay Round Agreement on Textiles and Clothing (ATC) took effect at the beginning of 1995. In the MFA’s last year of operation, six participants (Austria, Canada, EU, Finland, Norway, and the United States) applied quotas under the Agreement. The salient feature of the MFA was bilateral quotas. The MFA called on importing countries to endeavour to grow quota volumes by at least 6 percent per year. In practice quota growth was lower for established suppliers, while small and new exporting countries were generally granted more generous quota growth. MFA quotas act like bilateral export restrictions. The export tax equivalents of these quotas vary substantially across countries. The most competitive exporting countries, such as China and India, face more stringent restrictions than the less competitive countries. The export tax equivalents of quotas for textiles varied from zero in Japan to 7.8 percent in Canada (USA, 6.7 percent; EU, 4.5 percent). The tax 138 Cambodia, Macao SAR and Bangladesh, see IMF / The World Bank, (2002), 38. 139 See IMF / The World Bank (2002), 36.

284

Fritz Breuss

equivalents of quotas for clothing were higher (Japan zero; Canada, 16.8 percent; USA 11 percent; EU 5.3 percent).140 Under the ATC, quotas are to be phase out progressively over a 10-year period (ending in 2005). In the first stage (January 1, 1995), WTO members were required to integrate products representing not less than 16 percent in volume terms of their 1990 imports of textiles and clothing products. In stage 2 (January 1998), not less than a further 17 percent had to be integrated, and in stage 3 (January 2002), a further 18 percent. Finally, on January 1, 2005, all remaining products (amounting to a maximum 49 percent) are to be automatically integrated. After integration, regular GATT safeguards apply. In spite of the time plan of the ATC, market access barriers still remained. With the exception of Norway, whose T&C imports have been all freed from quotas, the major importers (the United States and the EU) eliminated only a small percentage of quotas originally in place during the first two stages (the USA from 750 quotas only 13, the EU from 219 only 14). Therefore, the vast majority of restrictions is left to be abolished at the end of the implementation period.141 In addition to the MFA quotas, T&C imports are subject to exceptionally high tariffs. During the Uruguay Round tariffs on T&C were cut less than those on other manufactures, and tariff peaks and escalation (tariffs on clothing are higher than those on textiles) remain common in this sector. In OECD import regimes, tariff peaks affect 27 percent of total tariff lines on T&C. Tariffs in the United States for textiles are 11.2 percent (for clothing 13.3 percent; for all manufactures 2.8 percent), in the EU 9.1, 11.9 and 3.6 percent respectively, in Japan 8.5, 12.5 and 1.4 percent respectively and in Canada 15.7, 21.2 and 3.9 respectively.142 However, tariffs on T&C are also very high in developing countries (16 percent). According to simulations with the GTAP model by the IMF and the World Bank, the cost of barriers to trade in T&C imports are a substantial burden on both developing and industrial countries. The combined income loss for developing countries quota and tariffs on industrial country imports amounts to US$ 24 billion, and the export revenue loss to US$ 40 billion. Industrial countries suffer 140 Ibidem, 39. 141 See ibidem, 41 for more details. 142 See ibidem, 39.

Does the ‘Development Round’ Foster Development?

285

around half the income loss but almost the same export shortfall as developing countries, namely around US$ 40-46 billion. As intradeveloping country trade accounts for about half of their total exports of textiles and 20 percent of clothing exports, the cost of barriers (mainly import tariffs) are larger for developing countries (US$ 28 billion) than for industrial countries (US$ 3 billion). The losses of exports amount to US$ 42 billion for developing countries and only for US$ 9 billion for industrial countries. Overall, the incomes loss due to MFA quotas and T&C tariffs for the world amount to US$ 66 billion, and the losses of export revenues are US$ 137 billion world-wide.143 The same model simulated indicates that as many as 27 million jobs are foregone in developing countries due to the combined effect of quota and tariffs.144 On average, each job saved in developed countries by tariffs and quotas is estimated to cost 35 jobs in developing countries, many of which are in China and India. MFA quotas and tariffs tend to be most hurtful to the poor, as T&C industries primarily employ low-skilled workers, often migrants from rural areas. Also, low-income households in industrial countries bear the brunt of the cost of MFA and high tariff restrictions as the poor spend a larger share of their income on necessities such as T&C. The delay in the integration plan according to the ATC leads to adjustment needs. The above mentioned back loading of effective liberalization turns what was planned as a gradual adjustment process into a shock at the end of the transition period (in 2005) – for both importing and exporting countries. According to the IMF and World Bank study, this raises concern that political pressures might spark greater recourse to other forms of protection once quotas are phase out, with trade remedy action and perhaps non-transparent ‘voluntary’ export restraints (prohibited in principle under the WTO) becoming a ‘new line of defence’.145 A sudden withdrawal of quota protection at the end of the transition period might also increase resistance to further reductions in tariffs. Accelerating the removal of quotas on textiles and clothing imports is an urgent priority. Additionally, the Doha round negotiations should lead to substantially lower tariffs on T&C trade, in both industrial and developing countries. 143 Ibidem, 42 et seq. 144 Ibidem, 3. 145 Ibidem, 44.

286

Fritz Breuss

Preferences for LDCs are not a long-term solution to problems of competitiveness, but schemes that provide LCDs with duty- and quota-free market access may ease problems of transition. Benefits of preferential market access can be substantially reduced and even negated by restrictive rules of origin, either because of the need to switch to higher-cost sourcing of intermediate goods, or because value added thresholds for preferential access are hard to meet. Mattoo et al. estimate that benefits under the AGOA could be reduced by as much as 75-80 percent because of onerous rules of region (there are similar problems with EU preferential schemes).146 In addition to the rules of origin, there is often a fine balance to strike on social and environmental conditions. As demonstrated in the case of minimum labour standards in connection with Cambodia’s exports of T&C, minimum wages and other conditions for preferential access may have limited the benefits the country has drawn from these preferential schemes.147 IV. Gains from Trade Liberalization under a Successful Doha Round

A. Formula Approaches to Market Access Negotiations 1. The General Procedure of Negotiations Concerning market access for products, there are 146 members negotiating on thousands of products. Under such circumstances, any means for simplifying negotiations are welcome. One instrument is the ‘formula approach’. A variety of approaches has been used in the past to negotiate the reduction of bound tariffs starting with the request and offer approach. This technique, grounded in the selected product by product approach proved cumbersome and yielded results that were not particularly ambitious. Two significant departures from this approach occurred during the Kennedy Round. Industrialised countries adopted a linear tariff reduction technique and developing countries were granted ‘less than full reciprocity’.148 In the Tokyo Round an explicit reference was made to ‘appropriate formulae’. A number of proposals were submitted in re146 Mattoo et al. (2002). 147 See IMF / The World Bank (2002), 46 for a case study on Cambodia. 148 Hoda (2001).

Does the ‘Development Round’ Foster Development?

287

sponse to the mandate, including some that had the effect of higher reductions for higher tariff rates in contrast to a linear reduction. The proposal from Switzerland was ultimately adopted by some countries. In the so-called ‘Swiss formula’149 the target tariff rate, t1 depends non-linearly on the initial tariff rate t 0 and a coefficient a : at 0 , (3) t1 a  t0 where a is a coefficient corresponding to the upper limit of desired tariff rates after the cut. In applying this formula some countries used a coefficient a equal to 14, others adopted 16. The formula was not universally applied by all countries and those that applied it did so with exceptions. Francois and Martin survey a range of formula options and examine both targeted and flexible applications of the Swiss formula that target tariff escalation and peaks, and would allow policymakers to directly target how far they will move towards free trade, while providing some flexibility for trading off reductions in peak tariffs against reductions in lower-tariff sectors.150 The mandate for the Uruguay Round negotiations and the DDA did not specifically mention the use of formulae as the core modality. However, during both negotiations proposals for modalities based on formulae have figured prominently. In the current Doha agriculture negotiations some Members proposed the Swiss formula with a coefficient a equal 25. In the non-agricultural market access negotiations the Swiss formula was proposed by the United States with a coefficient of 8 for certain phases of their proposed tariff reduction plan. In addition, variants of the Swiss formula that take into account the diversity of Members’ profiles were proposed.151

149 See WTO (2003), 150; Fontagné (2003), 6; Francois / Martin (2003). 150 Francois / Martin (2003), see also Panagariya (2002) for more details on the general properties of formulas that have been used for reciprocal negotiations. 151 See WTO (2003), 150 et seq. with references to the documents for the full variety of proposals.

288

Fritz Breuss

The Chair’s draft proposal for the Doha negotiations on agriculture followed the approach used during the Uruguay Round which was a target rate of reduction based on a simple average of out-quota tariff rates with a minimum cut per line. The reductions would apply across three different bands of tariffs with a higher average reduction for tariffs in the high range. Developing countries were proposed a similar approach, but with higher thresholds for tariffs to be reduced and lower percentage reductions. In the Doha non-agricultural market access negotiations the Chair proposed a number of elements for the reduction of tariffs. The core element is the following formula to be applied on a lineby-line basis: Bt a t 0 t1 (4) Bt a  t 0 where, t1 is the final rate, to be bound in ad valorem terms; t0 is the base rate for negotiations, ta is the average of the base rates and B is a coefficient with a unique value to be determined by the participants. Less than reciprocity in this context is incorporated into the formula through the ta coefficient. A higher coefficient implies a lower reduction and developing countries in general have higher average applied and bound tariffs (see Table 1). The Chair further proposed that WTO members could consider the elimination of tariffs in certain sectors of export interest to developing countries. As with agriculture, the Chair’s proposal in non-agricultural market access takes into account the issue of special and differential treatment (SDT) for developing countries. The impact of a 50 percent Swiss-formula based reduction of average tariff bindings for industrial and developing countries has been demonstrated152 (see Table 11a and Table 11b). With the implementation of the Uruguay Round commitments, average ad valorem tariffs in the industrial countries, generally are around 3 percent (see Table 11a, first column). However, there are important exceptions. One of these is textiles and clothing, where the average rate is roughly three times this overall average. This is reflected in the standard deviation and maximum tariff columns. With full implementation of current commitments, the estimated simple average industrial tariff in the United States is 3.2 percent, 152 Francois / Martin (2003); Francois / van Meijl / van Tongeren (2003).

Does the ‘Development Round’ Foster Development?

289

with a standard deviation of 4.3, and a maximum tariff of 37.5 percent. The European Union has a higher average (3.7 percent), but less dispersion (3.6 percent) and a maximum tariff rate of 17 percent. For the developing countries, average industrial tariffs range from a low of 3 to 4 percent to a high of more than 20 percent. In Table 11a the data for three developing countries are presented: Brazil, India, and Thailand. Brazil’s tariffs are all bound, though the average rate for industrial products is 14.9 percentage points above the currently applied rate (binding overhang). India and Thailand’s tariffs are partially covered by bindings, again with significant binding overhang. Table 11a: The Post-Uruguay Round and ITAi) Applied Tariffs in Selected Industrial and Developing Countries Post-UR and ITA tariffs (in %) Standard Maximum Binding deviation tariff overhangii) Agriculture EU 5.9 7.5 74.9 0.3 Japan 6.2 8.1 43.3 1.2 USA 3.5 7.4 90.0 0.5 Brazil 12.9 5.1 27.0 22.6 India 31.0 20.8 150.0 90.7 Thailand 26.5 14.4 65.0 7.1 Non-agriculture EU 3.7 3.6 17.0 0.4 Japan 2.3 3.4 30.9 0.1 USA 3.2 4.3 37.5 0.2 Brazil 15.9 6.0 35.0 14.9 India 19.2 16.5 40.0 3.9 Thailand 10.5 10.8 80.0 7.8 i) ITA = Information Technology Agreement. ii) Binding overhang is the gap between the bound rate and the current applied tariff rate. Source: Francois / van Meijl / van Tongeren (2003), Table 2.2. Simple average

Because of the small binding overhang in industrial countries a 50 percentage point tariff reduction according to the Swiss formula leads also to a comparable strong reduction in applied tariffs (see last column in Table 11b). In general, for developing countries, the binding overhang is large enough that reductions in the range of 50 percent are necessary to force any reductions in average applied rates for countries like Brazil. For many countries, even this will

290

Fritz Breuss

have little or no effect, as tariffs are largely unbound. This limits severely the negotiating leverage of developing countries in the WTO. This is also why the debate over using bound, applied, or ‘historic’ rates in the WTO as a starting point for negotiations is important.153 Table 11b: Effects of Basic Swiss Formula Reductions: Applied Tariffs after a 50 % Cut in Average Tariff Bindings Standard Maximum Binding ø% deviation tariff overhang reduction Agriculture: EU 3.0 2.9 10.9 0.1 -48.6 Japan 3.5 3.7 13.9 0.2 -43.0 USA 1.9 2.4 11.5 0.1 -46.6 Brazil 12.4 4.6 22.3 5.3 -3.7 India 29.5 14.9 70.8 31.3 -4.8 Thailand 15.1 6.3 30.1 1.7 -43.0 Non-agriculture EU 1.9 1.4 5.0 0.1 -47.7 Japan 1.2 1.4 5.6 0.0 -48.5 USA 1.7 1.6 6.1 0.0 -48.3 Brazil 13.5 4.2 16.7 1.9 -15.4 India 11.3 9.2 30.5 0.3 -41.3 Thailand 7.2 6.1 20.7 2.0 -31.6 Source: Francois / van Meijl / van Tongeren (2003), Table 2.2. Simple average

2. The Danger of Preference Erosion for the Developing Countries A formula approach fits well the objectives of the DDA and has advantages and caveats:154 1. Advantages: By strongly reducing tariff peaks, it offers better access to LDC exports in labour intensive and agricultural goods. It opens other developing countries’ markets that remain currently highly protected and thus stimulates SouthSouth trade. It allows a different a coefficient for developed and developing economies and, hence, respects the spirit of the SDT; last but not least it allows also a different coefficient for trade in manufactures and food products in order to match political economy constraints.

153 See Francois / van Meijl / van Tongeren (2003), 3. 154 See Fontagné (2003), 6 et seqq.

Does the ‘Development Round’ Foster Development?

291

Shortcomings: This means, however, ‘killing too many birds with one stone’.155 The advantage of differentiating the a coefficient contradicts the objective of making agricultural markets more open, or of enhancing South-South trade. Tariff peaks potentially affect exports of LDCs which are nevertheless conceded preferential market access (ACP countries, the GSP scheme, AGOA, EBA, etc.). Hence, any (non-linear) formula approach will have two effects: a) to eradicate the remaining peaks faced by LCDs exporters, and b) to erode the margin of preferences they had been conceded. The net effect may be negative. Fontagné by discussing a number of other subtle problems connected with the impact of formula based trade liberalization for developing countries simulates the different formula approaches with a multi-country CGE model (see Table 12).156

2.

Table 12: Formula Approaches – which Differences for Welfare Changes? (Simulations with a CGE world model) Linear Linear Swiss Swiss formula Long-run welfare formula formula formula + SDT change in % excl. peaks EU-25 0.38 0.14 0.55 0.47 USA 0.18 0.09 0.24 0.12 Japan 0.86 0.29 1.45 1.29 Cairns 0.30 0.14 0.35 0.39 Developing Asia 0.80 0.28 1.07 0.91 ACP countries 0.43 0.26 0.41 0.29 Other countries 0.55 0.20 0.79 0.70 World 0.42 0.16 0.61 0.51 Assumptions for coefficient a = 35 % for the linear formula t1 = at0 and a Swiss formula, and a coefficient a equal to 28 (manufactures) or 58 (food). Source: Fontagné (2003), 8.

Fontagné focuses on market access with a menu of scenarios in which developing countries are conceded SDT. Bilateral tariffs at the product level (HS-6 level), derived from MAcMaps, are cut

155 Ibidem, 6. 156 Ibidem, 7 et seq.

292

Fritz Breuss

according to a linear formula (where tariff peaks157 can be included or excluded from the liberalization), versus a truncated Swiss formula158 (applied to all tariffs). Coefficients of reduction are those suggested by previous rounds, and the SDT offered to developing countries is a lower coefficient of linear reduction and a larger a coefficient in the Swiss formula.159 The results of Table 12 highlight that benefits of increased market access at the world level are much higher with a Swiss formula and, in contrast, rather limited if one adopts a linear formula excluding peaks. Considering the Swiss formula combined with the SDT, the largest benefits accrue to Japan, where agriculture is highly protected. This is also why EU gains are much larger than US ones. Lastly, ACP countries record very limited gains, in particular in comparison to developing Asia which has in the past been conceded less preferences by industrialised importers. In summing up, Fontagné asserts that a formula approach leads to a sizeable erosion of preferences conceded to the poorest developing countries so far, with the aim of favouring exports of small and insufficiently diversified economies. The more specialised the exporters, the larger the benefits extracted in the past from preferential access schemes and the stronger the adverse effects of market opening they will have to cope with.160 B. How to Estimate Gains from Multilateral Trade Liberalization? The estimation in quantitative terms of the potential gains from trade liberalization is now done primarily with CGE models. The systematic use of CGE models to simulate the effects of trade negotiations started during the Tokyo Round.161 Since then rapid progress has been made as regards both modelling and data collection and assembly. Results from CGE simulations found a wide echo before the conclusion of the Uruguay Round, showing that nearly 157 Tariff peaks are defined as those superior to 15 percent in manufacturing, energy and raw materials, and those above 85 percent in agriculture and agrofood. 158 In a truncated Swiss formula, the reduction is linear up to the threshold defined as a tariff peak, and non-linear thereafter. 159 Fontagné (2003), 8. 160 Ibidem. 161 See Deardorff / Stern (1981); Deardorff / Stern (1986); Whalley (1985).

Does the ‘Development Round’ Foster Development?

293

all countries would have lost opportunities from a failure to reach agreement.162 In recent years, several CGE analyses of the effects of trade policy reforms in a future WTO Round have been produced.163 Some of them only consider agricultural liberalization (with welfare gains ranging from US$ 27 billion to US$ 384 billion, depending on the liberalization assumptions and the model type), others include manufacturing tariff reform (the welfare gains range from US$ 284 billion to US$ 1,210 billion). Only a few analyses consider the impact of service trade liberalization, mainly because of poor data on trade flows in the services sector and poor measurement of service trade barriers. The few studies considering liberalization in trade in services produce very large welfare gains.164 These large gains are due to two reasons: 1) Services account for a large share in consumption in most middle and high-income countries, much larger than that of agriculture. 2) Services are major inputs in the production of manufactures. Estimations of the World Bank of the effects of a 100 percent cut in merchandise protection and a 100 percent cut in service protection lead to a welfare gain in the static case of US$ 255 billion and in the dynamic version of US$ 830 billion.165 Brown et al. with the same liberalization experiment reach welfare gains of US$ 1,857 billion.166 There are several reasons for the large discrepancies in the estimations of welfare gains: a) assumption about the deepness of the liberalization; b) results are sensitive to model specifications: liberalization gains are higher in models allowing for increasing returns to scale and imperfect competition in the manufacturing sector. The gains are further enhanced in specifications allowing for dynamic (growth) effects of trade liberalization (trade-related changes in savings and investment or with development in productivity); c) also the chosen baseline influences the outcome; in most recent studies the GATP database (of 1997) is used to replicate the world economy. These most updated versions yield lower estimates of the 162 See the surveys of Harrison et al. (1997), and Francois et al. (1994) and (1996). 163 For a survey, see UNCTAD (2003), 39, and also Francois (2000). 164 See Brown et al. (2001); World Bank (2001). 165 World Bank (2001). 166 Brown et al. (2001).

294

Fritz Breuss

world-wide liberalization effects since the status-quo level of trade barriers is lower (after the tariff cuts of the Uruguay Round); d) also the dimensionality of the models (the number of sectors and regions considered) influences the outcome. Notwithstanding the notable differences in results from the various CGE analyses, there are a number of common findings:167 1) Global welfare results of agricultural liberalization are quite similar across models and studies. This is due to the consensus of modelling agriculture as a constant returns to scale sector where trade-related dynamic gains are quite limited. 2) A common feature of static, constant returns to scale CGE models is that the global gains associated with (full) agricultural liberalization are not very different from those originating from trade liberalization in manufactures. 3) Concerning the source of gains, almost all studies show that the major source of the gains accruing to each country is its own liberalization, rather than that of partner countries. 4) As for the distribution of the global gains between developed and developing countries, in the majority of the studies it was found that the gains are shared quite equally between the two groups. Among developing countries, Asian countries will reap the largest gains, while the gains for Latin American and African countries will be more limited. Even possible losses are found for Sub-Saharan countries associated with agricultural liberalization, markedly with terms-of-trade developments as a consequence of export subsidies removal. In the following two recent CGE model exercises are presented. One is a simulation of a world-wide cut of all merchandise tariffs (agricultural and manufactures) with a static model.168 The other one by Francois et al. is more ambitious as it considers a complete liberalization package (agriculture, manufactures, services) and uses a model with imperfect competition and dynamic elements.169 1. UNCTAD Simulations UNCTAD uses a static CGE model based on GTAP5 data base with a benchmark year 1997, disaggregated to six sectors (natural resources, manufactures, primary and processed agricultural products, 167 See UNCTAD (2003), 40. 168 Ibidem, 48 et seq. 169 Francois / van Meijl / van Tongeren (2003)

Does the ‘Development Round’ Foster Development?

295

textiles and apparel, services) and twelve world regions of which four are developing country regions.170 A worldwide 50 percent reduction of all merchandise tariffs (agricultural and manufactures) – a so-called comprehensive liberalization scenario – leads to the following results (see Table 13): The global welfare gain of around US$ 40 billion is almost twice as much than that arising from liberalization in agriculture only (see Table 10). The big gainers from adding manufacturing liberalization to agriculture liberalization are the Asian regions (+0.6 percentage points increase in welfare), followed by North Africa and Middle East (+0.4 percentage points), China and Oceania. Some countries, however, will not have an advantage from extending liberalization beyond agriculture. These are in particular North America (Canada, United States), transition economies (Hungary, Poland, the Rest of Central European countries, Former Soviet Union) and Sub-Saharan Africa, which would suffer from terms-of-trade losses by adding manufacturing liberalization. All these countries would see their market shares in textiles and clothing and other manufactures eroded by surging imports from Asia.

170 UNCTAD (2003), 48 et seq.

296

Fritz Breuss

Table 13: A Comprehensive Doha Round Liberalization Scenario: Results of Model Simulations of a 50 % Worldwide Cut in Tariffs on all Merchandise Trade Regions

Welfare effects Aggregate trade data Percentage Total value Exports Terms of trade change (‘97 US$ mill.) Percentage change 0.674 6,636.5 3.899 0.168 0.424 5,017.1 7.458 0.012 0.282 1,383.3 12.043 -1.747 0.075 5,489.6 1.105 0.078 0.023 1,778.0 2.591 -0.008 0.079 603.1 3.860 -0.483

Asian NICs China South Asia Western Europe North America Transition economies Sub-Saharan 0.004 Africa Oceania 0.386 North Africa and 0.476 Middle East Latin America 0.079 Japan 0.307 Rest of the world 0.281 Total Source: UNCTAD (2003), 48.

13.3

4.590

-0.927

1,584.1 3,735.8

4.265 5.004

1.435 -0.806

1,414.0 11,207.4 706.3 39.568.5

5.719 5.512 8.789

-0.734 0.752 0.091

The removal of all tariff protection boosts exports in all areas (see Table 13). The increase is in general much stronger than that associated with the elimination of agricultural tariffs only. The bigger increases in exports occur in low- to middle-income Asian countries (China, South Asia), followed by other developing countries and by Japan and Oceania. Western Europe and North America do not achieve major expansion of their exports. These results – not taking into account dynamic effects and imperfect competition in the manufacturing sectors – underline what was found in previous studies,171 namely that the inclusion of manufacturing liberalization in a ‘comprehensive round’ of negotiations would be especially interesting for the developing countries. This conclusion holds for developing economies taken as a single broad aggregate. There are, however, regions, in particular Sub-Saharan Africa, that might actually lose from extending liberalization from agriculture alone to all merchandise trade. 171 See Hertel / Martin (2000).

Does the ‘Development Round’ Foster Development?

297

2. CGE Model Simulations with Imperfect Competition Francois, van Meijl and van Tongeren with a CGE model with imperfect (monopolistic) competition and investment (dynamic) effects simulated the potential gains from a comprehensive DohaRound trade liberalization.172 They take the short-run (static) and the long-run (dynamic model) view. They include in their scenarios of trade liberalization tariff reductions (in agricultural, manufactures and in services trade), elimination of border controls, export subsidies, agricultural support and trade facilitation. The data on tariffs are taken from the WTO’s integrated database, with supplemental information from the World Bank’s recent assessment of detailed pre- and post-Uruguay Round tariff schedules and from the UNCTAD/World Bank WITS dataset. All of this tariff information has been concorded to GTAP model sectors. Services trade barriers are based on the gravity model estimates. Also the schedule of China accession commitments is implemented. While the basic GTAP dataset is benchmarked to 1997, and reflects applied tariffs actually in place in 1997, the authors want to work with a representation of a post-Uruguay Round world. This includes the accession of China, the enlargement of the EU; and Agenda 2000 reforms a part of the baseline. In the baseline scenario therefore the following inclusions have been made:173 x Implementation of the rest of the Uruguay Round tariff commitments, x implementation of the ATC phasing-out quotas, x implementation China’s accession to the WTO, x implementation of Agenda 2000, x implementation of the EU enlargement by 12 countries (EU27). The CGE model has been aggregated to 17 sectors and 16 regions. Perfect competition is assumed in the agricultural sectors. The manufacturing and services sectors involve imperfect (monopolistic) competition. There is a dynamic link, whereby the static or direct income effects of trade liberalization induce shifts in the regional pattern of savings and investment. These effects relate to classical models of capital accumulation and growth, rather than to endogenous growth mechanism. How much these ‘accumulation 172 Francois / van Meijl / van Tongeren (2003). 173 Ibidem, 8.

298

Fritz Breuss

effects’ will supplement static effects depends on the marginal product of capital and underlying savings behaviour. This specification allows to differentiate between short-run versus long-run effects. In the short-run capital stocks are fixed and in the long-run capital stocks adjust. The model includes the basic features of ‘economic geography’ models, including intermediate linkages, monopolistic competition, and returns from specialization.174 In the model estimations three scenarios are carried out:175 The first two are partial liberalization scenarios. In the ‘Linear 50%’ scenario all trade instruments are reduced by 50 percent. This involves a 50 percent reduction in agricultural and industrial tariffs and export subsidies, a 50 percent reduction in OECD domestic support for agricultures, a 50 percent reduction in the tariff-equivalent of services barriers, and a partial reduction in trading costs, related to trade facilitation measures (1.5 percent of the value of trade). The second partial liberalization experiment is called the ‘Swiss formula’ experiment. In this experiment the reduction in import tariffs in agriculture and manufactures is based on a straight Swiss formula with a coefficient of 25, meaning the maximum tariff is reduced to 25 percent. The third scenario simply involves full elimination of all trade barriers, implying in the case of trade facilitation a reduction of trade costs of 3 percent of the value of trade.

174 Ibidem, 7. 175 Ibidem , see also Table 14.

Does the ‘Development Round’ Foster Development?

299

Table 14: Three Liberalization Scenarios Scenario definition Instruments Linear 50% Swiss formula Full liberalization Import tariffs in 50% reduction Swiss formula 100% agriculture and Reduction (with a reduction manufacturing max 25% tariff) Estimated border 50% reduction 50% reduction 100% measures in services reduction 50% reduction 50% reduction 100% Export subsidies reduction Domestic agricul50% reduction 50% reduction 100% tural support in reduction OECD countries Trade facilitation 1.5% of value 1.5% of value of 1.5% of value of trade trade of trade Source: Francois / van Meilj / van Tongeren (2003), Table 2.5

The results are reported in Table 15.176 They refer to static welfare gains from a linear 50 percent liberalization (scenario I). The overall effects of agricultural liberalization are not clear-cut. Liberalization of domestic support in the OECD is generally positive for the OECD (in particular positive for EU countries and North America), however, negative for the food-importing SubSaharan Africa. But also other developing countries would lose. Even Australia and New Zealand, both net agricultural exporters gain only marginally from liberalizing domestic support. Agricultural liberalization of border measures sees all countries as winners if assuming constant returns to scale. If, however, increasing returns to scale are considered a number of countries are losers: Mediterranean, China, India, Australia and New Zealand, South Africa and Sub-Saharan Africa.

176 Ibidem, 11-15.

300

Fritz Breuss

Table 15: Static Welfare Effects (Equivalent Variations) from a Linear 50% Liberalization (Scenario I of Table 14), US$ mill. OECD Agriculture (border measures) Agriculture (domestic support) Manufactures (border measures) Services liberalization Trade facilitation Total

24.482 (16.818)

Developing countries 32.446 (11.083)

Other countries 4.630

Total (World) 61.558 [96.743]

8.744

-

711

9.455 [12.368]

12.057 (5.622)

22.230 (12.012)

2.789

37.076 [54.247]

17.225 6.907 1.963 26.095 (17.918) (5.609) [53.053] 46.159 26.152 5.881 78.192 (41.204) (21.953) [150.870] 108.667 87.735 15.974 212.376 (81.562) (50.657) [367.281] The figures in round brackets refer to the constant returns to scale case; all other figures refer to the increasing returns to scale case. The figures in square brackets refer to the overall welfare gains of a full liberalization (100 percent; scenario III of Table 10). Source: Francois / van Meijl / van Tongeren (2003), Table 2.6 and Tables 4.1 to 4.4.

The results for manufacturing liberalization are more consistent and generally positive. Generally, with increasing returns to scale the effects are twice that of constant returns to scale. Similar to the UNCTAD results the developing countries gain nearly twice as much as the OECD countries from opening-up the markets for manufactures. The gains in the industrial world must be lower because the OECD tariffs are, on average, already below 3 percent for manufacturing whereas the tariffs of the developing countries are five time as high (see Table 11a). China will be hurt by significant manufacturing liberalization. Once the WTO accession is fully implemented, the Doha round cannot be expected to yield much additional gains for China. The negative results for China follow primarily from an erosion of its terms of trade. Another important source of overall effects is services, which yield static income gains similar to those of liberalizing manufacturing. One obvious winner from services liberalization is the

Does the ‘Development Round’ Foster Development?

301

United States, which picks up more than half of the global welfare gains. Another big (not so obvious) winner is India. An astonishing result of the simulations by Francois, van Meijl and van Tongeren is that the gains from trade facilitation amount to nearly half of the total welfare gains (see Table 15).177 Biggest winners are North America and High Income Asia but also the EU. Overall the total welfare gain from a really comprehensive Doha round liberalization would be between US$ 212 billion (for a partial 50 percent liberalization) and US$ 367 billion. In terms of labour market effects, both unskilled and skilled labour gain from the partial and full liberalization scenarios in most regions, except for some cases in the CEEC economies and in China. Similarly, there are positive wage effects for unskilled workers in all regions, except for China in all scenarios and the CEEC in some cases.178 Overall export effects are clear-cut.179 Export growth, under all scenarios, is greatest in the developing countries, especially in Asia and the Pacific (including India and China), but also in the Mediterranean, African, and Latin American economies. The CEEC suffer from trade-erosion with respect to market access to the EU-15 economies. A decomposition of bilateral trade effects shows that much of the potential gains for developing countries depend on the realization of South-South trade opportunities. This would foster the positive development in this respect in the last decade.180 The bilateral trade effects – decomposed into the broad aggregate trade flows between the major world regions: North-North, North-South and South-South – resulting form the liberalization experiments by Francois, van Meijl and van Tongeren underpin the above raised question whether the EU needs the WTO at all (see chapter II.F.2).181

177 178 179 180 181

Ibidem. Ibidem, 13. Ibidem, 13 and Figure 4.2. See WTO (2003), 24 et seqq. Francois / van Meijl / van Tongeren (2003), 14.

302

Fritz Breuss

Table 16: Trade Effects of a Linear 50% Liberalization Experiment (percentage change in value of bilateral exports; short-run results with increasing returns to scale) ÈfromÆto

EU-27 NORTH -6

Non-OECD SOUTH 21

Other NORTH 13

EU-27 NORTH Non-OECD 30 39 25 SOUTH Other 12 26 8 NORTH Total 3 28 14 imports (5) (35) (15) The figures in brackets refer to the long-run (dynamic) results. Source: Francois / van Meijl / van Tongeren (2003), Table 4.5.

Total exports 2 (4) 30 (38) 14 (15) 12 (15)

The enlarged EU-27 can increase its total exports only by 2 percent compared to a 12 percent growth in world trade (see Table 16). One reason is that EU countries mostly trade amongst themselves. The benefits from removing the intra-EU barriers have already been realised in the past (customs union, Single Market) and there are no additional gains for intra-EU trade in a new WTO round. A second driver of this result is the increased competition for non-EU countries on EU markets. Simulated intra-EU-27 trade shrinks by 6 percentage points as other suppliers enter the EU markets. While in the past the share of trade within the EU (intraEU trade) was biased upward (trade creation) lowering external trade barriers by the EU in the Doha round will inevitable lead to the erosion of the intra-EU trade preferences. One can expect that the current bias towards intra-EU trade will be reduced. The most impressive growth in markets share is realized by suppliers from developing countries, who are simulated to expand their exports to the EU by 30 percent, compared to the 12 percent increase of imports from other developed countries. Because there is no positive growth to be expected from intra-EU trade, European exports can only be increased by expansion in non-EU markets. Exports to developing countries will grow with 21 percent and exports to the other regions will grow with 13 percent. Developing countries obtain the highest growth in exports (30 percent). They expand exports to all destinations, though the largest trade surge is observed for intra-developing country (South-South) trade. This will expand by 39 percentage points.

Does the ‘Development Round’ Foster Development?

303

The simulation experiments by Francois, van Meijl and van Tongeren, including nearly all possible issues of market access issues under the Doha Round show remarkable welfare gains. Nevertheless Singapore issues (the relationship between trade and investment policy, the interaction between trade and competition policies, transparency in government procurement practices, and trade facilitation practices) are only partly and then somewhat exaggerated (e.g. in the case of trade facilitation) included. Very little is known about the welfare and/or growth effects of a more secure network of regulations concerning Foreign Direct Investments (FDI), a cornerstone of globalization of multinational firms. V. Conclusions

The development agenda of the Doha Round may turn out to be a mere act of window-dressing. Most empirical studies confirm that there is no simple relationship between openness (trade liberalization) and growth and therefore development. Recent experience with the success and/or failure of poverty alleviation programs at the World Bank and the IMF may, however, help to teach the WTO that it should concentrate on its primal business, namely market access. The mandates of the WTO increased from Round to Round. In the Doha Round a climax in complexity seems to have been reached which may well be the reason that it is so difficult to reach an overall agreement. Given the heterogeneity of mandates, it is not always clear whether the various objectives (development, Singapore issues, market access in general) in the DDA are mutually compatible. The simulation studies discussed above give a clear picture of winners and losers of a delayed Doha Round. The large industrial countries and regions (United States with NAFTA) and the enlarged European Union with its extended regional trade agreements (e.g. with the ACP) have no distinctive interest in further liberalizing world trade. Its intra-trade shares are already – in particular in the case of the EU – so large that any further liberalization only hurts them. The big losers of further delayed trade liberalization are the developing countries. All simulations studies show that they would gain the most from further market access in particular when eliminating the still considerably high tariff peaks in products where they have comparative advantage vis-à-vis the industrial countries. Although the average industrial goods tariffs are already very low, there are still in place many subtle forms of protectionism (like

304

Fritz Breuss

tariff peaks and escalation and contingent protection) – in particular in the relations between the North and the South. Furthermore the tariffs in the developing countries are generally higher than in the industrial countries. It is therefore of utmost importance to eliminate the tariff peaks and generally to cut tariffs in developing countries. This would increase welfare in the developed world and spur South-South trade.

References

Daron Acemoglu / Simon Johnson / James A. Robinson (2001), The Colonial Origins of Comparative Development: An Empirical Investigation, in: The American Economic Review 91 (2001), No. 5, 1369-1401. Alberto F. Alesina / Dani Rodrik (1994), Distributive Politics and Economic Growth, in: Quarterly Journal of Economics 109 (1994), 465-490. Kym Anderson / Jikun Huang / Elena Ianchovichina (2004), Will China’s WTO Accession Worsen Rural Poverty?, CEPR Discussion Paper Series, No 4196, London, January 2004. Marc Bacchetta / Bijit Bora (2001), Post-Uruguay Round Market Access Barriers for Industrial Products, UNCTAD Policy Issues in International Trade and Commodities, Study Series No. 12, Sales No. E.01.II.D.23, New York / Geneva (United Nations) 2001. Marc Bacchetta / Bijit Bora (2002), Market Access for Industrial Products and the Doha Development Agenda, paper prepared for the World Bank Conference on ‘Informing the Doha Process: New Trade Research for Developing Countries’, Cairo, May 20-21, 2002. Harald Badinger / Fritz Breuss (2004), What Has Determined the Rapid Post-War Growth of Intra-EU Trade?, in: Review of World Economics (Weltwirtschaftliches Archiv), Vol. 140 (2004), No. 1, 31-51. Kyle Bagwell / Robert W. Staiger (1999), An Economic Theory of GATT, in: The American Economic Review 89 (1999), No. 1, 215-248.

Does the ‘Development Round’ Foster Development?

305

Scott L. Baier / Jeffrey H. Bergstrand (2001), The Growth of World Trade: tariffs, transport costs and income similarity, in: Journal of International Economics 53 (2001), 1-17. Paul Bairoch (1993), Economics and World History: Myths and Paradoxes, Chicago / IL (The University of Chicago Press) 1993. Robert J. Barro (1999), Inequality and Growth in a Panel of Countries, Harvard University, 1999. Antoine Bouët / Lionel Fontagné / Mondher Mimouni / Xavier Pichot (2001), Market Access Maps: A Bilateral and Disaggregated Measure of Market Access, CEPII Working Paper No. 18, Centre d’Etudes Prospectives et d’Informations Internationales, Paris 2001. Paul Brenton / Miriam Manchin (2002), Making the EU Trade Agreements Work: The Role of Rules of Origin, CEPS Working Document No 183, Center for European Policy Studies, Brussels 2002. Fritz Breuss (2003), WTO Dispute Settlement in Practice – An Economic Assessment, in: Breuss / Griller / Vranes (2003), 131-184. Fritz Breuss / Stefan Griller / Erich Vranes (Eds.) (2003), The Banana Dispute: An Economic and Legal Analysis (= Europainstitut Wirtschaftsuniversität Wien Publication Series Vol. 19), Wien / New York (Springer) 2003. Drusilla K. Brown / Alan V. Deardorff / Robert M. Stern (2001), CGE Modelling and Analysis of Multilateral and Regional Neogtiation Options, Research Seminar in International Economics, Discussion Paper No. 468, University of Michigan: Ann Arbor/MI 2001. Ha-Joon Chang (2002), Kicking away the Ladder, London (Anthem Press: Wimbledon Publishing Company) 2002. David T. Coe / Elhanan Helpman (1995), International R&D spillovers, in: European Economic Review 39 (1995), No. 5, 859887. David T. Coe / Elhanan Helpman / Alexander W. Hoffmaister (1997), North-south R&D spillovers, in: The Economic Journal 107 (1997), 134-149.

306

Fritz Breuss

Judith M. Dean / Seema Desai / James Riedel (1994), Trade Policy Reform in Developing Countries since 1985: A review of the Evidence, World Bank Discussion Papers 267, Washington/DC 1994. Alan V. Deardorff / Robert M. Stern (1981), A disaggregated model of world production and trade: An estimate of the impact of the Tokyo Round, in: Journal of Policy Modeling 3 (1981), 127152. Alan V. Deardorff / Robert M. Stern (1986), The Michigan Model of World Production and Trade: Theory and Application, Cambridge/MA / London (The MIT Press) 1986. Jared Diamond (1997), Guns, Germs, and Steel, New York (W. W. Norton & Co.) 1997. Betina V. Dimaranan / Robert A. McDougall (Eds.) (2002), Global Trade, Assistance, and Production: The GTAP 5 Data Base, Purdue University, Center for Global Trade Analysis, West Lafayette/IN, May 2002. David Dollar / Aart Kraay (2001), Growth is Good for the Poor, The World Bank Development Research Group, Macroeconomics and Growth, Policy Research Working Paper, WPS 2587, Washington/DC, April 2001. David Dollar / Aart Kraay (2002), Institutions, Trade, and Growth, Carnegie-Rochester Conference Series on Public Policy, 2002, available at http://www.carnegie-rochester.rochester.edu/April 02-pdfs/ITG2.pdf. Steve Dowrick (1994), Openness and Growth, RBA Annual Conference Volume, 2, Reserve Bank of Australia, 1994. Steve Dowrick / J. Bradford deLong (2001), Globalisation and Convergence, Paper for NBER Conference on Globalisation in Historical Perspective, Santa Barbara, California, May 4-5, 2001, available at http://www.nps.navy.mil/rmmcnab/intlecon/ intlecon01/Globalization%20and%20Convergence.pdf. European Commission (2002), Responses to the Challenges of Globalisation: A Study on the International Monetary and Financial System and on Financing for Development, Working Document of the Commission Services, Commission of the European Communities, SEC(2002) 185 final, Brussels, 14 February 2002.

Does the ‘Development Round’ Foster Development?

307

Simon J. Evenett (2003), The Doha Development Agenda: The Failure of the WTO Ministerial Meeting in Cancun: Implications for Future Research, CESifo Forum, Vol. 4, No. 3, Autumn 2003, 11-17. Lionel Fontagné (2003), The Doha Development Agenda: Market Access and Domestic Support Measures, CESifo Forum, Vol. 4, No. 3, Autumn 2003, 3-10. J. Michael Finger (1993), Antidumping: How It Works and Who Gets Hurt, Ann Arbor/MI (Michigan University Press) 1993. Jeffrey A. Frankel / David Romer (1999), Does Trade Cause Growth?, in: The American Economic Review, 89 (1999) No. 3, 379-399. Joseph F. Francois (2000), Assessing the Results of General Equilibrium Studies of Multilateral Trade Negotiations, UNCTAD Policy Issues in International Trade and Commodities, Study Series, No. 3, Sales No. E.00.II.D.24, New York / Geneva (United Nations) 2000. Joseph F. Francois / Will Martin (2003), Formula Approaches for Market Access Negotiations, CEPR Discussion Paper Series, No. 3720, London, January 2003. Joseph F. Francois / Brad McDonald / Håkan Nordström (1994), The Uruguay Round: A Global General Equilibrium Assessment, Working Paper, GATT, Geneva, November 1994. Joseph F. Francois / Brad McDonald / Håkan Nordström (1996), A Users’s Guide to Uruguay Round Assessments, CEPR Discussion Paper Series, No. 1410, London, June 1996. Joseph F. Francois / Hans van Meijl / Frank van Tongeren (2003), Trade Liberalization and Developing Countries under the Doha Round, CEPR Discussion Paper Series, No. 4032, London, August 2003. John L. Gallup / Jeffrey D. Sachs / Aandrew D. Mellinger (1998), Geography and Economic Development, NBER Working Paper, No. W6849, December 1998. David Greenaway / Wyn Morgan / Peter Wright (1998), Trade Reform, Adjustment and Growth: What does the Evidence Tell us?, in: The Economic Journal 108 (1998), Issue 450, 15471561.

308

Fritz Breuss

Spencer J. Henson / Rupert J. Loader / Alan Swinbank / Maury Bedahl / Nicole Lux (2000), Impact of Sanitary and Phytosanitary Measures on Developing Countries, Centre for Food Economics Research, University of Reading, Reading 2000. Robert Hall / Charles I. Jones (1999), Why Do Some Countries Produce So Much More Output per Worker than Others?, in: Quarterly Journal of Economics, 114 (1999), No. 1, 83-116. Glenn W. Harrison / Thomas F. Rutherford / David G. Tarr (1997), Quantifying the Uruguay Round, in: The Economic Journal 107 (1997), 1405-1430. Glenn W. Harrison / Thomas F. Rutherford / David G. Tarr (2003), Trade liberalization, poverty and efficient equity, in: Journal of Development Economics 71 (2003), 97-128. Tom W. Hertel / Will Martin (2000), Liberalizing agriculture and manufactures in a millennium round: Implications for developing countries, in: World Economy 23 (2000), 455-470. Anwarul Hoda (2001), Tariff Negotiations and Renegotiations under the GATT and WTO: Procedures and Practices, Cambridge (Cambridge University Press) 2001. Bernard Hoekman (2002a), Strengthening the Global Trade Architecture for Development: The Post-Doha Agenda, in: World Trade Review 1 (2002), No. 1, 23-45. Bernard Hoekman (2002b), Economic Development and the WTO After Doha, CEPR Discussion Paper Series, No. 3374, London, May 2002. Bernard Hoekman / Petros Mavroïdis (1996), Dumping, Antidumping and Antitrust, in: Journal of World Trade 30 (1996), No. 1, 27-52. Bernard Hoekman / Francis Ng / Marcelo Olarreaga (2001), Eliminating Excessive Tariffs on Exports of Least Developed Countries, Washington/D.C. (The World Bank) 2001 (published in: The World Bank Economic Review 16 (2002), No. 1, 1-21). Bernard Hoekman / Francis Ng / Marcelo Olarreaga (2002), Reducing Agriculture Tariffs versus Domestic Support: What’s More Important for Developing Countries?, CEPR Discussion Paper Series, No. 3576, London, October 2002.

Does the ‘Development Round’ Foster Development?

309

Elena Ianchovichina / Aaditya Mattoo / Marcelo Olarreaga (2001), Unrestricted Market Access for Sub-Saharan Africa: How much is it worth and who pays?, World Bank Working Paper, No. 2595, April 2001. IMF (2003), The tale of two giants: India and China, IMF Survey, International Monetary Fund, Vol. 32, No. 22, December 15, 2003. IMF and The World Bank (2002), Market Access for Developing Country Exports – Selected Issues, Prepared by the Staffs of the International Monetary Fund (IMF) and World Bank, approved by Timothy Geithner and Gobind Nankani, Washington/D.C., September 26, 2002. Daniel Kaufman / Aart Kraay / Pablo Zoido-Lobatón (2002), Governance Matters II – Updated Indicators for 2000/01, World Bank Policy Research department Working Paper No. 2772, Washington/D.C. 2002. Jane Kennan / Christopher Stevens (1997), From Lomé to the GSP: Implications for the ACP of Losing Lomé Trade preferences, Institute for Development Studies, Brighton (University of Sussex) 1997. Michael Knetter / Thomas Prusa (2000), Macroeconomic Factors and Antidumping Filings: Evidence from Four Countries, NBER Working Paper No. W8010, National Bureau of Economic Research, Cambridge/MA 2000. Simon Kuznets (1955), Economic Growth and Income Inequality, in: The American Economic Review, 45 (1955), No. 1, 1-28. Thomas Leoni / Wolfgang Pollan (2003), The Impact of Inequality on Economic Growth, WIFO Working Papers No. 211, November 2003. Giovanni Maggi (1999), The Role of Multilateral Institutions in International Trade Cooperation, in: The American Economic Review 89 (1999), No. 1, 190-214. Branko Milanovic (2002), Making Globalization Work, presentation at the Brookings/ILO/CEIP conference, Washington/D.C., December 2, 2002. Branko Milanovic (2003), Income Convergence During the Disintegration of the World Economy 1919-39, World Bank Policy Research Working Paper 1941, The World Bank, Development Research Group, Washington/DC, January 2003.

310

Fritz Breuss

Aaditya Mattoo / Devesh Roy / Arvind Subramanian (2002), The Africa Growth and Opportunity Act: Rules of Origin and the Impact on Market Access, IMF Working Paper WP/02, Washington/D.C., May 2002. Douglass C. North (1990), Institutions, Institutional Change and Economic Performance, New York (Cambridge University Press) 1990. OECD (1997), Indicators of Tariff and Non-Tariff Trade Barriers, Paris (Organization for Economic Co-operation and Development) 1997. OECD (2003), Agricultural Policies in OECD Countries – Monitoring and Evaluation, Paris (Organization for Economic Cooperation and Development) 2003. OECD (2004), Development Co-operation 2003 Report, in: The DAC Journal 5 (2004), No. 1, i-237. Caglar Ozden / Eric Reinhardt (2002), The Perversity of Preferences: GSP and Developing Country Trade Policy, 1976-2000, Emory University, Atlanta/GA 2002 (mimeo). Kevin H. O’Rourke (2000), Tariffs and Growth in the Late 19th Century, in: The Economic Journal, 110 (2000), Issue 463, 456-483. Arvind Panagariya (2002), Formula Approaches to Reciprocal Tariff Liberalization, in: Bernard Hoekman / Aaditya Mattoo / Philip English (Eds.), Development, Trade, and the WTO. A Handbook, Washington/D.C. (The World Bank) 2002, 535539. Roberto Perotti (1993), Political Equilibrium, Income Distribution and Growth, in: Review of Economic Studies 60 (1993), 755776. Roberto Perotti (1996), Growth, Income Distribution and Democracy: What the Data Say, in: Journal of Economic Growth 1 (1996), 149-187. Torsten Persson / Guido Tabellini (1994), Is Inequality Harmful for Growth?, in: The American Economic Review 84 (1994), 600621. Martin Ravallion (2003), The Debate on Globalization, Poverty, and Inequality: Why Measurement Matters, The World Bank

Does the ‘Development Round’ Foster Development?

311

Policy Research Working Paper, No. 3038, Policy Research Group, The World Bank, Washington DC, April 21, 2003. Jeffrey J. Reimer (2002), Estimating the Poverty Impacts of Trade Liberalization, GTAP Working Paper No. 20, Purdue University, 2002. Francisco Rodríguez / Dani Rodrik (2001), Trade Policy and Economic Growth: A Sceptic’s Guide to Cross-National Literature, in: Ben S. Bernanke / Kenneth S. Rogoff (Eds.), NBER Macro Annual 2000, Cambridge/MA 2001. Dani Rodrik (2003a), Growth Strategies, Harvard University, John F. Kennedy School of Government, Cambridge/MA, September 2003. Dani Rodrik (2003b), Institutions, Integration, and Geography: In Search of the Deep Determinants of Economic Growth, in: Dani Rodrik (Ed.), In Search of Prosperity: Analytic Country Studies on Growth, Princeton/NJ (Princeton University Press) 2003. Dani Rodrik / Aarvind Subramanian / Francesco Trebbi (2002), Institutions Rule: The Primacy of Institutions over Geography and Integration in Economic Development, unpublished paper, Harvard University, Cambridge/MA, October 2002. Andrew K. Rose (2003), Which International Institutions Promote International Trade?, CEPR Discussion Paper Series, No. 3764, London, January 2003. Adrew K. Rose (2004), Do We Really Know that the WTO Increases Trade?, in: The American Economic Review 94 (2004), No. 1 (March), 98-114. Jeffrey D. Sachs (2001), Tropical Underdevelopment, NBER Working Paper, No. W8119, February 2001. Jeffrey D. Sachs / Andrew Warner (1995), Economic Reform and the Process of Global Integration, Brookings Papers on Economic Activity 1995:1, 1-118. UNCTAD (2001), Duty and Quota Free Market Access for LDCs, UNCTAD and the Commonwealth Secretariat, Geneva 2001. UNCTAD (2003), Back to the Basics: Market Access Issues in the Doha Agenda, United Nations Conference on Trade and Development, New York / Geneva (United Nations) 2003.

312

Fritz Breuss

Athanasios Vamvakidis (1998), Regional Integration and Economic Growth, in: The World Bank Economic Review 12 (1998), No. 2, 251-270. John Whalley (1985), Trade Liberalization among Major World Trading Areas, Cambridge/MA / London (The MIT Press) 1985. World Bank (1993), Report on Structural Adjustment Lending III, Washington/D.C. (The World Bank) 1993. World Bank (2001), Global Economic Prospects and the Developing Countries, Washington/D.C. (The World Bank) 2001. World Bank (2002), Globalization, Growth and Poverty: Building an Inclusive World Economy, A World Bank Policy Research Report, Washington/D.C. / Oxford (The World Bank and Oxford University Press) 2002. WTO (2001), Ministerial Declaration, Ministerial Conference, Fourth Session, Doha, 9-14 November 2001, WT/MIN(01)/ DEC/W/1, 14 November 2001. WTO (2002), Market Access: Unfinished Business – Post-Uruguay Round Inventory and Issues, Special Studies No. 6, Geneva (World Trade Organization) 2002. WTO (2003), World Trade Report 2003, Geneva (World Trade Organization) 2003. WTO (2004), Doha Work Programme: Decision Adopted by the General Council on 1 Ausut 2004, Geneva (World Trade Organization), WT/L/579, 2. August 2004.

Does the ‘Development Round’ Foster Development?

313

Appendix: Tables and Figure on World Trade

Table A1: World Merchandise Exports by Product, 2002 (Billion US$ and percentage) Value 2002 6,272 583 468 114 788 63 615 110 4,708 142 660 460 2,539

Share 1995 100.0 11.7 9.0 2.7 10.7 1.2 7.3 2.2 74.3 3.1 9.7 7.9 38.8

2002 100.0 9.3 7.5 1.8 12.6 1.0 9.8 1.8 75.1 2.3 10.5 7.3 40.5

9.2 12.1

9.9 13.4

17.5

17.2

All productsi) Agricultural products Food Raw materials Mining products Ores and other minerals Fuels Non-ferrous metals Manufactures Iron and steel Chemicals Other semi-manufactures Machinery and transport equipment Automotive products 621 Office and telecom 838 equipment Other machinery and 1,080 transport equipment Textiles 152 Clothing 201 Other consumer goods 553 i) Includes unspecified products. They account for merchandise exports in 2002.

3.0 2.4 3.2 3.2 8.7 8.8 3.3 percent of world

Table A2: World Merchandise Exports by Region, 2002 (Billion US$ and percentage)

World North America United States Latin America Mexico Western Europe EU (15)

Value 2002 6,272 948 694 350 161 2,657 2,449

1990 100.0 15.4 11.6 4.3 1.2 48.3 44.4

Share 1995 100.0 15.5 11.7 4.6 1.6 44.8 41.5

2002 100.0 16.9 12.5 5.8 2.7 40.0 37.0

314

Fritz Breuss Value 2002 314

1990 3.1

Share 1995 3.9

2002 C./E. Europe/Baltic 4.3 States/CIS C./E. Europe 148 1.4 1.6 1.9 Russian Federation 107 1.6 1.7 Africa 140 3.1 2.2 2.3 South Africa 30 0.7 0.6 0.5 Middle East 244 4.1 3.0 4.3 Asia 1,620 21.8 26.0 26.4 Japan 417 8.5 8.8 7.6 China 326 1.8 3.0 4.0 6 East Asian traders 603 7.8 10.3 10.4 Memorandum item: NAFTA (3) 1,107 16.5 17.1 19.5 MERCOSUR (4) 89 1.4 1.4 1.4 ASEAN (10) 405 4.2 6.4 6.8 Source (A1 + A2): WTO, International Trade Statistics 2003, available at http://www.wto.org/english/res_e/statis_e/its2003_e/its03_bysubject_e.htm#sector

80 70 60 50 40 30 20 SOUTH

10

EAST

0 NORTH

NORTH EAST

SOUTH

Figure A1: Network of world merchandise trade, 2002 (in percent of world trade) North = North America, Western Europe, Japan, New Zealand; East = Central and Eastern Europe, Baltic States, Russian Federation; South = Latin America, Africa, Middle East, Other Asia.

Alan Matthews*

Agriculture after Cancún** I. Introduction II. The Road to Cancún III. Analysis of the Key Issues A. Market Access B. Export Subsidies C. Domestic Support IV. Evaluation of the State of Play of the Negotiations V. After Cancún A. The Regionalism Option B. Significance of the Peace Clause C. Moving towards Geneva References

315 317 320 320 324 326 328 331 331 332 334 336

I. Introduction WTO Members met for the Fifth Ministerial Conference at Cancún in September 2003 to undertake a mid-term review of the Doha round of trade negotiations. Agriculture is a key element of these negotiations as mandated in the Declaration launching the Doha Round: ‘Building on the work carried out to date and without prejudging the outcome of the negotiations we commit ourselves to comprehensive negotiations aimed at: substantial improvements in market access; reductions of, with a view to phasing out, all forms of export subsidies; and substantial reductions in trade-distorting domestic support. We agree that special and differential treatment *

**

Alan Matthews is Jean Monnet Professor of European Agricultural Policy Trinity College Dublin, Ireland. Correspondence to [email protected]. Revised Paper presented to the International Conference ‘The future of the World Trading System after the Failure at Cancún’, Vienna, Austria, November 20 and 21, 2003.

316

Alan Matthews

for developing countries shall be an integral part of all elements of the negotiations and shall be embodied in the Schedules of concessions and commitments and as appropriate in the rules and disciplines to be negotiated, so as to be operationally effective and to enable developing countries to effectively take account of their development needs, including food security and rural development. We take note of the non-trade concerns reflected in the negotiating proposals submitted by Members and confirm that non-trade concerns will be taken into account in the negotiations as provided for in the Agreement on Agriculture’.1 The failure of the Cancún Ministerial to reach agreement on a framework text on negotiating modalities for the market access agenda and on the extent of new rule-making left the agricultural negotiations, as with the other areas under negotiation, in suspension. A meeting of the WTO General Council on 15 December 2003 was unable to provide the necessary momentum to move towards a successful and timely conclusion of the negotiations. However, although there is little sign at the beginning of 2004 that countries are yet ready to re-engage in meaningful negotiations, it is too early to write off the Round as dead. Agriculture is one of the make-or-break issues in the Doha Round. Thus it is timely to review what progress was made, if any, in the discussions on agriculture in the run-up to Cancún and whether the shape of a final agreement can yet be discerned. This chapter addresses this question through an analysis of the successive drafts of the framework text on the modalities of the agricultural negotiations and a comparison of the positions of the major players in these negotiations.2 Particular attention is paid to the position of the developing countries and the way their concerns have been addressed in successive drafts. The paper is upbeat about the prospects of reaching an agriculture agreement. It argues that the main constraint to a successful resumption and conclusion of the Doha Round lies in broader political calculations in areas other than

1 2

WTO (2001). A similar exercise undertaken by the International Centre for Trade and Sustainable Development only came to my attention after the completion of this chapter (ICSTD (2003)). See also the agricultural backgrounder regularly prepared by the WTO Secretariat and updated 1 March 2004 (WTO (2004)).

Agriculture after Cancún

317

agriculture, but that progress in agriculture could greatly assist in improving the overall prospects for success. II. The Road to Cancún The first draft text on modalities was circulated to WTO Members by Stuart Harbinson, Chairman of the Special Session on Agriculture, on 12 February 2003. A mini-ministerial meeting in Tokyo later that month was the first opportunity for some of the WTO membership to comment on the paper. The proposal failed to get much support and, indeed, no country strongly supported the first draft. Members could not even agree on a term to describe the draft (whether as a starting point, reference or basis for the negotiations), eventually agreeing to call it a ‘catalyst’. Despite these disagreements, however, Harbinson’s second draft in the following month made few significant changes.3 A second mini-ministerial took place in Montreal on 28-30 July. Despite some limited signs of flexibility (for example, the EU increased its offer to cut trade-distorting support from 55% to 60%, while expressing its willingness to eliminate export subsidies and expand tariff rate quotas for an agreed list of products) the meeting failed to give a political momentum to the negotiations. Following that meeting, the US and the EU were asked to work together to break the deadlock. In August 2003, the EU and the US issued a joint framework suggesting some areas of agreement on how they thought progress could be made. Among other things, this text proposed a blended formula for tariff reductions, a special safeguard for developing countries to protect sensitive products and improved duty-free access for developing country exports. Many of the proposals from this paper were included in the framework text on agriculture contained in Annex A of the Draft Ministerial Text prepared by the WTO Secretariat and published on 31 August as the basis for dis-

3

This is not surprising. A statistical analysis by the Danish Research Institute of Food Economics (2003) which measured the distance between the negotiating proposals of each country showed that the Harbinson draft, if it were considered as the negotiating proposal of a separate ‘country’, was located in the middle of all other major proposals.

318

Alan Matthews

cussion at the Cancún meeting (called the Castillo draft after the Chairman of the General Council).4 Developing countries felt that this text leant excessively in the direction of developed country interests and it sparked a series of counter-proposals. A group of (initially) 21 developing countries, including Brazil, China, India and South Africa (now usually referred to as the G20 as its membership varied both during and after the Cancún meeting), put forward an alternative text,5 followed by the African and least developed countries, which proposed another.6 Four African cotton-producing countries pursued a Sectoral Initiative on Cotton demanding the removal of all cotton subsidies and financial compensation while the subsidies still existed. A revised framework text on agriculture was presented to the Conference by the Secretariat on 13 September (called the Derbez draft after the Mexican Foreign Minister who chaired the conference).7 Although the negotiations on this text broke down the following day because of incompatible positions on the Singapore issues,8 there was some optimism that a deal could be reached on agriculture on the basis of this draft. The manner in which it addressed the concerns of the major participating groups is discussed in the following section of the paper. The Cancún conference was notable for the capacity of developing countries to organise around common positions. While the Uruguay Round negotiations on agriculture were largely a US-EU affair, and were concluded once these countries reached the Blair House agreement, the Doha Round negotiations are more a NorthSouth issue although with sub-plots within each of these groupings. Already at the Doha Conference a loose alignment of developing countries calling themselves ‘Friends of the Development Box’ had 4 5 6 7 8

WTO (2003a). The G21 alternative text on agriculture can be found at www.icstd.org. WTO (2003b). WTO (2003c). The Singapore issues cover trade and investment, trade and competition policy, transparency in government procurement and measures to facilitate trade such as the simplification of customs procedures They are referred to as the Singapore issues because the Ministerial Declaration following the WTO Council meeting in Singapore in 1996 agreed to establish working groups to analyse them.

Agriculture after Cancún

319

been formed to promote acceptance of ideas for greater special and differential treatment for developing countries in the agriculture agreement. The G20 group was notable for bringing together countries which traditionally have held very divergent positions in the agricultural negotiations, particularly on the importance to be attached to the Development Box. While the G20 grouped the developing country heavyweights, the least developed countries, African, Caribbean and Pacific (ACP) and African Union countries came together in a group which was designated (by others if not by themselves, since the group contains only 61 WTO members, 2003) as the G90, led by Mauritius. This group, in addition to supporting the Development Box proposals, sought measures to address tariff peaks and escalation, the binding of preferential access to developed country markets, and a compensatory mechanism to compensate for the erosion of trade preferences due to tariff liberalisation. A final group of eventually 33 countries (dubbed the G33), led by Indonesia and the Philippines, formed the Alliance for Strategic Products and a Special Safeguard Mechanism, which emphasised the particular importance of strengthening measures to protect vulnerable farmers. Membership of these groups overlapped, and the fact that leading countries in each alliance are members of the Cairns Group of agricultural exporting countries is also significant. The Cairns Group, which had been dominated by Australia following a hardline free trade stance focusing on improved market access and the removal of subsidies, played a much more subdued role in Cancún as its developing country members followed a more developmentoriented path. CAFOD remarks on the fact that agriculture was the focus of all these developing country groupings, highlighting its central importance for them and the essential distraction which the Singapore issues represented.9 The cotton initiative proposed by four West and Central African LDCs – Benin, Burkina Faso, Chad and Mali – became the symbolic issue at Cancún, just as access to medicines at the Doha Conference. The African countries requested an ‘early harvest’ decision in Cancún to phase out all cotton subsidies and domestic support measures by 2005, along with the payment of compensation to LDCs during the transition period. The US and the EU were opposed to a decision of this kind, which they viewed as setting a 9

CAFOD (2003)

320

Alan Matthews

precedent for commitments on specific products, challenging the aggregate approach adopted for Amber Box trade-distorting subsidies in the AoA.10 A further feature of the Cancún conference was the role played by civil society organisations. Development NGOs turned up in strength and acted as a counterweight to the farm lobby representatives in influencing national delegations. The EU Agricultural Commissioner went so far as to partly blame their influence on developing country delegations for the breakdown of the talks. While the comment was seen as somewhat patronising with respect to developing country governments,11 it provides a further illustration of the changed dynamics of the conference negotiations. III. Analysis of the Key Issues This section of the paper traces how negotiating positions have been reflected in successive draft texts of the modalities, beginning with the Harbinson draft through the Castillo text at the outset of the Cancún conference and the Derbez text at the end of that conference. Areas of agreement as well as disagreement are highlighted in order to form a view on the degree of convergence which has been achieved in the negotiations to date. A. Market Access The US proposal on market access was to use a harmonising formula (the Swiss formula12) to reduce agricultural tariffs, ensuring that no individual tariff exceeds 25% after a five-year phase in period. Tariffs should be simplified to either ad valorem or specific, but not mixed. The EU, on the other hand, proposed to continue the Uruguay Round (UR) linear formula, reducing agricultural tariffs by 36% on average, with a cut of at least 15% per dutiable item. The Harbinson draft proposed a banded approach. Tariffs greater than 90% would be reduced by 60% on average, with a minimum cut of 45% per tariff line. Tariffs between 90 and 15% 10 11 12

The EU position has since become more supportive of the African countries’ position, see European Commission (2003). CAFOD (2003) The Swiss formula is Tn = (amax * T0)/(amax + T0) where T0 is the original tariff, Tn is the new tariff and amax is the upper bound on all resulting tariffs. With amax = 50, an initial tariff of 40 per cent would be reduced to 22 per cent.

Agriculture after Cancún

321

would be reduced by 50/30%, while tariffs lower than 15% would be reduced by 40/25%. Reductions would take place in equal instalments over five years. Developing countries would be given a ten year implementation period and lighter reduction commitments. Least developed countries would not be required to undertake reduction commitments but might be encouraged to do so on a voluntary basis, taking into account their development needs. Newly acceded countries (such as China), who often offered more ambitious tariff concessions than those required under the UR formula, would have the flexibility to defer the implementation period for new commitments by two years. The Harbinson formula was opposed both by the supporters of a Swiss formula approach (which would bring all tariffs down to a specific maximum level) and by those (such as the EU) which wanted to continue the flexibility to have smaller reductions for certain sensitive products. The joint EU-US framework proposal proposed a blended approach instead. Some proportion of tariff lines would be subject to a Swiss formula coefficient; some proportion would be subject to the UR formula of an average tariff cut subject to a minimum; and some proportion of tariff lines would be duty-free. Moreover, a maximum tariff level would be agreed; countries which wished to retain tariffs above this level would be required to offer effective additional market access through a request/offer process, including increases in tariff rate quotas. The G20 group accepted this blended approach although – for the UR element of the formula – they proposed a simple uniform tariff cut (instead of an average reduction subject to a minimum). However, for developing countries, they proposed the continuation of the UR formula across-the-board. They also called for an overall target for average tariff reductions by developed countries. Similar demands were included in the G90 proposal. The Castillo draft retained the blended formula approach based on the EU-US framework proposal, while adding the proviso that the resulting simple average tariff reduction for all agricultural products should be no less than a specified minimum as called for by the G20. It proposed to treat maximum tariffs as in the EU-US framework proposal, although the Derbez text offered a slight loophole that additional flexibility would be possible for a limited number of products on the basis of non-trade concerns. From the G20 perspective, an objectionable element of this text was that it re-

322

Alan Matthews

tained the use of the Swiss formula also for developing countries, implying a maximum tariff level for at least some products. Harbinson recognised the need for special and differential treatment for developing countries by proposing lower tariff reductions and longer implementation periods. However, both the Castillo and Derbez texts proposed that setting maximum tariffs also for developing countries should remain under negotiation, a proposal rejected by the developing countries. Reflecting the demand for a Development Box, Harbinson introduced the concept of ‘Special Products’ (SPs) which would be defined with respect to food security, rural development and/or livelihood security concerns. Harbinson suggested that the tariff reductions to be sought on SPs should be limited to an average of 10% (with a minimum reduction of 5%). The Castillo draft proposed a less attractive formulation from a developing country perspective, namely, a lower minimum reduction for SPs within the proportion of tariff lines which would be subject to the UR linear formula. Presumably, this would require compensating higher reductions on other tariff lines in order to achieve the overall targeted average reduction. The Castillo draft also proposed exempting SPs from any obligation to increase Tariff Rate Quotas (TRQs). The Derbez text added that, where existing tariff bindings were very low, no further reductions would be sought. Development Box proponents see these as very weak provisions to secure the protection they believe to be necessary for these particularly sensitive products. An important element of the market access proposals in the Harbinson draft proposals was a clear formula to address tariff escalation issues. Where the tariff on a processed product was higher than the tariff for the product in its primary form, the rate of tariff reduction for the processed product would be equivalent to that for the product in its primary form multiplied by, at a minimum, a factor of 1.3. The EU-US framework text did not address this issue, and in the Castillo draft, it was reduced to ‘best endeavour’ language, stating that ‘the issue of tariff escalation will be effectively addressed.’ The G20 draft proposed retaining the Harbinson formula, and it was reinstated in the Derbez text. The Harbinson draft proposed the elimination of the current special safeguard mechanism (SSG) which allows countries to levy an additional, time-limited import surcharge to protect domestic producers from a sudden surge in imports of certain products. This mechanism is only available to products which underwent tariffica-

Agriculture after Cancún

323

tion in the UR and where countries had reserved the right to use it in their schedules. In the EU-US framework text, the use of the SSG was left as a matter for further negotiation and this language was retained in the Castillo and Derbez texts. The G20 draft calls for the conditions and timetable for its elimination to be negotiated. The Harbinson modalities also included provision for a new Special Safeguard Mechanism (SSM) to be added to the modalities to enable developing countries to effectively take account of their development needs. The SSM is strongly supported by both the G20, the G90 and the G33 ‘subject to conditions and for products to be determined’. The EU-US framework draft also accepted a SSM but linked it to the notion of ‘import-sensitive products’ which would also benefit from lower tariff reduction commitments. Both the Castillo and Derbez texts contained a proposal to establish a SSM along the lines of the G20 proposal. Other developing countries in the G33, however, want the SSM to be available for all products, as well as ensuring that SPs would also have access to the SSM. On TRQs, the Harbinson text proposed that quota amounts would be increased up to 10% of current domestic consumption of each product, as well as calling for simplification and greater transparency in the administration of tariff rate quotas. It did not propose a reduction of in-quota tariffs, except for tropical products or products which might substitute for crops which were illegal or harmful to human health, or where fill rates over a recent period had been particularly low. The EU-US proposal did not address this issue, although it implied that TRQs would be increased in cases where countries made use of the flexibilities under the tariff reduction formula to maintain particularly high tariffs or to implement minimum reductions for import-sensitive products. Taking its cue from this draft, the Castillo and Derbez texts proposed a reduction of inquota tariffs while leaving the terms and conditions of further TRQ expansion under negotiation. The G20 text supports the Harbinson draft in calling for an expansion of TRQs, while the G90 text contents itself with a call for simplification and greater transparency of TRQ regimes. The erosion of trade preferences following from any reduction in agricultural tariffs in developed countries has emerged as an important issue for a number of developing countries. The Harbinson draft modalities called for the maintenance of the nominal margins of tariff preferences where technically feasible – this is obviously

324

Alan Matthews

not possible in cases where duty-free access is already granted. It proposed a longer implementation period and a two year moratorium for ‘tariff reductions affecting long-standing preferences in respect of products which are of vital export importance for developing country beneficiaries …’ How this would work in practice is not clear, as eligible products are defined as those which account for at least 20% of the total merchandise exports of any beneficiary. This would seem to open up a potentially large number of products for which this special exception would apply. The formulation is supported in both the Castillo and Derbez texts, as well as in the proposals from the G20. The G90, in addition, seek the development of a compensatory mechanism to address the erosion of preferences for countries adversely affected. The EU-US framework text addressed this issue by calling for duty-free access for a certain percentage of developing country imports to be provided ‘through a combination of Most Favoured Nation and preferential access’.13 This formulation was carried into the Castillo and Derbez texts, although using the ‘best endeavour’ language that all developed countries ‘will seek to provide’ this minimum amount of duty-free access. The G20 text strengthens this to the mandatory ‘shall provide’, with the implication that, where provided through preferential schemes, this access would be bound in the schedules of commitments entered by developed countries. The Harbinson draft modalities, in addition, had offered that developed countries either ‘should’ or (much stronger) ‘shall’ provide dutyand quota-free access to their markets for all agricultural imports from the least developed countries. This was weakened to ‘best endeavour’ language in the Castillo text that the objective of dutyand quota-free access for least developed countries ‘shall be expeditiously pursued’. Both the G20 and the G90 sought the stronger Harbinson alternative and, in the Derbez text, the Harbinson dual formulation was restored. B. Export Subsidies As was evident already in negotiating the Doha mandate in November 2001, the future of export subsidies is a hotly contested issue. The US, Cairns Group and developing countries have called 13

In its original negotiating proposal to the Special Session, the EU proposed zero duty access for 50% of total imports of agricultural products from developing countries, as well as unrestricted access for imports from the least developed countries.

Agriculture after Cancún

325

for their elimination. The EU is the single largest user of export subsidies, accounting for around 85% of the total by value notified to the WTO Committee on Agriculture. It had originally proposed a 45% reduction in expenditure on average, provided that all forms of export subsidies, including export credits and the activities of state trading exporters, were treated equally. The Harbinson draft proposed that export subsidies for products accounting for 50% of export subsidy expenditure would be phased out over five years, while the remainder would be phased out over nine years, with the greatest reductions occurring in the earlier years. Developing countries would be given 10 and 12 years respectively. In the EU-US framework text, the EU offered to eliminate export subsidies on certain products of particular interest to developing countries (to be determined) while subsidies for the remaining products would only be reduced. The Harbinson text had also proposed to discipline export credits. Non-conforming export financing subsidies would be subject to reduction commitments, with exceptions for emergency situations in importing Members. The EU-US framework text proposed that parallel disciplines should be applied to export credits, state trading export enterprises and food aid programmes. It also added that, without prejudging the outcome of the negotiations, reductions of, with a view to phasing out, all forms of export subsidies including export credits would occur in a parallel manner. The Castillo text combined these elements by affirming that reduction commitments shall be applied in a parallel manner to export subsidies, export credits, export state trading enterprises and food aid. It retained the commitment to phase out export subsidies for products of interest to developing countries, but added that, for the remaining products, ‘Members shall commit to reduce, with a view to phasing out, budgetary and quantity allowances for export subsidies’. The G20 draft proposed adding a definite time horizon to this last commitment, turning it from a ‘best endeavour’ to a mandated commitment, but the Derbez draft retained the Castillo wording. However, the Castillo draft contained a commitment (para. 3.6) that ‘the question of the end date for phasing out all forms of export subsidies remains under negotiation’. This was further strengthened in the Derbez draft to become ‘An end date for

326

Alan Matthews

phasing out all forms of export subsidies remains under negotiation’, thus changing the issue from ‘if’ to ‘when’.14 C. Domestic Support While high-support countries are broadly satisfied with the Green Box as defined in the URAA, intended to permit countries to continue support programmes to agriculture which are deemed to be not or minimally trade-distorting, agricultural exporters and developing countries have sought to limit Green Box payments, arguing that their sheer volume has a distorting effect on global trade. The Harbinson draft modalities did propose some minor tightening of the criteria for eligibility for the Green Box in Annex 2 of the URAA but did not propose a cap on Green Box spending or remove any income support payments from the Green Box. At the same time, it responded to the demands from the EU, Switzerland and others to allow payments for animal welfare programme to be classified as a Green Box measure. The Castillo text simply noted that the Green Box criteria remain under negotiation. This was slightly strengthened in the Derbez draft to state that Green Box criteria shall be reviewed with a view to ensuring that Green Box measures have no, or at most minimal, trade-distorting effects or effects on production. The G20 text altered this to propose that the Green Box disciplines should be strengthened with a view to ensuring that Green Box measures have no, or at most minimal, trade-distorting effects or effects on production. Developing country interest also focused on Article 6.2 measures to encourage agricultural and rural development which are exempted from reduction commitments under this paragraph. The Harbinson draft modalities proposed extending the range of exempted measures and attached a list of possible amendments for further consideration. Surprisingly, this issue was not explicitly included in either the Castillo or Derbez texts nor in the G20 draft. With respect to trade-distorting (Amber Box) support, the Harbinson draft proposed that the Aggregate Measurement of Support (AMS) should be decreased by 60% in five years for developed countries, and 40% in ten years for developing country members. The EU-US draft did not contain numbers for the extent of the reduction in trade-distorting support, but it did state that the reduc14

CAFOD (2003)

Agriculture after Cancún

327

tions would be ‘significantly larger than in the Uruguay Round’ (namely, 20% for developed countries). It also stated that Members with higher levels of trade-distorting subsidies would have to make greater efforts than others. The unspecified AMS reduction was carried into the Castillo draft, but a significant addition in the Derbez text was the commitment that product-specific AMS shall be capped at their respective average levels during a base period to be agreed. The G20 proposed, in addition, that the product-specific ceilings might be reduced in further negotiations and that, for heavily exported products, additional disciplines should be negotiated. The EU-US text also proposed a reduction in de minimis limits which was repeated in both the Castillo and Derbez drafts, although developing countries would be exempt from this reduction. The Uruguay Round Agreement introduced the Blue Box category of production-limited support. The Harbinson draft proposed to cap these payments at their 1991-2001 level and reduce them by 50% over five years. Alternatively, it suggested that the Blue Box should be merged with the Amber Box. The US had initially proposed to scrap the Blue Box entirely, but the EU-US framework text proposed to cap Blue Box spending at 5% of the total value of agricultural production at the end of the implementation period. It also proposed to relax the criteria for Blue Box payments by no longer requiring production limiting or supply management programmes. Under this new design, the US could seek to exempt its countercyclical payments under the 2002 Farm Act from its reduction commitments. The EU-US text also proposed a reduction in de minimis limits, as well as introducing a new ceiling where the sum of Amber Box, Blue Box and de minimis payments would be significantly reduced below the combined sum of these payments in 2004. This formulation was carried over into the Castillo draft (with the total value of agricultural production fixed at the level of the 2000-2002 period) with the cap to be achieved by a specified date and with the proviso that, following that date, such support would be subject to a further linear reduction for an additional period of years. The Derbez text added that there should be a specified cut in the first year of implementation in the total sum of support under the Amber Box, Blue Box and de minimis ceilings. The G20 draft sought stronger disciplines on Blue Box support, seeking to cap it at 2.5% of the value of production in the initial implementation phase

328

Alan Matthews

and proposing that the subsequent linear cuts in Blue Box payments would continue ‘with a view to its phasing-out’. IV. Evaluation of the State of Play of the Negotiations Despite the failure at Cancún, the positive movement in the negotiations on agriculture to date should be noted. With respect to market access issues, the tariff reductions on offer in the Derbez text are potentially larger than those agreed in the Uruguay Round.15 There is the prospect that tariff peaks would be addressed. A clear formula is on offer to address tariff escalation. Moreover, the principle of special products and a special safeguard mechanism for developing countries has been accepted by the developed countries, there is the possibility of mandatory binding of duty- and quota-free access for the least developed countries and acceptance of minimum levels of duty-free access for products from developing countries. Compared to the Uruguay Round Agreement, there has also been significant progress in the area of export subsidies. An offer to phase out export subsidies on a list of products of interest to developing countries has been made, as well as a commitment to significantly reduce remaining export subsidies in parallel with disciplines on other forms of export competition. The Derbez text also appears to move forward from the Doha mandate in requiring members to negotiate a final date for the end of all export subsidies, although the EU would still have difficulty in agreeing to this.16 Finally, much larger reductions in trade-distorting support, including Amber Box and de minimis payments, are now on offer than in the Uruguay Round. Furthermore, a cap will be placed on Blue Box payments. Special and differential treatment would apply to developing countries, including lower reduction commitments for Amber Box support, longer implementation periods and enhanced provisions under Article 6.2.

15

16

This partly depends on whether the reductions are compared in absolute or in proportionate terms. To achieve the same absolute reduction as the 36% reduction agreed in the UR would require a cut of 36/64 or 56% in the remaining tariffs. Any percentage average reduction less than 56% would imply that the absolute cut in tariffs would be smaller than in the UR. CAFOD (2003)

Agriculture after Cancún

329

Despite this evidence of convergence in positions, significant differences remain. In the market access negotiations, three conditions raised by the developed countries could still cause difficulties. First, the EU is adamant that, while developing countries should continue to receive special and preferential treatment, this should not imply a different set of rules, but different implementation periods and extent of commitments based on the same rules. The EU does not agree to the blended formula approach (including applying the Swiss formula to a certain percentage of tariff lines) for developed countries and the Uruguay Round approach alone for developing countries. A second potential stumbling block is the insistence of both the EU and the US, in their framework text, that the rules and disciplines for special and differential treatment should be adjusted for significant net-exporting developing countries. This idea of differentiation (‘graduation’) – which is also promoted by the EU and others across all areas of negotiation in the Doha Round – is very opposed by the developing countries. However, the EU has highlighted that much of the benefits of the Round lie in the potential for increased South-South trade which is currently restricted by high developing country tariffs. It is also proposing that the more advanced developing countries should commit to provide duty-free and quota-free access for the agricultural exports of the LDCs, partly as a way of compensating for the erosion of their preferential access to developed country markets. Third, the EU has made very little progress to date on its demands on non-trade concerns such as consumer protection, food safety, extension of Geographical Indications, the environment and animal welfare. The Harbinson text merely called for further consideration. The EU-US framework text simply noted that these were issues of interest that were not yet agreed, and this language has been carried into the Castillo and Derbez drafts as well. From the developing country perspective, the blended formula approach to tariff reductions may not provide the sharp reductions in developed country tariffs they want to see. For example, a characteristic of the EU’s bound tariffs is the great variance in individual rates. Large numbers of tariff lines have zero or relatively low MFN duties, but for a small number of tariff lines MFN duties are very high. By putting the relatively low tariffs into the grouping subject to the Swiss formula and a maximum tariff while dealing with the high tariff items through the linear Uruguay Round formula subject to a minimum reduction for sensitive commodities, the

330

Alan Matthews

EU could avoid making significant reductions in its highest tariff rates (although the EU-US text did provide for expanded TRQ access where Members took advantage of the minimum rate reductions). Many developing countries, including India, also argue that the current proposals on Special Products would need to be strengthened considerably before they are acceptable. The export subsidy area still has the potential to cause difficulties, particularly for the EU. However, it is my view that all those involved in EU agricultural policy now accept that export subsidies are an indefensible form of trade policy instrument; the question is not whether they should be eliminated, but when. The key issue for the EU is the question of parallelism. If sufficiently watertight commitments can be given to equally discipline export credits, food aid and the activities of single desk exporters, the EU would find it very difficult to resist pressure to agree a date for their elimination, and the only issue would be the timeframe for this to happen. In the area of domestic support, the contentious areas which remain are whether Blue Box payments should be further reduced and whether Green Box criteria should be further strengthened. Developing countries argue that the sheer scale of these payments distorts trade and have called for the elimination of the Blue Box and reductions in Green Box support. The EU insists that all forms of support are not equally trade distorting and that this distinction must be maintained. It is concerned that its recent efforts in the Mid-Term Review of the Agenda 2000 CAP reforms to shift spending from Blue Box to Green Box (decoupled) payments should not be undermined. It does not want to see the Blue Box ceiling further reduced or phased-out in the future, and it is concerned that the Green Box criteria should remain sufficiently flexible to allow continued support to EU farmers to enhance environmental, food quality or safety and animal welfare objectives. With sufficient movement on market access and export subsidies, it is unlikely that these differences on domestic support would prove a stumbling block to an ultimate agreement. Through the smokescreens of individual countries’ negotiating positions, the broad outlines of the deal are clear. Developed countries will commit to significant reductions in tariffs and trade-distorting domestic support payments and the elimination of export subsidies, in return for a commitment by the developing countries to accept some lowering of their bound tariffs, subject to a deal on Special Products where lesser commitments would apply, while

Agriculture after Cancún

331

least developed countries would be exempt from reduction commitments. The fact that, for most developing countries, bound tariffs are well in excess of applied tariffs so that this discipline will not have any immediate impact, will make it easier to accept. Various side-payments would need to be made to ensure the agreement of key players, e.g. addressing the EU’s concern over GIs, developing countries’ concerns over special safeguards and the concern of preference recipients over preference erosion. However, these would not alter the basic parameters of potentially a very farreaching and worthwhile agreement, also from the point of view of developing countries. V. After Cancún A. The Regionalism Option Following the Cancún talks breakdown, US Trade Representative Robert Zoellick famously attributed the breakdown to a divide between the ‘can do’ and ‘won’t do’ countries. ‘The rhetoric of ‘won’t do’ overwhelmed the concerted efforts of the ‘can do.’ he said. He subsequently developed this theme in a Financial Times article in which he suggested that the US would place more effort in trying to secure its trade objectives through bilateral or regional deals with the ‘can do’ countries if the multilateral process remained bogged down.17 The US is currently negotiating free trade area agreements with Morocco, Australia and the Central America Free Trade Agreement but these are hardly significant in global terms. As Lamy gleefully pointed out in a Washington speech in early November, ‘I am not going to lose any sleep about Bob Zoellick pursuing an FTA with Morocco, where the US starts from a tiny percentage of Morocco’s total trade … And did you know how many FTAs the US has? Six, including counting NAFTA as two. The EU … has got a fair few more than that …’.18 A number of Latin American countries, including Columbia and Peru, have since left the G20 group in order to be able to pursue bilateral agreements with the US, but the big prize for the latter is the Free Trade Area of the Americas (FTAA) that would give it greater access to markets in the Western Hemisphere. However, the US, backed by Canada, 17 18

Financial Times, 22 September 2003. Lamy (2003).

332

Alan Matthews

has insisted that it will only negotiate concessions on agricultural subsidies (and anti-dumping rules) – both key concerns of Latin American countries – in the Doha Round and not in the FTAA.19 On the other hand, US demands that the talks should cover issues such as rules on services, investment, government procurement and intellectual property have been resisted by Brazil. Both sides appeared to set a fairly low threshold for success at their ministerial conference in Miami in mid-November.20 For the EU, with its extensive network of regional arrangements, using the threat of regionalism to try to make progress on its multilateral agenda is less convincing. However, it continues to extend this network (for example, most recently with Mexico) and has also been pursuing talks with MERCOSUR. Lamy has argued that, if the multilateral process fails, the EU would look again at pursuing the bilateral option more actively but he does not see this as undermining the multilateral approach. ‘For the foreseeable future, we in Europe, you in the US and, increasingly, they in Asia, are going to pursue a mix of multilateral and bilateral trade agreements. I think this is healthy’.21 The Achilles heel of any attempt by the EU to wield the regionalism stick is, again, its reluctance to grant sufficient agricultural concessions to its negotiating partners to make such deals attractive. Regionalism will not prove an attractive option to the major developing countries as an alternative to reviving multilateral negotiations. It would seem strangely ironic if these countries were individually to accept in bilateral deals what they have collectively rejected in the multilateral process. B. Significance of the Peace Clause A more attractive option for developing countries to pursue their objectives would be to use their existing rights under the WTO Agreements to attack developed country agricultural subsidies. Currently, Article 13 of the AoA, known as the ‘Peace Clause’ puts some restraints on the exercise of the normal rights which WTO Members have to react against the unfair effects of other countries’ subsidies. Under Articles XVI and VI of GATT 1994 and the Agreement on Subsidies and Countervailing Measures, two types of remedies are available against subsidies: 19 20 21

Financial Times, 24 September 2003. Financial Times, 15 November 2003. Lamy (2003).

Agriculture after Cancún

333

x

imposition of countervailing duties, if there is material injury or the threat of material injury to domestic production resulting from these subsidies; x making use of the dispute settlement mechanism against the subsidising Member if, as a result of the subsidy, there is either serious prejudice to the domestic industry, material injury or threat of injury to the domestic industry, or nullification or impairment of benefits accruing to that Member even where the behaviour of the subsidising Member does not infringe a GATT provision (these latter are called non-violation complaints). The Peace Clause limits the right of countries to impose countervailing duties or initiate dispute settlement proceedings. In the case of domestic subsidies, a distinction is drawn between Green Box and trade-distorting subsidies. Green Box subsidies are exempted from countervailing duty action and the dispute settlement process. In the case of trade-distorting subsidies, due restraint should be shown in initiating countervailing duty investigations, the dispute settlement process cannot be activated provided that the measures do not grant support to a commodity in excess of the level during the 1992 marketing year and the measures are also exempted from non-violation complaints. Conforming export subsidies have the protection that due restraint should be shown in initiating countervailing duty investigations, and action through the dispute settlement process cannot be taken. These provisions are valid for the ‘implementation period’ which is defined in Article 1(f) of the AoA as the nine-year period commencing in 1995. Thus the Peace Clause protections are set to expire on 1 January 2004.22 The implications of this expiry is that countries aggrieved by, for example, the EU’s use of export subsidies or the US use of domestic subsidies could challenge these in the dispute mechanism process. Developing countries and other agricultural exporters, if frustrated by the apparently slow progress in liberalising farm trade through negotiation, will be tempted to seek more rapid progress towards this objective through the dispute resolution mechanism. For example, Australia, Brazil and Thailand 22

Some have tried to interpret the ‘beginning of the implementation period’ as meaning the time when their implementation commitments began, which in the case of the EU would be 1 July rather than 1 January.

334

Alan Matthews

have already initiated cases against the EU and US on cotton and sugar. They may indeed be successful in this objective, although there is a serious risk that the ensuing series of trade disputes would lead to a further poisoning of the world trade climate.23 Countries, such as the EU, which have benefited from the Peace Clause will insist that its renewal is a precondition for continuing negotiations. The EU has warned that exporting countries face a mutually exclusive choice between achieving their objectives either through litigation or through multilateral negotiation. However, it appears to have no real leverage to secure its continuation, as it is hard to see what they might offer to developing countries, in particular, who have relatively few subsidies and thus have little interest in allowing an extension of the Clause. It hardly seems credible that the EU, having invested so much in the Doha Development Round, would be willing to walk away from further negotiations if it did not get its way on this issue, particularly as it would take at least two years before the consequences of any successful complaint against it would be felt.24 However, even if there is no formal waiver or moratorium, there could be a tacit agreement to show restraint if there was clear evidence that the developed countries were prepared to move closer to developing country demands on agricultural issues.25 C. Moving towards Geneva The meeting of the WTO Council in Geneva on 15 December 2003 was mandated to take stock of the fallout from the failure of the Cancún meeting, and ‘to take the action necessary at that stage to enable us to move towards a successful and timely conclusion of the negotiations’.26 A number of countries, such as the APEC countries (including the US), and those of the G20 meeting in Buenos Aires in early October 2003, had indicated that they were willing to 23

24

25 26

For a view on the prospects for bringing a successful challenge to agricultural subsidies under the GATT 1994 and the Agreement on Agriculture following the expiry of the Peace Clause, see Steinberg / Josling (2003). This includes the time for the consultations, panel hearing and report, appeal to the Appellate Body and the time allowed to a country to bring its laws and regulations into compliance. Bridges Vol. 7, No. 31, 25 September 2003. WTO (2003d).

Agriculture after Cancún

335

resume negotiations on the basis of the Derbez text (even though the latter roundly rejected this text at the Cancún meeting).27 The EU reacted more cautiously, with Lamy expressing amazement at ‘the amazing race over the last week to endorse the text produced in Cancún on 14 September’.28 Noting that many WTO members had attacked the document in Cancún, he said: ‘I am left to wonder, rather, what magic dust has been shaken over a text so roundly rejected in September, to find it so roundly endorsed in October’.29 However, he also stated that he thought the Cancún text was a ‘pretty good effort even if it caused us some real grief on agriculture’.30 Subsequently, it has been reported that the EU and the US are prepared to use the Derbez text as the basis for further discussions.31 While the outline of a potential deal on agriculture can be discerned, the way forward for the resumption of the Doha Round negotiations remains unclear. Removal of the roadblock caused by the Singapore issues (the EU, at least, has now indicated that it is prepared to adopt a more flexible approach on these issues)32 would certainly facilitate progress in the agricultural negotiations. More broadly, the frustrations of the developing countries that their concerns, on implementation and on improving special and differential treatment provisions, have not been addressed in the period since Doha despite the promises on that occasion, will also need to be tackled.33 The cotton issue remains important in this regard. But there remains a sense that success might be elusive. Following the failure of the Seattle WTO Ministerial Council to launch a comprehensive round of trade negotiations, success at Doha in 2001 was achieved in part because of its proximity to the terrible 27

28 29 30 31 32 33

Financial Times, 22 October 2003. There is speculation that the apparent softening in the position of the G20 group may reflect the haemorrhaging of members since the Cancún meeting. Only 12 countries signed the political declaration at the end of the Buenos Aires meeting, see Bridges Vol. 7, No. 34, 15 October 2003. Lamy (2003). Ibidem. Ibidem. Bridges Vol. 7, No. 34, 15 October 2003. See European Commission (2003). Chetaille / Tavernier (2003).

336

Alan Matthews

events of 11 September that year and the desire of countries to demonstrate solidarity at that time. Business pressures in the major countries, and the pressures of exporting interests in particular, have appeared much more ambivalent about the importance of successfully concluding a new round than was the case in the Uruguay Round. With a more divided international climate in 2003 over policy towards Iraq, the necessary momentum behind the talks has been slow to build up. The US commitment to a successful round, which had never seemed wholehearted, has cooled markedly as the US has been put on the defensive over the 2002 Farm Bill and cotton subsidies. The escalating number of US-EU trade disputes (steel, FSC, hormones, GMOs, the US Bio-Terrorism and Buy American Acts, the new EU chemicals policy), the US Presidential election in 2004 and the replacement of Pascal Lamy as the chief negotiator on the EU side when the new Commission takes office next year also add to the general uncertainty. Progress on agriculture may depend on whether an improvement in the overall negotiating climate can be achieved, but there is no doubt that progress on agriculture would also help greatly to bring about that needed improvement. Negotiations on agriculture are set to resume in March 2004. Although some Members have expressed the view that sufficient progress may be made to permit at least an informal ministerial by year end, it seems more likely that meaningful negotiations will not resume until the next Ministerial meeting planned for Hong Kong in 2005. Such a delay, while regrettable, would draw parallels with the Uruguay Round experience. If a good agreement on agriculture is attainable, it is worth waiting for.

References CAFOD (2003), The Cancun WTO Ministerial Meeting, September 2003: What happened? What does it mean for development?, CAFOD Policy Paper, UK (available at: www.cafod.org). Anne Chetaille / Karine Tavernier (2003), Failure of the Fifth WTO Ministerial Conference in Cancun: a Looming Crisis in the Multilateral Trade System?, Agritrade and Solgral, September 2003 (available at: http://agritrade.cta.int/cancun-analysisen3.pdf).

Agriculture after Cancún

337

Danish Research Institute of Food Economics (2003), Note on the Harbinson Draft on Modalities in the WTO Agriculture Negotiations, Copenhagen 2003. European Commission (2003), Reviving the DDA Negotiations – the EU Perspective, Communication from the Commission to the Council, to the European Parliament and to the Economic and Social Committee, Brussels, 26 November 2003. International Centre for Trade and Sustainable Development (ICTSD) (2003), Agriculture Negotiations at the WTO: PostCancún Outlook Report, Geneva, November 2003 (available at: http://www.ictsd.org/issarea/atsd/products/docs/Agriculture Negotiations9.pdf). Pascal Lamy (2003), Trade Crisis?, Speech delivered to the European Institute, Washington/D.C., 4 November 2003 (available at http://europa.eu.int/comm/commissioners/lamy/speeches_ articles/spla198_en.htm). Richard H. Steinberg / Timothy E. Josling (2003), When the Peace Ends: The Vulnerability of EC and US Agricultural Subsidies to WTO Legal Challenge, in: Journal of International Economic Law 6 (2003) No. 2, 369-417. WTO (2001), Ministerial Declaration, Ministerial Conference, Fourth Session, Doha, 9-14 November 2001, WT/MIN(01)/ DEC/W/1, 14 November 2001. WTO (2003a), Draft Cancún Ministerial text submitted by General Council chairperson Carlos Pérez del Castillo and DirectorGeneral Supachai Panitchpakdi 31 August 2003, JOB(03)/150/Rev.1. WTO (2003b), Consolidated African Union / ACP/ LDC Position On Agriculture: Communication from Mauritius, WT/MIN(03)/W/17. WTO (2003c) Draft Cancún Ministerial text submitted by Ministerial Council chairperson Luis Derbez 13 September 2003, JOB(03)/150/Rev.2. WTO (2003d), Ministerial Statement following the Cancun Ministerial Council (available at: http://www.wto.org/english/ thewto_e/minist_e/min03_e/min03_14sept_e.htm#statement). WTO (2004), WTO Agriculture Negotiations: The issues, and where we are now, Updated 1 March 2004 (available at: www.wto.org).

Markus F. Hofreither

Cancún and beyond A European Perspective of Agricultural Issues I. II.

Introduction Technicalities, Procedures and Tactics A. Official Reactions after Cancún B. Procedural Deficiencies and Tactics C. EU’s Agricultural Approach for Cancún III. Substance of Negotiations A. Past Experience and Current Problems B. CAP Reform Trends and Doha Development Agenda C. Approaching Cancún: the Cap Reform of June 2003 D. The Derbez Draft – a Promising Compromise? IV. Cancún and beyond V. Conclusion and Outlook References Appendix

339 342 342 344 346 348 348 349 353 356 359 363 364 368

I. Introduction The Fifth WTO Ministerial Conference has taken place in Cancún, Mexico, from 10 to 14 September 2003.1 In WTO’s own wording the primary objective of this meeting was ‘to take stock of progress in negotiations and other work under the Doha Development Agenda’. On Sunday, 14 September, the meeting ended without consensus on any of the items on the agenda.2 In agriculture, de1

2

The previous four WTO Ministerial Conferences took place in Singapore (1996), Geneva (1998), Seattle (1999) and Doha (2001). With negotiations still in progress, and some narrowing of the gap between the positions of countries apparent, the Mexican Chairman,

340

Markus F. Hofreither

spite some progress made, WTO Members were unable to adopt an already downsized modalities ‘framework’.3 No doubt, this meeting was a clear failure with respect to the objective of pushing forward the current Doha Round of trade negotiations, and thus ‘the second black eye for the WTO in four years’.4 Failures of important mid-term meetings in trade negotiations are not uncommon: during the Uruguay Round the meeting in Montreal, which was comparable with Cancún regarding its objectives, was a blunt failure. Two years later, the scheduled final meeting in Brussels brought no agreement at all. And, last but not least, the unfortunate occurrences associated with the 3rd WTO Ministerial conference in Seattle are still an issue. Yet, for the European Union (EU) the failure of Cancún must have represented a novel and probably surprising experience. During the Uruguay Round (UR) the Union – at least with respect to agriculture and until the scheduled final meeting in Brussels 1990 – has always been in a state which may be described as ill-prepared and rather unwilling. Indeed, the Brussels meeting failed mainly due to the inability of the EU to accept the Hellstrom compromise, although this proposal was very close to the final outcome three years later (‘Uruguay Round Agreement on Agriculture’, URAA). This conundrum is unravelled by the fact that the EU simply needed some extra time to rearrange its internal agricultural

3

4

Foreign Minister Luis Ernesto Derbez, closed the conference abruptly, by announcing that despite considerable movement in consultations 'the conference was at an end' when no deal could be struck on the Singapore issues. The official ministerial statement can be found on the WTO’s website (http://www.wto.org/english/thewto_e/ minist_e/min03_e/min03_14sept_e.htm). According to paragraph 14 of the Doha Declaration, Members were required to establish modalities by 31 March 2003. After missing this March deadline on agricultural modalities, chances were considered to be quite low that full agricultural modalities could be agreed in the Cancún Ministerial Conference. So, as a less ambitious alternative, negotiators tried to achieve at least a ‘framework’ for those modalities. Schott (2003), 2. While until the ‘fist black eye’, the disaster of the Seattle Ministerial of 1999, the WTO looked unstoppable, since then a continuous downsizing of expectations is recognized (Josling (2001)).

A European Perspective of Agricultural Issues

341

framework in a way, which subsequently paved the path to the Marrakech Accord 1994. With respect to the Doha Round - and particularly Cancún things have been different: in agriculture, the EU has put considerably more preparatory work into this round, has set up important agreements with the developing countries as well as the US in a timely fashion, and even reformed the Common Agricultural Policy (CAP) a few months before the Ministerial. As a result, compared to the UR the Union was in a much better negotiating position. Unfortunately, the outcome of Cancún was identical to Brussels 1990: the meeting was a failure, and some participants again put the blame for this outcome on the EU. However, playing the ‘blame game’ is a common ingredient of many negotiation processes and normally does not trigger enduring costs. The failure of Cancún, however, could bring about some real economic costs as well: as an immediate reaction important negotiating parties announced to reorient their priorities towards bilateral agreements. As a matter of fact, the US could try to deepen North American economic integration, and probably may look for agreements with Colombia, Peru, Thailand and others. The European Union could reinforce its activities with Mediterranean neighbours, and China could probably put more effort in a ‘North-East Asia FTA’.5 If these activities get momentum and end up with concrete deals, the failure of Cancún may bear long lasting costs in the form of rising trade diversion and discrimination through bilateral and regional agreements. Such a development would also make it more difficult to get back to serious multilateral negotiations in the future. Attempts to analyse and understand what happened in Cancún are undoubtedly necessary in order to facilitate a ‘successful and timely conclusion of the negotiations’, as required by the concluding Ministerial statement. Depending on the depth of the current crisis, such efforts could also become a highly rewarding undertaking with respect to the long-term development of the global economy. Basically, two main areas are considered with respect to the question of what may have caused the failure of Cancún: the first one focuses on the technicalities, procedures, and tactics of these 5

Schott (2003), 1 et seq.

342

Markus F. Hofreither

negotiations, while the second dwells on the substance of what has been discussed – or not discussed – between the groups engaged. This paper tries (1) to summarize and interpret both the procedural as well as the substantive issues of the negotiation process in agriculture, and (2) to draw some conclusions with respect to the future of these negotiations. Both tasks are accomplished with special emphasis on the viewpoints and positions of European agriculture. II. Technicalities, Procedures and Tactics A. Official Reactions after Cancún Immediate reactions after the closure of the Ministerial covered quite a broad range of opinions about the substance of the talks. In many cases deficiencies of the negotiation process or the strategies of the parties involved were mentioned: x Most probably aiming at the G20+,6 US Trade Representative Robert B. Zoellick criticized ‘certain developing countries’ for their inflexibility during negotiations and expressed his disappointment about the fact that the readiness of the US to negotiate on a wide range of issues was blocked by countries who seemed more interested in ‘tactical rhetoric’ than concrete progress. Zoellick also emphasized that henceforth, in the absence of multilateral movement forward, the US would engage in bilateral and regional trade agreements. x Speaking for the G20+, Brazil, Argentina, South Africa, Ecuador and Egypt also considered the failure of Cancún as a setback, but refused the notion that the group's inflexibility in agriculture was the main reason for bringing the negotiations to a halt. For them, the current situation must be considered as the beginning of a new phase. They emphasized that Cancún helped to solidify the group and thus made it a serious player 6

This paper uses the label ‘G20+’ (Group of 20-Plus) for a new coalition of developing countries, which together represent more than half of the global population. After the dropping out of Peru and Columbia the G20+ consists of Argentina, Bolivia, Brazil, Chile, China, Costa Rica, Cuba, Ecuador, Egypt, Guatemala, India, Indonesia, Mexico, Nigeria, Pakistan, Paraguay, the Philippines, South Africa, Thailand and Venezuela. The enormous weight of this group is partly attenuated by heterogeneous interests among its members with respect to important areas (e.g. agriculture).

A European Perspective of Agricultural Issues

x

x

x

7

8

343

not only in the area of agriculture. This would improve the chance to better incorporate the interests of a large part of the population of developing countries into the Doha Round. The ACP-AU-LDC Group,7 an alliance of poorest countries, mainly pointed to the Singapore issues as the main stumbling block. Due to their opposition to launching negotiations on any of the Singapore issues they would have had no option but to reject the EU's offer to drop only two of them. Moreover, these countries expressed their dissatisfaction that the negotiations never got to seriously deal with their priorities (agriculture, non-agricultural market access, S&D, and cotton). Compared to its impact during the UR the Cairns Group8 was relatively quiet in Cancún. Probably weakened through the exit of some weighty developing countries, which moved over to the G20+, after Cancún only Australia published a statement expressing its disappointment over a missed opportunity for the farmers of the world. For the EU, Trade Commissioner Pascal Lamy diagnosed the Doha Round as definitely being ‘in intensive care’. For him this outcome was both a severe blow for the WTO and a lost opportunity for developed and developing countries. Lamy mainly put the blame for the failure on the procedures and rules of the WTO by stating that the WTO would be ‘a medieval organisation’, which is not able ‘to steer discussions among 146 Members in a manner conducive to consensus.’ He also put across that in the near future the EU could take a more proactive stance.

This group unifies African, Caribbean and Pacific countries (ACP) as well as the African Union (AU) and Less Developed Countries (LDC) and comprises more than 90 countries. In a press statement of Australian Trade Minister Mark Vaile of 20 February 2004 (http://www.trademinister.gov.au/releases/2004/ mvt011_04.html) the members of the Cairns group were listed as Argentina*, Australia, Bolivia*, Brazil*, Canada, Chile*, Colombia, Costa Rica*, Guatemala*, Indonesia*, Malaysia, New Zealand, Paraguay*, Philippines*, South Africa*, Thailand* and Uruguay. However, during the Cancún Meeting the countries marked with an asterisk were counted as members of G20+.

344

Markus F. Hofreither

Yet, these official reactions have to be interpreted as what they genuinely are, namely playing the ‘blame game’: attempts to strategically put the blame for the collapse of Cancún on others. In fact, the unwillingness of the US to make concessions in the cotton case,9 the inability of the EU to go beyond the positions of the modest EU-US-proposal, or the static position of Japan with respect to its rice farmers, just to mention a few examples, have been important stumbling blocs for a swift progress. On the other side, parts of the developing world widely misjudged the objectives and possibilities of the Cancún meeting. So, in sum, nearly all participating groups have their share in making this meeting a failure. However, although it is important to understand the reasons for such a failure, putting too much effort in playing the ‘blame game’ sharply reduces the chances for a fruitful continuation of trade negotiations, and then protectionism is a likely substitute.10 B. Procedural Deficiencies and Tactics Far in advance it was known that the Ministerial Meeting in Cancún will be a huge event, with about up to 10 000 participants expected. Yet, this challenge was mainly taken on through attempts to streamline the local organization, but far less, if at all, by thinking about the adequacy of the structures and procedures of the negotiation process itself.11 Many countries have complained about information deficits and a lack of time during the Cancún meeting.12 As an example,

9

10

11

12

The US may have underestimated the perception of damage done to a large numbers of very poor farmers in some of the sub-Saharan African countries through its subsidies to less than 20,000 US cotton farmers (Srinivasan 2003, 32). If the US actually goes back to bilateral trade deals it is unlikely that the EU will seek leadership in liberalizing trade within a multilateral framework, because without sufficient ‘external’ pressure domestic interest groups, particularly in agriculture, would obstruct such a course with all their power. ‘In a body of 148 in which consensus is the norm for decision making, a long agenda with items on many of which there is no widespread agreement is a prescription for failure.’ Srinivasan (2003), 37. One notable example is the complaint of some developing countries that they did not get the information about the EU’s willingness to

A European Perspective of Agricultural Issues

345

until the morning of the last day of the conference, there was no substantial discussion of the highly controversial Singapore issues. However, depending on the capacities available, evaluating the consequences of new proposals and developing feasible counter proposals may take quite a lot of time, which further increases with the complexity of the issues discussed. Another reason for time pressure may have come from the changes with respect to the composition of negotiation parties before and even throughout Cancún: new coalitions emerged (G20+, ACP-AU-LDCs) and, partly caused by this, the composition of established groups changed.13 The necessary build-up of infrastructure and internal links within these coalitions clearly slows down the process of getting information across or finding out who is an ally with respect to a particular issue. A more general argument, which was mainly brought forward by developing countries in response to accusations they would behave hesitantly or at least seek too many tiered obligations,14 referred to the difference between the previous GATT framework and the WTO setting: The old GATT functioned with less frictions only because the obligations per se were not binding on everyone. So countries could join or not join agreements without losing their MFN rights. However, as all members must agree on all issues for the Doha Round to be successful (‘single undertaking’), all members have to be able to participate and understand the consequences before they are able to agree.15 Some observers suppose that the demanding position of developing countries may have been induced by the impression that they have not got a fair return out of the UR.16 The implicit emphasis of

13

14 15

16

drop some of the Singapore issues in the last stages of the negotiations in advance. The Cairns group, e.g., lost Argentina, Brazil, Chile, Columbia, Philippines through their move into the newly established G20+ group. European Commission (2003c), 2 et seq. Third World Network (2003). In recognizing this problem the EU has proposed switching to a mode of ‘optional participation’ with respect to the Singapore Issues (European Commission (2003g), 5). For a detailed and quite balanced analysis of this problem area please refer to Wang / Winters (1997).

346

Markus F. Hofreither

the Doha declaration on development, which was also corroborated by the rhetoric of many participants, may have induced them to expect that Cancún could function as a ‘donor conference’.17 In sharp contrast to such beliefs, the developed world arrived at Cancún with the firm intention to negotiate trade interests. Many participants put the blame for the surprising end of the Cancún negotiations on the inexperience of Chairman Derbez in handling such a big international conference. This argument explains the unsatisfactory end of this meeting with a simple accident. However, for quite a few observers it is difficult to believe that the host country Mexico would have acted without being sure of support from the United States.18 A general strategic problem of this Ministerial meeting has been the ‘contamination’ with geopolitical issues and interests which sometimes made international politics appear to dominate the genuine objective, namely trade negotiations. A more specific obstacle for the future continuation of ‘business as usual’ trade negotiations within the WTO could emanate from the - sometimes distinctly demonstrated - scepticism of important US representatives with regard to the efficacy of the current WTO procedures and processes.19 In sum, many participants agreed that some changes with respect to the structures and the procedures of the WTO would be necessary for a successful re-launch of the failed trade talks. C. EU’s Agricultural Approach for Cancún Focussing mainly on agricultural issues, EU Commissioner Franz Fischler stressed that he does not directly blame particular instances for the failure, but would like to attribute responsibility for this unsatisfactory outcome of Cancún to four more general factors:20 1. the specific approach taken by the Chairman, 17 18

19 20

Schott (2003), 5. In fact, there have been rumors of some hidden conversation between Mexican trade minister Derbez and US officials. The main argument in this respect was that with an unsuccessful end the United States – at least temporarily - would get rid of uncomfortable issues like cotton, where it was quite isolated at this time (Third World Network (2003)). Schott (2003), 3. Agra Facts (2003), 1 et seq.

A European Perspective of Agricultural Issues

347

2. 3.

the inflexible way of negotiating by specific countries, the negative impact of NGOs (e.g. ‘no deal is better than a bad deal’), and 4. a lack of proper preparation, where he obviously meant the WTO itself. Commissioner Fischler accentuated the fact that Cancún has not changed the attitude of the EU – or at least the Commission – with respect to the future course of negotiations, because ‘the principles have not changed’. He underlined this position by formally rejecting recent suggestions – mainly linked to statements of US officials – that bilateral agreements would be a viable alternative to the WTO process. The main strategic lapse of the EU may have been trying to negotiate – starting from a well prepared position – important details on a mainly technical level, without or at least too late recognizing that the state of information of a substantial number of its trading partners was insufficient and in a few cases even purposely biased. Understandably, in a situation with acute problems in defining and organizing its own position, the technical stance of the EU with ready-made proposals and positions made it appear as an opponent for many developing countries. In some cases this development may have been supported by blunt arguments of some NGOs, which recognized this situation as a good chance for getting broader media coverage.21 In putting together the evidence from a procedural viewpoint, the key drivers of this collapse quite certainly have been a political deadlock between important players, which could not be resolved in a few days and, in the wording of Pascal Lamy, the ‘medieval’ structure of negotiations imposed by the WTO.

21

Among other groups, ‘Food First’, an activist NGO, expressed the failure of Cancún as a ‘victory to the people’ (Economist, 20 September 2003). However, as many NGOs are organized as private enterprises it should not come as a surprise that they also behave as such in trying to serve their customers.

348

Markus F. Hofreither

III. Substance of Negotiations A. Past Experience and Current Problems So far, all previous rounds have been successful as in the end always the substance of the negotiations has prevailed over officially determined procedures. Negotiations have gone as long as required to obtain results, which are acceptable for all groups involved. The ultimate compromise across the differing interests often came in the form of a last-minute balancing of concessions after a long period of negotiations. The institutional design mainly determines the ease and thus the speed at which these results are obtained.22 With respect to the substance of this meeting it is apparent that things have changed to some extent. Since the creation of the GATT in the late 1940s, trade negotiations have been dominated by advanced industrialized countries. During this round, however, the centre of gravity has shifted towards the divide between North and South. Compared to narrowly defined quarrels about mainly technical aspects of trade policies between the big players of the developed country group, this new focus brings in a much broader problem set.23 There is also an increased emphasis on ‘behind the border’ measures, mainly through the Singapore issues, and as a completely new feature the challenge of reconciling the traditional areas of trade negotiations with the promotion of development.24 Depending on the boundaries of the eventual problem definition in the course of the current negotiations, the disputes could become an intense and broad battle about North-South conflicts. However, such a scenario would clearly overload the whole proc22

23

24

Technically speaking such negotiations are repeated games with many players. The negotiation process runs over an ex ante unknown time period which is only structured by a number of official stages. Information gathering and transmitting, coalition building, reputation and trust, but also history does matter in repeated games. Once negotiations have officially started, in the course of time there is considerably increasing pressure on negotiators to produce useful results (Hoekman / Kostecki (1995), 79 et seq.). The politically agreed background of the Doha Round is a development perspective. The problem set in this respect is impressively framed by the seven so-called ‘Millennium Development Goals’ to be achieved by 2015 (United Nations (2000)). Evenett (2003), 4.

A European Perspective of Agricultural Issues

349

ess, as GATT/WTO-negotiations are merely a means to balance differences in country interests related to trade. In this case the recent positions of the various groups with respect to the hotly disputed farm subsidies or the so-called ‘Singapore issues’ could become a matter of secondary importance. Yet, shaking the very foundations of the current framework of multilateral agreements regarding trade issues could end in more than less inequity, because a faltering WTO would probably further shift the balance of global politics from poor to rich countries. In sanguinely assuming that the current round will continue more or less within the boundaries of the Doha declaration, the definition of the problems to be solved may remain unchanged. Hence, also the concessions of negotiating parties will remain within the boundaries of their current, long-term strategy. In the case of the EU, these boundaries are set by the basic building blocks of the CAP as well as the political possibilities to change them within a reasonable time span. The following sections mainly deal with the basic interests of the EU concerning agriculture in the Doha Round, and the critical limits with respect to restricting these interests through the final results of these negotiations. B. CAP Reform Trends and Doha Development Agenda Already the discussion with respect to the pros and cons of the CAP reform of 1999 (‘Agenda 2000’) took reference to Article 20 of the URAA. Another official justification was the pending decision on Eastern Enlargement of the Union. Despite the fact that in the end this reform mainly concentrated on rearranging the financial setting of the EU (‘financial perspective 2000-2006’), the agreed changes of the CAP can be considered as a modest step into the direction demanded by the international authorities with respect to a more liberal setting regarding agricultural trade. The period until the formal decision on 13 December 2002 to include ten new member states by 1 May 2004 was characterized by a lengthy dispute between the Commission and EU member states whether Eastern Enlargement could be realized without a further reform of the CAP and the Structural Policies. Again, a key issue in this respect was the question of the availability of sufficient funds, mostly related to the current financial perspectives fixed by the Agenda Reform. In an attempt to strengthen the Commissions view that a further CAP reform would make Eastern enlargement easier and would also support the position of the Union in the upcoming

350

Markus F. Hofreither

trade negotiations, Commissioner Fischler presented his Mid-term Review of the CAP on 10 July 2002. With a formal decision on these issues pending, the Member States agreed to include ten countries as new members by May 2004. Only half a year later, in June 2003, the Council of agricultural ministers made a decision for another reform of the CAP. As this move partly changed the previously agreed setting for Eastern enlargement, there was considerable perturbation among the new member states. Although there was minor reference to international trade issues during the decision process, these two important decisions were largely treated as internal affairs.25 This is remarkable, because WTO negotiations more or less took place in parallel. The negotiations began in early 2000 with reference to Article 20 in ‘Special Sessions’ of the Agriculture Committee. This first phase ended with a meeting on 26 – 27 March 2001, which tried to achieve some stock-taking. At this time, 126 member governments had submitted 45 proposals, which mainly described these countries starting positions for the negotiations. At this time, the differences between the positions taken were considerable. In building on work carried out until then, on 14 November 2001 the Doha Ministerial Declaration set a new mandate, which made objectives more precise and set a new timetable with clear deadlines. The agenda agreed in November 2001 in Doha was mainly based on the conclusions of the URAA, but added also some new objectives, e.g. x to enable developing countries to effectively take account of their development needs, including food security and rural development; x to confirm that non-trade concerns will be taken into account in the negotiations as provided for in the Agreement on Agriculture.

25

The press statement of 13 December 2003 related to the successful conclusion of the negotiations to include ten new member states as from 2004 e.g. does not contain a single reference to the WTO issues (Enlargement and agriculture: Summit adopts fair and tailor-made package which benefits farmers in accession countries, IP/02/1882, Press Release, RAPID database, Brussels/Copenhagen 13 December 2002).

A European Perspective of Agricultural Issues

351

In the second phase, mostly informal meetings covered and compared technical details of the member’s proposals in order to finally reach a consensus about changes of rules and commitments in agriculture. This second phase ended in March 2002, however, without substantially narrowing down the gaps between country positions. According to the Doha Declaration, the deadline for the modalities in agriculture was 31 March 2003. This ‘modalities phase’ was considered to be a very critical stage of the negotiations, as these agreements are expected to widely determine the shape of the final outcome, which according to the Doha declaration is due by 1 January 2005. With having ‘the aims of the Doha Declaration in mind’ the EU submitted a proposal with quantitative details of its starting position for the upcoming negotiations in January 2003.26 The EC’s initial proposal for modalities in the WTO Agriculture negotiations contained the following objectives:27 x Further substantial liberalization on fair and equitable basis by cutting trade-distorting domestic support by 55% from the URAA levels measured by the AMS,28 tariffs by 36% on average with a minimum per tariff line of 15%, budgetary outlays on export subsidies by 45% and a ‘substantial cut in the volume of export subsidies’, with a view to even phasing out export subsidies for certain products which are important for developing countries, provided ‘that others remove their subsidies for the same products’; x The special needs of developing countries should be adequately addressed by demanding duty free and quota-free access for all imports from the Least-Developed countries, zero 26

27 28

One can safely assume that the EU has proposed this offer with a close look at the room to move provided by the Agenda reform 2000, but also the pending reform of June 2003. European Commission (2003a). The AMS (Aggregate Measure of Support) is the sum of the amount of market price support (based on a fixed, historic world price average), ‘non-exempt’ direct payments, and other subsidies not exempt from reduction, with levies or fees imposed on agriculture deducted. Details on the calculations of the AMS are specified in Annex 3 of the URAA.

352

Markus F. Hofreither

duty access for at least 50% of the developed countries imports from developing countries, and lower reduction rates as well as a longer implementation period for developing countries. x Recognition of a model of sustainable agriculture based on environmental protection, rural development, food safety and other consumer concerns by exempting domestic support measures aimed at addressing these concerns from reduction commitments, clarifying the use of the precautionary principle with respect to the level of food safety chosen by individual countries, and introducing mandatory labelling schemes with the objective of informing consumers about the characteristics and the production methods of agricultural products. Not surprising, the EU linked this offer to a number of conditions.29 Commissioner Fischler repeatedly criticized the de minimis rule as being a considerable loophole – primarily for US agriculture – in domestic support commitments, which must be closed.30 Similarly, export credits, State Trading Enterprises and the dumping of surpluses under the cover of food aid have to be brought under WTO disciplines similar to export subsidies. Last, but not least, due to the restructuring of the CAP since 1992, non-trade concerns have become an important justification for agricultural support coming in the form of direct payments. Hence, a very important point for the EU is the continuation of the blue and the Green Box. The Commission also maintains that the Peace Clause should be prolonged, as otherwise the sustainability of future WTO agricultural agreements would be jeopardized.31 The EU offer does not contain any concessions with respect to the more sensitive aspects of market access (high tariffs, tariff peaks and tariff escalation), as well as for other WTO-related issues in agriculture. The main reason for this may be that the EU expects to extract the maximum concessions from its trading partners before 29

30 31

Probably also as a reaction to internal critics of his offensive WTO strategy, Commissioner Fischler repeatedly emphasized that ‘there will be no unilateral disarmament’ and also EU Trade Negotiator Commissioner Lamy has stated that the EU will demand concessions from its trading partners. European Commission (2003b), 2. European Commission (2003a); Steinberg / Josling (2003), 7 et seqq.

A European Perspective of Agricultural Issues

353

it translates its reform efforts into WTO commitments. Though, compared to the positions in the UR, the January 2003 EU proposal has to be seen as a step which indicates the willingness of the EU to make contributions towards more liberalized agricultural trade, at least within the boundaries of acceptance within the Union.32 Despite intensive negotiations in February and March 2003 the Member countries were not able to find a compromise on the key parameters for an agricultural framework accord. Hence, on 31 March the chairman of the special (negotiating) session of the WTO Committee on Agriculture, Stuart Harbinson, formally declared that Members had failed to agree on agricultural modalities. C. Approaching Cancún: the CAP Reform of June 2003 In realizing that a consensus on modalities, including ‘hard’ numbers, was very unlikely before and even in Cancún,33 the negotiating parties changed their strategy. They now tried to at least agree on a modalities ‘framework’ in the form of an attachment to the draft Cancún Ministerial Text. As a contribution to break this deadlock, on 13 August the EU and the US presented a ‘Joint Text’ which also proposed parameters for a modalities ‘framework’34. This attempt initiated some activity in this area, as other countries also put forward own proposals. An important prerequisite for this joint offer on the side of the EU was the June 2003 decision to reform the CAP, because this ‘historic’ reform brought about a fundamental conceptual shift in the way subsidies are provided. The key elements of this reform are x a single payment for EU farmers, which is independent from production (‘decoupling’) and leaves some few forms of limited coupled elements only to avoid abandonment of production (‘cross-compliance’),

32

33

34

A simplified timeline with respect to internal as well as external drivers influencing the concrete way of the EU towards Cancún can be found in the Appendix (see Table A-1). According to the initial timetable, until the Cancún meeting countries should already have prepared their ‘schedules’, which means country-specific detailed commitments based on the ‘modalities’. ICTSD (2003), 11 et seqq.

354

Markus F. Hofreither

x

a reduction in direct payments (‘modulation’) for bigger farms with the money retained being transferred to the ‘second pillar’ (rural development programs); x a strengthened rural development policy with more money, and new measures to promote environment, quality and animal welfare; x an improved financial discipline to secure farm budget until 2013; x some changes in market policy.35 By breaking the link between subsidies and production for quite a few commodities, this reform is in line with long-term demands of international bodies with respect to the preferred way of reducing trade-distorting support. Commissioner Fischler presents the CAP reform decision of June 2003 as a move towards a tradefriendly farm policy,36 which at the same time has positive consequences for the environment as well as for consumers. Indeed, the ‘Fischler reform’ reduced producer support and thus made the CAP significantly less trade distorting.37 Commissioner Fischler stresses that the cumulative effect of the last three reforms is a reduction of the most-trade distorting forms of agricultural support by 70%, and its export subsidies even by 75%.38 The perceivable impact of these reforms is – among other facts – corroborated by a decline of the net export position of the EU. The most visible effect occurred within the wheat market, where the EU’s world market share has fallen from 22.4% to 12.8% over the last ten years.

35

36

37

38

Among them, asymmetric price cuts in the milk sector for butter and skimmed milk powder, a reduction of the monthly increments in the cereals sector by half, and reforms in the market orders for rice, durum wheat, nuts, starch potatoes. European Commission (2003a), 3. ‘Europe has also shown that we not only talk the talk, we are also ready to walk the walk towards a trade-friendly farm policy’, European Commission (2003f), 2. European Commission (2003e), 1. This holds at least under the assumption that it is indeed possible to distinguish between support measures with respect to their trade-distoring effects, as the various ‘boxes’ of the URAA suggest. Fischler (2003), 2.

A European Perspective of Agricultural Issues

355

In influencing the potential to make offers in line with its Doha commitments the CAP reform 2003 has a strong bearing on the negotiating position of the EU. In fact, the shift of support criteria from output criteria to quality and standards (animal welfare, animal and plant health, food safety, environmental indicators) facilitates a move towards less trade distorting forms of support. Moreover, the EC stresses the fact that this reform will be beneficial for developing countries, as it will lead to ‘substantial reductions in EU exports of subsidised production and, in any case, to less spending on export subsidies.’39. Commissioner Fischler regularly points to the EBA initiative40, which gives free market access for 49 of the least developed countries of the world. Additionally, with the enhanced generalised system of preferences41 more than 140 countries profit from preferential market access arrangements. In general, the Commission sees the CAP on a clear path through its changes towards more transparent and non trade-distorting instruments in line with the Doha Declaration Agenda. As mentioned above, based on the Luxembourg decision in June 2003 the EU put forward its joint initiative with the US on 13 August 2003. Among other things, this initiative called for a special safeguard for developing countries to protect sensitive products, suggested a system of preferential treatment, and urged other countries

39 40

41

European Commission (2003a), 5. The EU’s ‘Everything but arms’ initiative (EBA), introduced in 2001, grants duty-free and quota-free access for all goods, excepted arms, originating in least developed countries. Critics maintain that the benefits are relatively minor, because over 99 per cent of EU imports from the LDCs are in products with already liberalized EU access, while at the same time the removal of barriers for remaining key products, rice, sugar and bananas, has been delayed (Cernat et al, (2003), 29). In some cases, increased market access for LDCs comes at the expense of other preference-receiving countries. The Generalized System of Preferences (GSP) was first introduced in 1971 and allows industrialized countries to grant non-reciprocal tariff reductions to developing countries. The underlying motivation was helping the developing world to industrialize. The current version of the EU’s own GSP is effective from 1 January 2002 and – in the words of Pascal Lamy – is the most generous GSP package ever put forward by the European Union.

356

Markus F. Hofreither

to also engage in duty-free access initiatives similar to the EBA initiative.42 In mainly following this EU-US-proposal, on 24 August WTO General Council Chair Pérez del Castillo presented a revised draft for a ‘Cancún Ministerial Text’ containing a framework for establishing modalities in agriculture in a separate annex. Yet, until the Cancún Ministerial, the reaction of the various negotiating groups to this text was modest. Mainly, countries reiterated previously held positions. D. The Derbez Draft – a Promising Compromise? This draft text was tabled by the Conference Chair, Luis Ernesto Derbez, on 13 September 2003. It was based on previous drafts prepared by Harbinson in early 2003 and Perez del Castillo shortly before Cancún43. During the conference quite a few negotiation parties expected that an agreement in agriculture would be possible on the basis of this draft. The following Table 1 illustrates selected important positions of the Derbez text: Table 1: Selected proposals of the ‘Derbez Text’ Topic Position held in Derbez Text Domestic … demands substantial reductions of both Final Support Bound Total AMS as well as product-specific AMS, with members having ‘higher trade-distorting subsidies making greater efforts’ (1); … direct payments [Blue Box definition] shall not exceed 5% of total value of agricultural production and shall be subject to linear reductions (1.3); Green Box criteria to be reviewed with respect to trade-distorting effects (1.5).

42

43

Yet, these steps can be and actually have been seen as an attempt to conduct ‘business as usual’ in coordinating interests with the US before an important Ministerial Meeting, which caused resistance among developing countries, mainly the G20+ coalition. The way in which the Derbez draft addressed the concerns of the major participating groups is presented in detail in Matthews (2003), 4 et seqq.

A European Perspective of Agricultural Issues

357

Position held in Derbez Text … includes developing countries needs with respect to rural development, food security and/or livelihood security, and de minimis domestic support (1.6); introduced by longer implementation periods and enhance provisions unter Art. 6.2. and Green Box (1.7) Market Access … substantial improvements in market access (2.) Tariff Reduc- … mixed approach using both the Swiss formula tion (2.1 (i)) as well as the UR formula (2.1 (ii))with some additional flexibility possible for a ‘very limited number of products’ on the basis of nontrade concerns (para 2.2) Tariff Escala- … clear formula to address this issue (rate of tion tariff reduction for the processed product is equivalent to that for the product in its primary form times a factor of, at least, 1.3 (para 2.3) Tariff Rate … reduction of in-quota tariffs proposed, but Quotas conditions for further expansion of TRQ expansion remains open (2.4). Erosion of … retarded implementation of tariff reductions Preferences for products being of high export importance for developing countries (2.7). … duty-free access for a share of developing country imports still to be determined Special Safe- … elimination of SSG is left to further negotiaguard Mecha- tion (‘subject to conditions and for products to be nism determined’) (para 2.9) Export Com- … reductions of, with a view to phasing out, all petition forms of export subsidies (3.) Forms of sub- … reduction commitments shall be applied sidies equally to export subsidies, export credits, export state trading enterprises, and food aid (3.4), with an ‘end date for phasing out’ remaining under negotiations. (3.6) … longer implementation periods (3.8); Special and … after phasing out of export subsidies, develdifferential oping countries shall continue to benefit from treatment S&D provisions of Art. 9 of the AoA. (3.9). Topic Special and differential treatment

358

Markus F. Hofreither

Topic Least Developed Countries

Position held in Derbez Text … exempt from reduction commitments (4.); … developed countries grant duty-free and quota-free market access from LDCs (4.).

Remarks: figures in brackets indicate the number of the relevant paragraph in the original text.

In general, the Derbez text is close to the EU-US-proposal. Hence, for the EU many points of the Derbez Text should not pose insurmountable problems. Critical issues are the implicit threat to the volume of the Blue Box44 as well as the possible problems emanating from a revision of the criteria of the Green Box, as both boxes are of fundamental importance for the current setting of the CAP. Since 1992, the strategy of the EU in developing its agricultural policy concentrates on ‘box switching’, which stands for making support less trade-distorting by moving subsidies from the Amber Box to the blue box and the Green Box.45 Hence, the proposed tightening of criteria and probably also the volume with respect to the Green Box could jeopardize the CAP reform of June 2003 with its strong move towards decoupled payments.46 Also the remaining steps with respect to enlarging the EU will critically depend on room to move with respect to the blue and particularly the Green Box. For the EU, it is also important that the AMS remains non-product specific, as otherwise the loss of flexibility could pose severe problems in meeting current WTO commitments. The Derbez text also does not support the EU in its efforts to achieve some progress with respect to non-trade concerns (e.g. food safety, envi-

44

45

46

Shortly after Cancún, the G20 dropped the demand for the total elimination of the Blue Box, but suggested to limit this form of support to 2.5 percent (instead of 5 percent) of all agricultural production ‘with a view to phasing out’. In making agricultural support less trade-distorting another benefit comes in the form of getting rid of reduction requirements, as with respect to the Blue Box as well as the Green Box currently no reduction requirements exist. Currently, total support provided through the CAP budget is allocated to the Amber Box with approximately 20%, the Blue Box with 70%, and the Green Box with 10%.

A European Perspective of Agricultural Issues

359

ronmental protection, animal welfare, consumer protection).47 Another problem area with a clear North-South-dimension is the question of how in detail to implement special and differential treatment for developing countries.48 Despite these reservations, for the EU the Derbez text may be a feasible starting point for the continuation of the negotiations, because it contains a number of starting points for acceptable compromises. A very compelling argument in favour of the Derbez draft is that no feasible alternative to this text exists, as the joint proposal of the EU and the US has been dismissed by developing countries. Hence, dropping this text could push back the trade talks to the start, with risky consequences for the whole round. In summing up the previous analysis, it seems hardly correct to put the blame for the failure of Cancún on a particular topic or group. It has been a complex mix of negotiation accidents through information deficits, shortness of time, ill-constructed strategies and differences with respect to the substance. Here, besides agriculture mainly the cotton conflict as well as the different approaches with respect to the Singapore issues have been the main stumbling blocks. With increasing distance to the events in Cancún it becomes less useful to speculate about the possible routes of the negotiations in Cancún if the talks would not have stopped on Sunday. As a matter of fact, the talks have stopped and the key question is about the future prospects of this round. IV. Cancún and beyond As already mentioned above, the events in Cancún went not the way the EU had expected. Although the Derbez draft at first has been considered as an acceptable starting point, the assessment later on changed and EU officials were anxious to stress that this draft was never seriously discussed at any Cancún session and so cannot be considered to be the latest state of discussions. A possible reason for this position could be not to loose too much ground before a 47 48

European Commission (2003d), 6 et seqq. In general, the EU insists on this is to be done by varying implementation periods and the extent of commitments, while developing countries would like to see a completely different set of rules (Matthews (2003), 13).

360

Markus F. Hofreither

formal restart of the negotiations, because other groups have indicated that the Derbez text still is too modest in order to meet the objectives set out in the Doha declaration.49 Depending on the future course of negotiations, the Fischler reform of June 2003, despite some ambition towards trade liberalization, may be not sufficient to meet the requirements of upcoming WTO agreements. However, for the EU this reform has considerably narrowed the room to move in the forthcoming negotiations, as the key elements of the CAP depend on particular minimum levels of protection as well as the general acceptance of particular instruments, e.g. to support non-trade concerns. A key issue for European agricultural policy is the notion of ‘multifunctionality’. This position sees agriculture as a provider of public goods and positive external effects, which are not remunerated by markets. In order to provide these services, which are understood as demanded by society, often the public hand has to step in. Hence, in being compensations for public good provision, the resulting payments should not be counted as trade-distorting subsidies. The obvious problem in this case is that internationally agreed rules, which allow differentiating between truly ‘multifunctional’ activities and disguised protectionism, do not exist. Until now this problem has been solved mainly on a ad hoc case-by-case basis, and it remains an open question whether in the future this can be replaced by accepted general rules. However, if core requirements of the CAP are at stake, the EU could become a huge obstacle on the way to a successful conclusion of the Doha Round. The implicit threat in the wording of Commissioner Fischler with respect to the possibility of ‘dozens of panels against the EU’ after the expiration of the Peace Clause highlights this view50. If in fact the EU gets into a position where only another 49

50

In mainly reflecting the position of Australia, Podbury et al. ((2003), 2) demand, among other things, the elimination of export subsidies as well as the special safeguard mechanism for developed countries, an expansion of the volumes under tariff quotas, and a changed methodology for calculation of the AMS in order to better reflect actual support. The Peace clause has already expired by the end of 2003 and thus primarily the export subsidies of the EU may come under fire, as under the current legal framework no straightforward defence mecha-

A European Perspective of Agricultural Issues

361

CAP reform makes it possible to agree to a new WTO Agreement in Agriculture, this round may last very long. But also the US is facing a challenging year, with presidential elections taking place in November 2004, which may imply a reduced willingness to deal with subtle trade problems. Despite the fact that the continuation of the negotiations in the Doha Round will not become easier in 2004, the odds for a constructive continuation of the round after the drawback of Cancún appear to be fair. Certainly, the setback of Cancún will delay the round by at least one year.51 But this is neither a new experience in trade negotiations nor is it granted that without the failure of Cancún a conclusion of the Doha Round by the targeted date would have been guaranteed. Nevertheless, one cannot completely exclude the possibility that Cancún has caused a structural break in the overall framework of upcoming trade negotiations. This break may come in the form of a heightened self-esteem of developing country groups and their experience that effective coalitions are an important means in order to get a position through. In addition, developing countries may increasingly recognize that, at least in a relative perspective, what industrialized countries offer is much less than what is at stake for themselves.52 This comprehension could severely influence the mental framing of the negotiation object. In agriculture the question, whether the distinction between trade-distorting and least-trade distorting support really makes sense, may become a key issue in future negotiations between the North and the South. This distinction was at the core of work of

51

52

nism is available. At present, the EU holds more than 80% of global export subsidies, with dairy, beef meat, and coarse grains being the most important positions. Even with an optimists view the next important step – the modalities framework – will not be concluded before summer 2004, as official negotiations will hardly continue before March or April. As an example, an offer to cut tariffs under the ‘blended’ approach would allow developed countries to cut ‘import sensitive products’ by only 15%, if the UR setting is to be repeated. However, as only a few percent of more than 1000 tariff lines of the US and the EU are indeed ‘import sensitive’ this flexible approach will hardly create any additional market access (Podbury et al. (2003), 9).

362

Markus F. Hofreither

influential international organizations in this field during the last decades. In the meantime, this distinction has also become the backbone of agricultural reforms in the EU and other countries as well. Implementing agricultural reforms relying on ‘box-switching’ has allowed developed countries to become ‘less trade-distorting’ without having to cut agricultural subsidies in total. Developing countries increasingly seem to question the rationale of this construction in maintaining that all forms of agricultural support have significant trade effects.53 In this case, the south could in fact start to insist on a full dismantling of all the traditional policies, which now for decades have supported and protected the farm lobbies of rich countries. Such a scenario, however, would make the continuation of the Doha Round in a ‘business as usual’ mode extremely difficult, as ‘box switching’ has become the key approach of agricultural reforms in many developed countries and in the meantime has also achieved broad acceptance among farm interest groups. No matter how justifiable the claim for total support reduction may be from a welfare economic, environmental or equity viewpoint, such a setting could cause a detrimental deadlock in international trade talks, which would severely postpone any progress regarding the international trading system.54 Given the extremely uneven distribution of political and economic powers between North and South, pressure through external factors will hardly be enough to induce significant changes of agricultural policies in developed countries. Domestic political pressure to remove trade-distorting policies is the other indispensable ingredient.55 Only in a scenario where non-agricultural interest groups in the industrialized world see fundamental trade interests at stake, 53

54

55

A very common channel linking decoupled payments and production is the income effect of such support, which helps the farmer to raise money for further cost-reducing investments. At the same time such payments also reduce the effective risk of farmers with respect to market fluctuations. In discussing the difference of trade interest from a North-South perspective it would be misleading to follow a pure ‘black and white’-perspective. It should not be overlooked that quite a number of developing countries have policies in place, which are similarly trade-distorting as those of particular developed countries. Hoekman (2003), 5.

A European Perspective of Agricultural Issues

363

and thus exert strong pressure on agricultural decision makers to give in, a new order for domestic agricultural support in high-income countries could take force. However, at the time being there are no signs of sufficient pressure from internal factors like budget constraints or civil society groups. V. Conclusion and Outlook The reason for the failure of the 5th Ministerial Conference of the WTO in Cancún is a complex mix of tactical and strategic mistakes in combination with partly strong divergent positions with respect to the substance of important issues. Primarily the active role of developing countries, which was inspired by the Doha declaration, made these talks less straightforward in comparison to previous rounds. Given the insufficient on site capacity to analyse new proposals, the mass of issues on the agenda made it difficult for many countries to position themselves appropriately in a setting of intense time pressure. This problem was further aggravated by - in the words of Pascal Lamy - the ‘medieval’ structure of negotiations imposed by the WTO. Regarding the substance of this meeting, clearly the Singapore issues as well as agriculture and the cotton dispute have played the most important role for the ‘Tequila sunset’ in Cancún.56 In agriculture, the joint EU-US approach of mid-August was dismissed by developing countries due to its lack of ambition, mainly with respect to the elimination of export subsidies. In the end there was a political deadlock between important players, which could not be resolved in a few days. Compared to previous rounds, the European Union had approached this Ministerial meeting with a more open stance with respect to trade liberalization and also a much better pre-meeting preparation. Nevertheless, for the EU the costs of the failure of Cancún are modest, if they exist at all. Many observers from different camps agree that a successful conclusion of the Doha Round will primarily benefit the developing countries, at least if the label ‘successful’ describes an outcome with improvements in market access and sharply reduced export subsidies in agriculture.

56

Das (2003), 17.

364

Markus F. Hofreither

The worst case scenario after Cancún is that this breakdown will weaken the WTO in a way which makes it widely irrelevant for the conduct of trade policy in the future. Trade flows within the developed world would increasingly be regulated in the form of bilateral agreements between existing trade blocks, with the commitments of the UR waning in the course of time. In this case, the poorest countries could be the main victims of the partly applauded failure of Cancún: for the wealthy nations there is no time pressure to eliminate their farm subsidies or their barriers for textile imports, while developing countries will have to bilaterally negotiate the needed changes step by step, which may get a very cumbersome task. A likely result could be the creation of more solid coalitions among them, with positive effects for their negotiating authority. Nevertheless, the successful development of the international trading system since the creation of the GATT would very likely come to a complete halt. At the beginning of the year 2004 there is some evidence that this worst case scenario will not come true, as countries seem to recognize the importance of a well-functioning international trading system as one of the fundamental factors determining the well-being of the people living both in developing and developed nations. The immediate questions seem to be from what basis and when the talks will resume. In the course of future negotiations within the Doha Round also more general questions – e.g. regarding the adequacy of the construction and the objectives of the WTO if a development mandate is adopted – will emerge. However, it is not a question that in future any advance in the existing international trading regime will only be possible through a sincere recognition of the problems of developing and least-developed countries.

References Agra Facts (2003), Cancún Outcome will not change EU approach to reform, says Fischler, No 81-03, 19 September 2003, Paris 2003. Lucian Cernat / Sam Laird / Luca Monge-Roffarello / Alessandro Turrini (2003), The EU’s Everything But Arms Initiative and the Least-developed Countries, Discussion Paper No. 2003/47,

A European Perspective of Agricultural Issues

365

June 2003, World Institute for Development Economics Research, Helsinki 2003. Dilip K. Das (2003), The Doha Round of Multilateral Trade Negotiations: Causal Factors behind the Failure of Cancún, (mimeo), Missisauga 2003. European Commission (2003a), EU Agriculture and the WTO, Doha Development Agenda, Cancún – September 2003, Directorate-General for Agriculture and Rural Development, September 2003, Brussels. European Commission (2003b), Commissioner Fischler calls for fair burden sharing in WTO farm talks, News Releases, No. 07/03, Delegation of the European Commission to the United States, Washington, February 4, 2003. European Commission (2003c), Singapore Issues – Options postCancún, Directorate-General for Trade, Ref. 514/03, Brussels, 30 October 2003. European Commission (2003d), Reviving the DDA Negotiations – the EU Perspective, Communication from the Commission to the Council, to the European Parliament and to the Economic and Social Committee, Brussels, 26 November 2003. European Commission (2003e), CAP reform strengthens the EU position in WTO talks, Newsletter, No 56, Brussels, July/August 2003. European Commission (2003f), EU is leading the way to a tradefriendly farm policy, September 2003, downloadable from: www.delmys.cec.eu.int/en/european_union/key_external_polic ies/eu_agricultural_trade/cap_reform_june_03.pdf. European Commission (2003g), The Doha Development Agenda After Cancún, Directorate-General for Trade, Brussels, 25 September 2003. Simon J. Evenett, (2003), The Failure of the WTO Ministerial meeting in Cancún: What implications for future research on the world trading system? mimeo, to be published in CESifo Forum, Volume 4, No 3, Autumn 2003, 11-17. Franz Fischler (2003), From Cancún: The road ahead for the trade and agriculture negotiations, Speech/03/413. Ontario Ministry of Agriculture and Food, Guelph, Canada, 16 September 2003, downloadable from: http://www.europaworld.org/week145/ speechfischler19903.htm.

366

Markus F. Hofreither

Bernard Hoekman (2003), Cancún: Crisis or Catharsis? Paper presented at the Sept 17th meeting of the Brookings-George Washington Trade Roundtable, downloadable from: http:// siteresources.worldbank.org/INTRANETTRADE/Resources/H oekman-CancunCatharsis-092003.pdf Bernard Hoekman / Michel Kostecki (1995), The Political Economy of the World Trading System, From GATT to WTO, Oxford University Press, Oxford 1995. International Centre for Trade and Sustainable Development (ICTSD) (2003), Agriculture Negotiations at the WTO: PostCancún Outlook Report, Geneva, November 2003 (available at: http://www.ictsd.org/issarea/atsd/products/docs/Agriculture Negotiations9.pdf). Tim Josling (2001), WTO: Successes and Failures, Presentation prepared for the 1st International Agricultural Trade and Policy Conference, IFAS, University of Florida, downloadable from http://www.iatpc.fred.ifas.ufl.edu/conference1/joslingpapaer.ht ml, Gainesville, October 31-November 2, 2001. Allan Matthews (2003), Agriculture After Cancún, Trinity Economic Paper No. 17, 2003, Dublin 2003. Troy Podbury / Shirshore Hagi Hirad / Neil Andrews / Wayne Gordon (2003), WTO Agriculture Negotiations, The way forward from Cancún, Australian Commodities, Vol. 10, No. 4, December quarter 2003, 1-14. Jeffrey Schott (2003), Unlocking the benefits of world trade, Economist.com, Oct 30th 2003, Washington, DC. T. N. Srinivasan (2003), The Future of the Global Trading System – Doha Round, Cancún Ministerial and Beyond, Paper presented at the conference ‘The Future of Globalization: Explorations in Light of Recent Turbulence’, Yale/NJ, October 10-11, 2003. Richard H. Steinberg / Timothy E. Josling (2003), When the Peace Ends: The Vulnerability of the EC and US Agricultural Subsidies to WTO Legal Challenge, Comment, Bridges, No. 8, ICTSD, November 2003, 7-10 (available at: http://www.ictsd. org/issarea/atsd/products/docs/Steinberg_Josling.pdf). Third World Network (2003), Process and substance caused failure at Cancún, Geneva, 16 September 2003, downloadable from http://www.ongd.lu/article.php3?id_article=360.

A European Perspective of Agricultural Issues

367

United Nations (2000), A better world for all, Progress towards the international development goals, Washington 2000. Zhen Kun Wang / L. Alan Winters (1997), Africa’s Role in Multilateral Trade Negotiations: Past and Future, Working Paper, Economic Development Institute and Development Research Group, World Bank, Washington D.C., 1997.

368

Markus F. Hofreither

Table A-1: Time line of internal and external factors determining the WTO-position of the European Union Period Internal Factors External Factors 07/1999 ‘Agenda Reform’ of the CAP 11/1999 3rd Ministerial Conference (Seattle, United States) 03/2001 WTO-Meeting on stocktaking regarding Art 20 URAA 11/2001 4th Ministerial Conference (Doha, Quatar) ‘Doha Declaration’ 07/2002 Mid-term Review of Agenda 2000 (‘Communication’ presented by Commissioner Fischler) 12/2002 Copenhagen Decision on ‘Eastern Enlargement’ 01/2003 EU proposal for agricultural ‘modalities’ 06/2003 CAP Reform 2003 (‘Fischler Reform’) 08/2003 US-EU joint proposal G16/22 proposal 09/2003 5th Ministerial Conference (Cancún, Mexico) Source: own compilation.

Wolfgang Weiß*

GATS and Domestic Regulation – A Threat to Democracy?** I.

GATS and Democracy A. National Regulation and the Peculiarities of Services B. Legitimacy in International Law C. Adjudicating Necessity D. Sovereignty and Interdependance E. Liberalization, Democracy and Human Rights II. Some Remarks on GATS and Domestic Regulation A. Liberalization and Cross-border Trade in Services B. Balancing Liberalization and Legitimate Interests by Regulatory Disciplines C. Questions Regarding the Application of Art. VI GATS 1. Legal Value of the Disciplines Adopted According to Art. VI:4 GATS 2. Applicability of the Disciplines Adopted According to Art. VI:4 GATS 3. Relationship between Disciplines and Specific Commitments 4. Regulatory Measures not Included D. Mutual Recognition References

* **

370 370 371 373 375 376 376 376 377 380 380 380 381 381 381 382

Prof. Dr. Wolfgang Weiß, University of Erlangen-Nuremberg. This paper is a comment on the paper ‘GATS and Democratic Legitimacy’ by Rudolf Adlung that was presented at the Vienna conference and is published in Aussenwirtschaft 59 (2004) No. 2, 127-149.

370

Wolfgang Weiß

The paper presented here is a comment on the thoughts presented by Rudolf Adlung and, furthermore, adds some general remarks on the issue of GATS and domestic regulation. Mr Adlung discussed the problem as to whether GATS is a threat to democracy. In choosing this topic, Mr Adlung spoke on one, maybe the most important particular aspect of the broader issue of GATS and domestic regulation. The issue of democracy and in particular the fear that GATS threatens democracy is one of the concerns raised by opponents of the GATS. I. GATS and Democracy A. National Regulation and the Peculiarities of Services Opponents to the GATS and to the WTO in general proclaim that GATS restricted domestic regulation as it limited governments’ ability to regulate the supply of service, in particular regarding standards of qualification, licensing and standardization requirements. For this reason, GATS was a threat to the sovereignty of the WTO members and a threat to democracy.1 Indeed setting international standards for regulating services gets closer to the heart of democratic decision making than it is the case with trade agreements on goods. The potential threat to democracy is greater in GATS than in GATT, if there is any at all. The reason for this lies in the peculiarities of services. Services are much more complex when compared to goods. Mr Adlung mentioned it. This can be seen from the fact that GATS defines and regulates the different modes of service supply. GATS knows four modes of supplying services (see Art. 1:2 GATS). Three of them are connected with the move of persons (Art. 1:2 b-d: consumption abroad, commercial presence or presence of natural persons), and only one mode refers to the service itself crossing the border (crossborder supply), see Art. 1:2a GATS.2 This highlights that the supply of services usually does not meet border measures regarding the service itself but regulations that regulate the pre- and side-conditions of the supply of services like qualifications, licensing, authorizations and standards. Thus, protectionism in services does not take the form of border measures which apply to the trade in goods: tariffs and non-tariff border measures. But in contrast to goods, 1 2

See e.g. Joy (2001) and Woodroffe (2001). See Matsushita / Schoenbaum / Mavroïdis (2003), 236.

GATS and Domestic Regulation – A Threat to Democracy?

371

protectionism in services trade takes almost exclusively the form of internal regulatory interventions3 that provide for requirements for the supply of services and not of border measures. As poor services, unlike goods, can not be taken back, governments have to protect the consumer in the most appropriate way, i.e. by prior regulation. Additionally, there are some particular requirements which refer to the supply of services from abroad and which restrict the service supply from abroad in case it is connected with persons crossing the border: the requirements for entry and stay of foreigners. Thus, national regulation has a great importance for the supply of services. This circumstance, first, means that in order to liberalize cross border service supply one has to focus on the internal regulation of service supply and, second, makes it much more complex to detect protectionism and to discern protective internal regulations from regulations that pursue legitimate social or environmental policy goals. Thus, the peculiarities of services make the relation between domestic regulation and international requirements much more complex with regard to services than to goods.4 An international agreement like GATS that was created for the liberalization of cross border service supply has to address the issue of domestic regulation and has to try to find common standards and denominators for the domestic regulation of the supply of services by WTO members. This influences and restrains the regulatory options of the WTO members and makes GATS particularly sensitive for democracy concerns as the national regulatory autonomy comes under greater scrutiny. Mr. Adlung dealt with the different aspects of this discussion. In general, I agree with his view that GATS is not a threat to democracy. B. Legitimacy in International Law The background of this debate is the question whether decisions taken in international fora are legitimate. When addressing this issue one has to be aware that the requirements for legitimacy are different in international settings. The debate on the constitutional requirements of the European Integration Process witnesses the same concerns. International organs usually are not elected by people, but nevertheless their decisions are democratically legitimate although specialized international agencies like the WTO do not 3 4

Djordjevic (2002), 305. Krajewski (2003), 2 et seq.

372

Wolfgang Weiß

and cannot derive their constitutional legitimacy in the same way as nation states.5 First of all, we have to bear in mind that the WTO is the creation of states. Its legitimacy is derivative in character and based on state sovereignty.6 In the established international order the states are the primary actors. Therefore, the question of legitimacy can only be discussed from the perspective of the individual government as Mr Adlung rightly mentioned. The WTO members have agreed to the obligations enshrined in any of the WTO agreements, including those in GATS. Mr Adlung analyzed in a detailed way the possible democratic deficits such agreement could face, like for example the possibility of being outvoted by majority, the imbalances between WTO members and the pressure created by domestic pressure groups and lobbies. He rightly pointed out that the latter dangers are not a peculiarity of international policy making but are a challenge to domestic regulatory processes as well.7 The challenges are the same, at least comparable. Therefore, there is no reason for blaming the GATS to be undemocratic as long as the states are free to take up only those obligations they agreed to. Opponents to the GATS when raising concerns on the prevalence of the rich nations in the international trade negotiations pretend to speak in the interest of small nations and developing states. But they tend to forget that international organizations like the WTO are better than none. Multilateralism is always better than unilateralism; multilateralism in particular favours small countries. The poor states will lose if the WTO is abolished and unilateralism is reintroduced. This deliberation, however, does not mean that there is no need to improve the rules and structures of the WTO. An additional remark on GATS and developing countries can be added: Opponents raise concerns because developing countries are at a disadvantage, as they, for example, do have only a limited negotiating capacity. From the point of view of democracy, this could be welcomed, because many, maybe most of the developing countries 5 6 7

Matsushita / Schoenbaum / Mavroïdis (2003), 608 et seq. Matsushita / Schoenbaum / Mavroïdis (2003), 609. Thus, when Krajewski (2001), 177 states that the WTO decision making process does not meet the conditions of deliberation and rational discourse due to package deals and wrong type of compromises, one has to remember that this applies to national legislative processes as well.

GATS and Domestic Regulation – A Threat to Democracy?

373

are not really democratic states (the governments were not elected at all or the elections were manipulated; the political system is authoritarian or autocratic rather then democratic). Thus, it could be proclaimed that democratic legitimacy of the WTO would be weakened if the developing states had increasing influence.8 This argument highlights that the issue of legitimacy in international settings is quite intricate. Second, legitimacy is not only an issue of democracy and election, but also an issue of values and citizens’ rights. The more an organization or an agreement respects human rights the higher is its legitimacy (output legitimacy9). GATS is blamed to fundamentally undermine citizens’ right to determine their own social and environmental rights and priorities for the future. But GATS recognizes the sovereign right of Member governments to determine the regulatory objectives they want to pursue and to define the interests they want to protect, and to introduce new regulations on the supply of services within their territories in order to meet national policy objectives. This right, which respects policy choices in accordance with the democratic will and perceived needs and cultural traditions of each society, is mentioned in the preamble of GATS and has been re-emphasized at the Doha Ministerial10 and was reaffirmed in the recently adopted Guidelines and Procedures for the Negotiations on Trade in Services. This clear statement has to be welcomed all the more since Art. VI:4 GATS could be understood to protect solely the quality of the service (see Art. VI:4 lit. b)) and no other domestic policy objectives. C. Adjudicating Necessity There is, however, one issue that has to be looked at slightly differently to the view Mr Adlung presented: the dispute settlement in the WTO. Mr Adlung rightly pointed to the fact that international agreements like the GATS offer more scope for interpretation. Thus, the question has to be raised as to the exact meaning of the criteria of Art. VI:4 GATS. When is a domestic regulatory requirement more burdensome than necessary? Who defines it? 8

9 10

This argument is used here only for the sake of argumentation. It is not the interest of the author to drive the developing states out of the WTO. See Krajewski (2001), 169. See the Doha Ministerial Declaration, WT/MIN(01)/DEC/1, para. 7.

374

Wolfgang Weiß

It cannot be contested that the necessity tests contribute to the philosophy of better regulation (like, for example, transparency and proportionality do as well). One can, however, be extremely concerned about the fact that these ‘necessity tests’ are very much open to interpretation and that the WTO dispute settlement procedure could be used to enforce a strongly deregulatory interpretation in favour of business interests. One could argue that it is wholly unacceptable that a group of lawyers dedicated to the liberalization of trade could be given the right to declare regulations that have been enacted by democratically elected governments ‘excessive’ or ‘unnecessary’ as there is always a complex judgement to be made about the right balance between freedom of trade and public interest regulation.11 Why should WTO adjudicators be seen to make sounder assessments as regards the necessity of domestic regulation? It seems that it is more the responsibility of elected governments to make those judgements, and not of unelected lawyers working for a body which is seen by some as being ideologically committed to maximizing economic liberalization. In order to meet this concern there is need for the panels and the Appellate Body to adopt interpretations that are deferential to state sovereignty. It seems to be doubtful whether this always has been the case. The other side of the coin is the as well indubitable fact that the WTO members agreed to broad terms in the WTO agreements and to WTO dispute resolution mechanism. Thus they were aware that the WTO panels and the Appellate Body were responsible for correctly interpreting the WTO agreements when settling disputes unless the WTO members (i.e. the General Council and the Ministerial Conference) adopt an authentic interpretation according to Art. IX:2 WTO Agreement. Furthermore, the Dispute Settlement Body may deny a panel or Appellate Body report if it does not share the legal opinions given there. In the general perspective of democracy, legitimacy and division of powers, the problem of legitimacy of the interpretation of norms is well known in domestic law as well. Is it undemocratic 11

The EC holds the view that a measure should not be considered more burdensome than necessary if it is proportionate to the objective pursued. This is intended to make clear that the degree of ‘necessary’ trade restrictiveness will depend on, and be assessed against, the technical and economic context of a specific domestic policy objective, but would not question its central validity or rationale.

GATS and Domestic Regulation – A Threat to Democracy?

375

that a judge interprets a particular norm contrary to the will of the legislator? Or is this not an expression of division of powers which no doubt contributes to legitimacy and expresses basic constitutional principles? The problems of legitimacy of ‘government by judges’ or judicial policy-making that the WTO and the international level are faced with do have their counterparts in the national level where legitimacy in the end is not doubted, at least not to the same extent.12 Thus, again, the problems are much more intricate and too complex than rashly to blame the WTO to have a legitimacy gap. Another contentious issue of the necessity test which can only briefly be mentioned here is the issue of cross-subsidization. The necessity test could become a threat to cross-subsidization (which seems to be particularly necessary for developing countries) by involvement of private companies in public services. The importance of cross-subsidization as a mechanism for developing countries to realize social objectives seems to be widely acknowledged.13 The problems of adjudicating necessity raised here can be solved by introducing flexibility. The balancing test provided for in Art. VI:4 GATS gives room for the necessary flexibility.14 Either the adopted disciplines or the test which assesses the validity of domestic regulations in GATS should assimilate flexibility.15 D. Sovereignty and Interdependance When blaming the WTO to be undemocratic, opponents also seem to forget that the WTO Members get the possibility to influence the behaviour of other nations. Of course, binding commitments restrict the sovereignty of any state. It is bound and no longer allowed to behave differently. The regulatory independence becomes limited. But in exchange for these restrictions the states can contribute to the international negotiating process and invest their interests there. It is a mutual give and take situation. There are, of course, imbalances between WTO members. But they would even be worse without WTO. Weak and poor countries have, as Mr Adlung rightly pointed out, the possibility to share information, to coordinate their policy 12 13 14 15

This issue seems not to be reflected by Krajewski (2001), 182. See Hilary (2001), 47 et seq. See also Verhoosel (2002), 111 et seq. Djordjevic (2002), 309.

376

Wolfgang Weiß

and to build coalitions. One problem however remains: in international affairs governments have more powers than national parliaments. Whereas governments actively are involved in the negotiations and thus determine their content and result, the parliaments only can adopt or reject the result of the negotiations.16 This, however, again is a problem of national constitutions and not of the structure of an international organization like the WTO and its agreements. The national constitutions have to take care of the democratic accountability of the governments and to find ways for it. Here transparency in the decision-making mechanisms becomes important. The WTO can facilitate this task of the national constitutions by increasing the transparency of the negotiating process and of the decision making mechanisms. This is a necessary international safeguard against abuse of power by governments. E. Liberalization, Democracy and Human Rights In the discussion about democratic legitimacy, one must not forget that opening up of markets will speed up democratization processes and the respect to human rights. Markets can not increasingly be opened and liberalized without giving more rights to people. Open and liberalized markets require more individual economic rights because it is the citizenship of a WTO member that has to make use of increased opportunities. And this inevitably will call for more political rights and freedom. People that are economically active become more and more aware of and interested in the political requirements and pre-conditions of their trade chances in particular and in politics in general. This in the end will result in strengthening democracy in any WTO member. This is a constitutional function and benefit of international economic law leading to open markets that might not yet have been fully recognized. II. Some Remarks on GATS and Domestic Regulation The following thoughts give some basic insights into the inter-linkage of trade in service liberalization and regulation in GATS. A. Liberalization and Cross-border Trade in Services As described above, due to the particularities of services, domestic regulation has a great importance for the supply of services. The domestic prerequisites and requirements both apply to domestic and 16

See Krajewski (2001), 175 et seq.

GATS and Domestic Regulation – A Threat to Democracy?

377

foreign services supply alike on the assumption that they are administered in an impartial way (which is provided for by Art. VI:1 GATS). There are some particular requirements which refer to the supply of services from abroad and which restrict the service supply from abroad in case it is connected with persons crossing the border: the requirements for entry and stay of foreigners. The latter might have protectionist intents whereas the former domestic requirements usually are not motivated by protectionism but by legitimate concerns like, for example, consumer and health protection. The more liberalized a national market for the supply of services is, the lower are these requirements. And the lower these requirements are the easier is the supply of services for foreigners and from abroad. In other words: The more liberalized a market is at home, the more open it is for service suppliers from abroad. B. Balancing Liberalization and Legitimate Interests by Regulatory Disciplines Art. VI:4 GATS empowers the Council for Trade in Services to develop any necessary disciplines with a view to ensuring that measures relating to qualification requirements and procedures, technical standards and licensing requirements do not constitute unnecessary barriers to trade in services. Such disciplines shall aim to ensure that such requirements are, inter alia, based on objective and transparent criteria (such as competence and the ability to supply the service), not more burdensome than necessary to ensure the quality of the service and, in the case of licensing procedures, not in themselves a restriction on the supply of the service. This provision expresses the rationale that national regulation of the supply of services usually is legitimate as and insofar as it tries to safeguard legitimate national policy concerns. The WTO members restrict the supply of services on their territory due to considerations inter alia of consumer protection, functionality of the markets or of ensuring the quality of services. In principle they enjoy full choice as to which legitimate interests to follow. As mentioned above (see supra I. B.) the legitimate objectives have not to be solely concerned with the quality of the service. The limitations international attempts to liberalize service supply from abroad put on the national regulatory sovereignty therefore have to balance the interest of increased international liberalization against the legitimate national policy objectives which call for

378

Wolfgang Weiß

regulating the service supply and thereby create obstacles to cross border supply. A balance has to be found between guaranteeing quality of service while not making regulation so onerous that, for example, it becomes impossible for qualified foreign professionals to practice. Disciplines should in principle contribute to creating more appropriate, trade friendly and transparent regulatory frameworks, and thereby facilitate and promote trade in services, while taking full account of the legitimate policy objectives pursued by government regulation. For this reason, the disciplines have to be biased. Disciplines on domestic regulation should help ensuring that legitimate regulation is not applied with a view to undermine commitments negotiated. However, it seems neither to be the intention nor to be the purpose of such disciplines to prescribe or impose certain regulatory approaches, or even the substance of any domestic regulations. But they should take care that the domestic regulations do not go beyond what is necessary in order to protect or meet the legitimate national objective. Disciplines under Art. VI:4 should be sufficiently clear and precise to allow clear implementation. They should ensure, for example, that the procedures under which a licence can be obtained are subject to some basic, common rules so that they do not become unnecessary barriers to trade in services. Furthermore, Articles VI:1-3 GATS provide for additional limitations to the national sovereignty as regards administration of regulations and review procedures: All measures of general application that affect trade in services have to be administered in a reasonable, objective and impartial manner. This rule wants to avoid that the way of application of national measures weakens a country’s GATS commitments. For this reason the way how generally applicable measures are administered is limited. If an authorization is required, the competent authorities of a Member shall, within a reasonable period of time, inform the applicant of the decision concerning the application. At the request of the applicant, the competent authorities of the Member shall provide, without undue delay, information concerning the status of the application. The limitation enshrined in Art. VI:2 GATS requires review procedures: The Member States shall establish tribunals or procedures which provide, at the request of an affected service supplier, for the prompt, objective and impartial review of administrative decisions. Thus, it is said that Art. VI GATS (as does Art. X:3 GATT 1994) reflects – like in a nutshell – the conflict between the objec-

GATS and Domestic Regulation – A Threat to Democracy?

379

tives of liberalized trade in services and national sovereignty rights that may legitimately restrict the supply of services.17 Art. VI GATS as well proves that GATS can already discipline ‘non-discriminatory’ regulation.18 The disciplines mentioned in Art. VI:4 GATS apply to domestic and foreign service suppliers alike; they are of a predominantly qualitative nature. Because of its importance domestic regulation is one of the key elements of the ongoing GATS work programme. In this regard work started in 1995 to establish disciplines on domestic regulations. The development of regulatory disciplines is an important component in the current round of negotiations. The focus of the work is on qualification requirements and procedures, technical standards and licensing requirements. By December 1998, WTO members (i.e. the Council for Trade in Services) had agreed Disciplines on Domestic Regulations for the Accountancy Sector.19 Since then, WTO members have been engaged in developing general disciplines for all professional services and, where necessary, additional disciplines for specific sectors. The disciplines can have both a horizontal approach and can be developed on a sector specific basis. Currently, the Working Party on Domestic Regulation which was asked by the Council to develop disciplines on licensing requirements and procedures, technical standards and qualification requirements20 discusses whether the Disciplines in the Accountancy Sector are suitable for other professions as well. The Working Party holds consultations with international professional services organizations, as selected by WTO members, in this regard.21 All 17 18

19 20 21

Barth (1994), 457. Weiß (1999), 510. The WTO Secretariat has been given the task of collecting examples of regulations that would violate the new disciplines, and has compiled a list that includes zoning, ‘unnecessarily burdensome’ standards, government-imposed limits on fees that can be charged. It excluded any examples submitted of regulations that discriminate against foreign service suppliers, because these would already be violations of GATS national treatment commitments. Disciplines on Domestic Regulation in the Accountancy Sector, S/L/64, dated 17 December 1998. S/L/70. See Report of the Working Party on Domestic Regulation to the Council for Trade in Services (2003), S/WPDR/6, dated 3 December 2003.

380

Wolfgang Weiß

the agreed disciplines are expected to be integrated into GATS and to become legally binding by the end of the current services negotiations. C. Questions Regarding the Application of Art. VI GATS In the following some problems of the interpretation of Art. VI GATS will be raised.22 1. Legal Value of the Disciplines Adopted According to Art. VI:4 GATS As Art. VI:5 GATS implies an entry into force of the disciplines, the disciplines themselves are not binding. Furthermore, the GATS Council does not have le gislative powers. Otherwise the GATS really would become a threat to democracy. Therefore, the disciplines become legally binding only in case they were adopted by the Member States according to their constitutional requirements. Thus, the Member States have to integrate the disciplines into their schedule of special commitments.23 The Disciplines on Domestic Regulation in the Accountancy Sector24 are supposed to enter into force when integrated into the GATS.25 2. Applicability of the Disciplines Adopted According to Art. VI:4 GATS The problem has to be solved as to whether the adopted disciplines are applicable only in case of specific commitments or whether they do apply to all services sectors as Art. VI GATS belongs to Part II on general obligations. The Accountancy Disciplines were made applicable only to WTO Members who had entered specific commitments in accountancy in their schedules (see para. 1 of the Disciplines).26 Art. VI GATS explicitly refers to sectors of specific commitments in its paragraphs 1, 3, 5 and 6 whereby paragraph 5 refers to paragraph 4. Furthermore market access only takes place in sectors of specific commitments.

22 23 24 25 26

For more detail see Krajewski (2003), Chapter 5. Pitschas (2003), 544, Mn. 140. S/L/64. See Council for Trade in Services, S/L/63. Djordevic (2002), 315.

GATS and Domestic Regulation – A Threat to Democracy?

381

3. Relationship between Disciplines and Specific Commitments Mr Adlung presented the view that schedules of commitments and the restrictions on domestic regulation according to Art. VI are independent. Indeed it is considered that there should be no overlap between Articles XVI and XVII GATS, which belong to Part III of the GATS, and Art. VI, which belongs to Part II. Therefore, measures subject to scheduling under Articles XVI and XVII of the GATS should not be addressed by disciplines under Art. VI:4.27 As regards the Accountancy Disciplines, measures subject to scheduling under Art. XVI and XVII GATS were excluded from their application (see para. 1 and 2 of the Disciplines). 4. Regulatory Measures not Included There are regulatory measures that fall outside the scope of Art. VI as they are not addressed by Art. VI:4 GATS, such as for instance the question of the independence of regulators, universal service obligations, access to networks or essential facilities. Does Art. VI:4 preclude work on such measures and principles as the competence of the Working Party and the GATS Council is limited to the ones described in Art. VI:4 GATS (qualification requirements and procedures, technical standards and licensing requirements; licens28 ing procedures )? The comparison with TBT Annex 1 shows that in particular voluntary standards seem not to be addressed by Art. VI:4 GATS. D. Mutual Recognition Another area of interest is the mutual recognition of professional education, experiences, licenses and certifications. The area of education and certification is also part of national regulatory sovereignty and of interest for service suppliers from abroad. The GATS allows for recognition according to Art. VII GATS. Such recognition, which may be achieved through harmonization or otherwise, may be based upon an agreement or arrangement with the country concerned or may be accorded autonomously. In this regard the 27

28

See the Communication from the European Community and its Member States regarding a proposal for disciplines on licensing procedures, S/WPDR/W/25, para. 6. These categories are defined in a note by the WTO secretariat, Art. VI:4 of the GATS: Disciplines on Domestic Regulation Applicable to all Services, S/C/W/96.

382

Wolfgang Weiß

question was raised as to the possibility that such recognition agreements (MRA) detract from or even violate the MFN principle.29

References Dietrich Barth (1994), Das Allgemeine Übereinkommen über den internationalen Dienstleistungshandel (GATS), in: Europäische Zeitschrift für Wirtschaftsrecht 5 (1994), 455-460. Margareta Djordjevic (2002), Domestic Regulation and Free Trade in Services – A Balancing Act, in: Legal Issues of European Integration 29 (2002) No. 3, 305-322. John Hilary (2001), The Wrong Model: GATS, trade liberalisation and children’s right to health, London 2001. Clare Joy (2001), GATS and Domestic Regulation, July 2001 (available at http://www.gatswatch.org/GATSandDemocracy/ domestic.html). Markus Krajewski (2001), Democratic Legitimacy and Constitutional Perspectives of WTO Law, in: Journal of World Trade, 35 (2001) No. 1, 167-186. Markus Krajewski (2003), National Regulation and Trade Liberalization in Services, London / Den Haag (Kluwer) 2003. Mitsuo Matsushita / Thomas J. Schoenbaum / Petros C. Mavroïdis (2003), The World Trade Organization, Oxford (Oxford University Press) 2003. Christian Pitschas (2003), Allgemeines Übereinkommen über den Handel mit Dienstleistungen (GATS), in: Hans-Joachim Prieß / Georg M. Berrisch (Eds.), WTO-Handbuch, München (C.H. Beck) 2003, 495-564. Gaetan Verhoosel (2002), National Treatment and WTO Dispute Settlement, Oxford (Hart Publishing) 2002. Wolfgang Weiß (1999), Gibt es eine EU-Inländerdiskriminierung? Zur Kollision von Gemeinschaftsrecht mit Welthandelsrecht und Assoziationsrecht, in: Europarecht 34 (1999), 499-516.

29

Matsushita / Schoenbaum / Mavroïdis (2003), 244.

GATS and Domestic Regulation – A Threat to Democracy?

383

Jessica Woodroffe (2001), GATS and the Right to Regulate, July 2001 (available at http://www.gatswatch.org/GATSand Democracy/r2regulate.html).

Jacques H. J. Bourgeois

Post Cancún WTO TRIPs – A Bumpy Road I. II.

Introduction TRIPs and the Developing Country WTO Members A. Extending the Transitional Period B. Traditional Knowledge C. Access to Essential Medicines D. Capacity Building and Technical Assistance III. Extending the Scope A. Biotechnology and Relationship with Agreements on Biotechnology B. Improved Protection of Geographical Indications IV. Enforcement A. Causes of Inadequate Enforcement B. Cost of Enforcement C. Enforcement Through the WTO Dispute Settlement System VI. Conclusions References

385 387 388 388 389 392 392 392 393 394 394 395 396 396 397

I. Introduction The TRIPs Agreement is as of now the most comprehensive international agreement for the protection of intellectual property rights as to the number of States that are party to it – all WTO Members are bound by it, with transitional arrangements for developing country WTO Members. It is also the most comprehensive one as to coverage: it incorporates existing standards of intellectual property protection – the 1967 Paris Convention and the 1971 Berne Convention (literary and artistic works) – and it refers to the 1961 Rome Convention (performers and producers of phonograms and broadcasting organizations) and the 1989 Washington Agreement (integrated circuits), it provides additional standards for such intellectual property rights and includes protection of geographical indications

386

Jacques H. J. Bourgeois

and industrial designs. It also contains important and far-reaching obligations on enforcement of intellectual property rights in the national legal systems.1 Finally, being part of the Marrakesh Agreement, the TRIPs is subject to the WTO integrated dispute settlement system with its characteristic features of binding jurisdiction, time limits and enforcement systems. Agreed to originally by 123 countries, the TRIPs has helped to foster a broadly accepted understanding of rules governing IP protection and the increasing importance of IP in the global economy. This is perhaps most important when extended to States acceding to the WTO. Each potential WTO Member is aware that joining the WTO requires discussions at national level and, often, a realignment of IP rules and practices, if not introducing protection of IP, in order to successfully complete the accession process. By making TRIPs an integral part of the accession process, the WTO is helping to extend internationally accepted IP standards and furthering the creation of a level playing field in international trade. While the issues that did not allow progress to be made at the Cancún WTO Ministerial meeting did not relate to the TRIPs Agreement, one major TRIPs issue having been provisionally resolved before Cancún (see infra), this agreement is on the table of the Doha Development Round of Multilateral Trade Negotiations. There are a number of TRIPs issues that are, explicitly or implicitly, on the agenda. Not all such issues are dealt with in this short contribution. Those examined are presented under three headings: TRIPs and developing country WTO Members, extending the scope and enforcement. Before addressing these issues, it should be recalled that views vary considerably among countries about the legitimacy of IP protection, or, where they agree on such legitimacy, about the scope of the protection.2 An IP right that is perceived as perfectly legitimate in one country may be perceived as abusive in another one. In addi1

2

These obligations go into the nooks and cranies of national legal systems. For an interesting example involving the question whether the Netherlands procedure of ‘kort geding’ is in line with Article 56 TRIPs, see Court of Justice of the EC, Hermès v. FHT, C-53/96 [1998] ECR, I-3603. See e.g. P. Klemperer, America’s patent protection has gone too far, The Financial Times, March 2, 2004, a UK view shared widely in Western Europe.

Post Cancún WTO TRIPs – A Bumpy Road

387

tion, even within countries views on the respective merits of various IP rights may vary (e.g. patents v. copyright and trademarks). In negotiating clarifications of, or additions to, the TRIPs Agreement, WTO Members should be aware of this and avoid pushing other WTO Members to enter into commitments which will subsequently appear to be unsustainable. At the same time, certain WTO Members should keep a proper balance between their trade interests and the evolving views at national level on the legitimacy of the various IP rights. II. TRIPs and the Developing Country WTO Members The Uruguay Round of Multilateral Trade Negotiations and the TRIPs Agreement have led to and continue to feed a debate about the appropriateness of IP protection for developing country WTO Members. In this debate it is sometimes forgotten that IPRs are also in the interest of developing country WTO Members. Apart from the impact the absence of IP or insufficient IP protection in certain countries may have on exports to these countries and on imports from these countries, there is a link between IP protection in a country and transfer of technology to, and diffusion of technology in, that country. Moreover, IP protection is an important factor of foreign direct investment.3 For most developing country WTO Members complying with their TRIPs obligations often requires introducing unfamiliar or novel norms and on matters taking the back seat of their policy concerns. They accepted such obligations as the price to pay to be part of the ‘single undertaking’ of the Marrakesh Agreement establishing the WTO which was expected to give them benefits in trade in agricultural and textile products. In the context of the Doha Round, the question arises how, if not a balance better for, at least one more acceptable to developing country WTO Members could be achieved. 3

Edwin Mansfield, Intellectual Property Protection, Direct Investment and Technology Transfer: Germany, Japan and the United States, IFC Discussion Paper 27, The World Bank, Washington / DC 1995, cited by Elangi Botoy (2004). According to anecdotal evidence, an important developing country introduced IP protection of pharmaceuticals as a quid pro quo for sizeable investment with a view to producing pharmaceutical products.

388

Jacques H. J. Bourgeois

A. Extending the Transitional Period Developing country WTO Members benefit from a five, viz. ten years (already renewed once) transitional period to establish an IP protection system in line with the TRIPs Agreement. A transitional period is not only justified by the novelty of IP protection with its administrative and financial constraints and its costs for those WTO Members. In the view of those WTO Members, which point toward similar experiences of industrialized countries in the past, a transitional period is also required to allow them to develop a viable technological base without being constrained by IP protection: it appears doubtful that the current transitional period gives those WTO Members sufficient time in this respect.4 B. Traditional Knowledge Among the items put on the agenda of the TRIPs review is the protection of ‘traditional knowledge’.5 This concept refers to genetic resources, indigenous medicinal knowledge and other resources.6 The problem of protection is the following: traditional medicinal knowledge based on plants is usually not patentable – it is either obvious or it is in the public domain. However, a pharmaceutical product derived from plants via that traditional medicinal knowledge is patentable. And the patent goes to the pharmaceutical company. Under the TRIPs rules – in Daniel Gervais’ terminology – there is a double ‘exclusionary effect’: a ‘negative exclusionary effect’ in that traditional knowledge is not protected, and a ‘positive exclusionary effect’ in that IP rights are acquired by non-traditional knowledge holders. In dealing with these exclusionary effects, most of the options for introducing protection of traditional knowledge probably involve far-reaching adaptations of the TRIPs Agreement. In a paper submitted in 2002 to WIPO, the EC contemplated three possible approaches: protection via existing IP rights, prevent 4

5 6

Cfr. Elangi Botoy (2004); see also the interesting findings of Carlos Correa on the exclusion in the past of certain fields from patentability for developed WTO Members in Correa (1998); see also for an interesting Dutch example with the Netherlands suspending patent law protection from 1869 until 1910 Bronckers (1994), fn. 6. Doha Ministerial Declaration dated 20 November 2001, para 19 (WT/MIN(01)/DEC/1). For a good analysis, see Gervais (2003), with a wealth of references; also O’Connor (2003) with interesting examples.

Post Cancún WTO TRIPs – A Bumpy Road

389

inappropriate patenting or other types of misappropriation and ensure that benefits from interventions based on traditional knowledge are shared with providers of traditional knowledge or develop specific IP protection7. Quite apart from the difficulty on agreeing on an operational definition of traditional knowledge, it is difficult to see how it could be protected via existing IP rights, the concepts used by the TRIPs Agreement being what they are. Suffice it to mention: x many forms of traditional knowledge are in the public domain; x who is the ‘right holder’? x certain forms of traditional knowledge are by their nature hardly capable of being protected by current intellectual property rights, e.g. biological and genetic resources in their natural state. The second approach appears to be too complicated to be laid down in hard and fast rules. The third approach, a specific IP protection for traditional knowledge, will require much imagination and many ad hoc rules to include it in the TRIPs Agreement. Yet, by including traditional knowledge in IP protected by it, the TRIPs Agreement would be addressing also specific developing country interests. C. Access to Essential Medicines Infectious diseases kill over 10 million people each year, more than 90% of whom are in the developing world.8 The main causes of death in Africa, Asia and South-America are HIV/AIDS, respiratory infections, malaria and tuberculosis. There are several different aspects to the lack of access by developing countries to medicines essential to combat these diseases.9 The TRIPs Agreement recognizes that there are instances when the societal goods served by IP protection must give way to other pressing social needs: e.g. if in a case of public emergency, a pharmaceutical manufacturer is not able to produce enough of a needed medicine for which it has a patent, the WTO Member can under 7 8 9

WIPO/GRTKF/IC/3/16. WHO (2001), 144. See i.a. ’t Hoen (2003) that updates an earlier account of the TRIPsAgreement and medicines negotiation published in the Chicago Journal for International Law in 2002.

390

Jacques H. J. Bourgeois

TRIPs rules require that company to license its medicines to another domestic manufacturer in order to supplement any anticipated shortfall. This practice, commonly referred to as compulsory licensing, may seem particularly useful today, given heightened concerns over bio-terrorism for example. Given the spreading of Anthrax by mail in the United States, and the reliance on the drug Cipro to combat it, it is not hard to see how WTO Members might value the ability to resort to compulsory licensing. However, Article 31(f) of the TRIPs Agreement, which deals with the issue of compulsory licenses, provides that production under compulsory licensing must be predominantly for the domestic market. The basic problem is that many developing country WTO Members simply have no capacity to produce, even under license, the necessary medicines. For them, the only realistic means of access is direct import of generics from a WTO Member company producing generic medicines under compulsory licenses. So, there is a conflict: on the one hand, a massive public health problem, and on the other, a system that was not set up to adequately deal with global public health crises. During the past few years, the WTO Members have sought to resolve this issue. The 2001 Ministerial Declaration refers to TRIPs and public health. It states that ‘[we] agree that the TRIPs Agreement does not and should not prevent Members from taking measures to protect public health. Accordingly, we affirm that the Agreement can and should be interpreted and implemented in a manner supportive of WTO Members’ right to protect public health and, in particular, to promote access to medicines for all.’10 The Declaration further stated, in paragraph 6, that ‘We recognize that WTO Members with insufficient or no manufacturing capacities in the pharmaceutical sector could face difficulties in making effective use of compulsory licensing under the TRIPs Agreement.’11 The TRIPs Council was instructed to find an expeditious solution to this problem. 10 11

WTO, Doha Ministerial Declaration on the TRIPs Agreement and Public Health (WT/MIN(01)/DEC/2), para 4. Ibidem, para 6.

Post Cancún WTO TRIPs – A Bumpy Road

391

Several avenues were explored to find a solution, pending the conclusion of the Doha Round.12 Amending the TRIPs Agreement was too ambitious: for some WTO Members it must have appeared as a down payment on the conclusion of the Doha Round which remains uncertain. Adopting an ‘interpretation’ of Article 30 did not provide much legal certainty. Agreeing not to initiate dispute settlement proceedings in case of departure from Article 31(f) of the TRIPs Agreement was highly dubious from a systemic point of view. The only appropriate solution was a ‘waiver’ within the meaning of Article IX(3) of the WTO Agreement.13 This solution was adopted in a Decision of the WTO General Council, which allows WTO Members to grant compulsory licenses with a view to exporting pharmaceutical products to countries with no or insufficient manufacturing capacities.14 This decision is accompanied by a Chairman Statement reflecting the Members’ understanding on some particular issues. This decision has been criticized: it has been pointed out that compulsory licenses for the domestic market (Art. 31(f)) as opposed to licenses for export, are not subject to WTO procedures, that the limitation of the scope of the Decision – it should not be ‘an instrument to pursue industrial or commercial objectives’ – contradicts the objectives and principles of the TRIPs Agreement and that the measures provided to prevent re-exportation make this solution costly and inflexible.15 However, it is normal to provide measures against the use of the system to other ends than those for which it was established: the inclusion of compulsory licensing was not designed to pursue industrial or commercial objectives. Moreover, some safeguards against re-exportation are obviously needed otherwise compulsory licensing will not achieve its purpose.

12 13 14

15

On the issues and the negotiation see Abbott (2002); Bourgeois / Burns (2002); Correa (2002). Which empowers the WTO Ministerial Conference to waive an obligation imposed on a Member by any WTO Agreement. Decision of 30 August 2003, Implementation of Paragraph 6 of the Doha Declaration on the TRIPs Agreement and Public Health (WT/L/540 of 2 September 2003). For a good analysis of this decision and of the Chairman Statement see Vandoren / Van Eeckhoute (2003). I.a. Gopakumar (2004).

392

Jacques H. J. Bourgeois

The negotiations on how to incorporate this transitional solution in the TRIPs Agreement will offer the opportunity to fine-tune this system. Rather than provide for regulatory measures in developing countries to prevent re-exportation or to control re-importation in developed countries on the basis of ‘non-exhausted’ IP rights, one could try to ensure that the medicines are effectively used in developing countries, e.g. by insisting that they are administered to patients in health centers rather than distributed in pharmacies and, if need be, with the assistance of NGOs active in the health sector. D. Capacity Building and Technical Assistance Developed country WTO Members undertook in Article 66(2) of the TRIPs Agreement to provide incentives to their companies to transfer technology to least-developed country WTO Members. More is probably needed. Capacity building and technical assistance are increasingly on offer bilaterally and unilaterally or through international organizations to help developing countries to catch up and to adapt to the regulatory requirements of international trade. This is particularly relevant for the setting-up or the reform of IP protection systems and their management. Engineers and scientists, who are a scarce resource in certain developing countries, should be employed in research and production of goods rather than in evaluating patent applications. Shouldn’t this be done by consultants put at the disposal of these countries by developed WTO Members? III. Extending the Scope A. Biotechnology and Relationship with Agreements on Biotechnology The advances in the field of biotechnology and the relationship with a series of fairly recent international agreements and arrangements16 dealing with such advances may need to be addressed. Paragraph 19 of the Doha Declaration put on the negotiations agenda, on the one hand, the review of Article 27.3(b) of the TRIPs Agreement17 with reference to biotechnological inventions, plant 16 17

See i.a. the series of contributions in Cottier / Mavroïdis (2003). Which excludes from patentability i.a. ‘plants and animals other than micro-organisms, and essentially biological processes for the produc-

Post Cancún WTO TRIPs – A Bumpy Road

393

varieties and possible exclusion of patentability and, on the other, the relationship between the TRIPs Agreement and the Convention on Biological Diversity.18 The first one of these topics raises technical issues (such as: how to define the terms used in Article 27.3(b)? what is meant by sui generis protection of plant varieties?) and ethical issues relating to the patentability of life-forms. The second one raises the question whether and how the TRIPs Agreement can further the objectives of the Convention on Biodiversity, and more particularly whether IP can be an instrument for implementing the Convention on Biodiversity, i.a. on the sharing of benefits resulting from the use of genetic resources and on the disclosure of the geographical source and origin of genetic material. The further issue is how this can be achieved without altering the principles of the TRIPs Agreement. B. Improved Protection of Geographical Indications Protection of geographical indications is a specific European concern. Already in the Roman Empire, there was case law on the illegal use of geographical denomination of origin.19 Geographical indications are defined by the TRIPs Agreement as ‘indications which identify a good as originating in the territory of a Member, or a region or a locality in that territory, where a given quality, reputation or other characteristic of the good is essentially attributable to its geographical origin’ (Art. 22.1). They are basically to be protected against ‘any use which constitutes an act of unfair competition within the meaning of Article 10bis of the Paris Convention (1967) (Art. 22.2(b)) and against registration of ‘a trademark which contains or consists of a geographical indication with respect to goods not originating in the territory indicated’ provided this is such as to mislead the public as to the true place of origin (Art. 22.3). Interestingly, where there are

18 19

tion of plants or animals other than non-biological and microbiological processes’ but which also provides for the protection of plant varieties ‘either by patents or by an effective sui generic system or by a combination thereof’’. On the issues which Article 27.3(b) TRIPs raises, see i.a. Llewelyn, (2003), at 306-308. 31 ILM 818 (1992). Flury (2003), 6.

394

Jacques H. J. Bourgeois

homonymous geographical indications for wines, each indication is to be protected (subject to Art. 22.4). However, geographical indications can often not be registered as trademarks, to the extent that they lack distinctiveness. Protection via unfair competition laws, e.g. misrepresentation, may not be adequate, as geographical indications do more than simply guarantee the geographical origin. In line with its own legislation,20 the EC has proposed to introduce in the TRIPs Agreement a system of ‘registered geographical indications’ requiring not only proof of geographical origin but also compliance with product standards. This can be expected to be a difficult negotiation as such system has also a number of disadvantages.21 Quite apart from the paradox that poor European emigrants used in their new home countries their skills, applied their product standards and tended to use geographical indications of their country of origin, a system of registered geographical indications may grant a right of exclusive use over a word that is commonly used or has become generic in another jurisdiction.22 IV. Enforcement Much work has been done by WTO Members to adapt their national laws to the TRIPs Agreement. However, enforcing the rules appears to be a considerable challenge in a series of sectors. A. Causes of Inadequate Enforcement There seems to be two main causes of inadequate enforcement. First, intellectual property rights are not universally accepted and recognized by public opinion, not only in developing country WTO Members but also in developed country WTO Members, and the degree to which IP rights are accepted differs according to IP 20

21 22

Council Regulation (EC) 2081/92 on the protection of geographical indications and designations of origin for agricultural products and foodstuffs (OJ 1992 L 208/1). For a good analysis of the various issues and the pros and cons, see van Caenegem (2003). Such a phenomenon also occurred within the EC, where Catalan producers had to abandon the use of the word ‘crémant’, reserved for certain quality sparkling wines produced in France and Luxembourg, for their sparkling wines following Spain’s accession to the EC. See Codorniú SA v. Council, C-309/89 [1994] ECR, I-1853.

Post Cancún WTO TRIPs – A Bumpy Road

395

rights. Particularly in developing country WTO Members the proposition that certain ideas, technologies and processes should be protected is far from being accepted. Establishing such protection and enforcing it is rather unpopular and infringing the relevant rules is socially acceptable. The phenomenon is not limited to developing countries: piracy of compact disks and of DVDs is occurring on a large scale in developed countries as well.23 Second, the global explosion of information technology, media and access to technology has created stresses on the system in the face of continued technological innovation and the ever increasing sophistication of those who seek to infringe IP rights and sidestep protection. The most obvious, albeit not the only, sector affected is that of music and video recording, where the industry is waging a battle, i.a. against online downloading. Although the technical means to identify music file-swappers exist, e.g. with Internet search engines – ‘crawler/spider’ – targeting websites and users of peer-to-peer file-sharing networks sites, their lawfulness is being contested:24 there is a clear need for legislative solutions, if not at the international level, then at least at national level.25 The difficulty will be to strike the right balance between the rights of IP holders and the privacy rights of citizens, who are not all pirates. B. Cost of Enforcement Ensuring that a country’s legal, administrative and enforcement infrastructure is adequate to implement the TRIPs Agreement may represent a fairly considerable cost.26 Developing country WTO Members should be encouraged to pool their resources and to establish regional IP offices, as e.g. the Organisation Africaine de la Propriété Intellectuelle.27 23

24 25

26

It has been reported that US Secretary of Commerce Evans told his Chinese hosts last year that he bought a pirated DVD of Kill Bill in China for US$1. But he could probably find this pirated DVD in New York as well if not for US$1 then at least for US$5. E.g. The Financial Times, Two rulings hit crackdown in internet piracy, December 20/21, 2003. In the EC the matter could and should be dealt with under EC Directive of 24 October 1995 on the protection of individuals with regard to the processing of personal data and the free movement of such data (OJ 1995 L281/31). Lybbert (2002), at 310. There may also be an indirect cost: large transfers of income between countries (McCalman (2001)).

396

Jacques H. J. Bourgeois

C. Enforcement Through the WTO Dispute Settlement System As already indicated, one of the achievements of the TRIPs Agreement is the application of the WTO DSU to disputes on IP matters covered by the agreement, at least with respect to most TRIPs provisions. In fact, a fair number of dispute settlement proceedings have been initiated. At the time of writing, 22 disputes had been brought.28 One of them has created an interesting precedent as to remedies: in US – Copyright of Music in Bars,29 the US agreed to pay compensation to EC copyright holders. Yet, not subjecting some obligation to the DSU is seen by some as a way out of a diplomatic stalemate. The example set by Article 6 TRIPs Agreement (‘For the purposes of dispute settlement under this Agreement … nothing in this Agreement shall be used to address the issue of the exhaustion of intellectual property rights’) may be followed if at the end of the day no other way is found to reach agreement on a substantive rule. E.g. as already indicated, at some point in time this was one of the solutions contemplated for the purposes of implementing paragraph 6 of the Doha Declaration on the TRIPs Agreement and Public Health. Such a solution might be an elegant way out from a diplomatic ‘impasse’. From a lawyer’s perspective, excluding the application of the DSU to a given WTO obligation is tantamount to degrading the obligation agreed upon to what John Jackson once called ‘a norm of aspiration’. V. Conclusion This brief and incomplete review of issues arising in the review of the TRIPs Agreement in the framework of the Doha Round has highlighted the need to better take account of the specific difficulties developing country WTO Members are facing in implementing the TRIPs Agreement and to include a form of IP protection of traditional knowledge offering more balanced benefits. It has drawn the attention to the issue of compulsory licensing of pharmaceutical patents with a view to export. It has dealt with the challenge of including a system of registered geographical indications in the TRIPs Agreement. It has addressed the enforcement issue in some of its different aspects. 27 28 29

See Nwauche (2003). For an analysis, see Abbott (2004). WT/DS160/R of 15 June 2000.

Post Cancún WTO TRIPs – A Bumpy Road

397

Finding reasonable and balanced solutions to these issues at WTO level will require considerable efforts and skills, different from those needed for GATT and GATS matters. It should be borne in mind that the legal structures of the TRIPs Agreement differ in many respects from those of the GATT and of the GATS, in that i.a. it includes not only negative (prohibitive) rules but also a large number of positive (affirmative) obligations to introduce measures for the protection of IP rights and it is designed to protect private rights.30

References Frederick M. Abbott (2002), Compulsory Licensing for Public Health Needs: the TRIPs Agenda at the WTO after the Doha Declaration on Public Health, Quaker U.N. Office Occasional Paper No. 9, Geneva 2002. Frederick M. Abbott (2004), WTO Dispute Settlement Practice Relating to the Agreement in Trade-Related Intellectual Property Rights, in: Federico Ortino / Ernst-Ulrich Petersmann (eds.), The WTO Dispute Settlement System 1995-2003, Den Haag / London / New York (Kluwer Law International) 2004, 421453. Marco C. E. J. Bronckers (1994), The Impact of TRIPs: Intellectual Property Protection in Developing Countries, in: Common Market Law Review 31 (1994), 1245-1281. Jacques H. J. Bourgeois / Thadeus J. Burns (2002), Implementing Paragraph 6 of the Doha Declaration on TRIPs and Public Health – The Waiver Solution, in: The Journal of World Intellectual Property 5 (2002) No. 5, 835-864. Carlos M. Correa (1998), Implementing the TRIPs Agreement in the Patent Field – Options for Developing Countries, in: The Journal of World Intellectual Property 1 (1998) No. 1, 75-99. Carlos M. Correa (2002), Implications of the Doha Declaration on the TRIPs Agreement and Public Health, WHO Health Economics and Drugs EDM Series No. 12, Geneva 2002.

30

See Petersmann (2003), at 23.

398

Jacques H. J. Bourgeois

Thomas Cottier / Petros C. Mavroïdis (eds.) (2003), Intellectual Property, Trade, Competition and Sustainable Development (= World Trade Forum Vol. 3), Ann Arbor / MI (The University of Michigan Press) 2003. Ituku Elangi Botoy (2004), From the Paris Convention to the TRIPs Agreement – A One-Hundred-and-Twelve-Year Transitional Period for the Industrialized Countries, in: The Journal of World Intellectual Property 7 (2004) No. 1, 115-130. Andrea E. Flury (2003), Grundprobleme des Rechts der geografischen Herkunftsbezeichnungen, Bern / Stuttgart / Wien (Haupt) 2003. Daniel Gervais (2003), TRIPs, Doha and Traditional Knowledge, in: The Journal of World Intellectual Property 6 (2003) No. 3, 403-419. K. M. Gopakumar (2004), The WTO Deal on Cheap Drugs. A Critique, in: The Journal of World Intellectual Property 7 (2004) No. 1, 99-113. Margaret Llewelyn (2003), Which Rules in World Trade Law – Patents or Plant Variety Protection?, in: Cottier / Mavroïdis (2003), 303-339. Travis J. Lybbert (2002), On assessing the cost of TRIPs implementation, in: World Trade Review 1 (2002) No. 3, 309-321. Phillip McCalman (2001), Reaping What You Sow: An Empirical Analysis of International Patent Harmonization, in: Journal of International Economics 55 (2001), 161-186. Enyinna S. Nwauche (2003), An Evaluation of the African Regional Intellectual Property Rights System, in: The Journal of World Intellectual Property 6 (2003) No. 1, 101-138. Bernard O’Connor (2003), Protecting Traditional Knowledge. An Overview of a Developing Area of Intellectual Property Law, in: The Journal of World Intellectual Property 6 (2003) No. 5, 677-697. Ernst-Ulrich Petersmann (2003), From Negative to Positive Integration in the WTO: the TRIPs Agreement and the WTO Constitution, in: Cottier / Mavroïdis (2003), 21-52. Ellen F. M. ’t Hoen (2003), TRIPs, Pharmaceutical Patents and Access to Essential Medicines: Seattle, Doha and Beyond, avail-

Post Cancún WTO TRIPs – A Bumpy Road

399

able at http://www.accessmed-msf.org/documents/chicago journalthoen.pdf. William van Caenegem (2003), Registered Geographical Indications. Between Intellectual Property and Rural Policy, in: The Journal of World Intellectual Property 6 (2003) No. 5, 699-719 and No. 6, 861-874. Paul Vandoren / Jean Charles Van Eeckhoute (2003), The WTO decision on paragraph 6 of the Doha Declaration on the TRIPS Agreement and public health, in: The Journal of World Intellectual Property 6 (2003) No. 6, 779-793. WHO (2001), The World Health Report 2001, Geneva 2001.

Abbreviations ACP: ADP: AFTA: AG: AGOA: AIDS: al.: AMS: AoA: APEC: Art.: ASEAN: ATC: ATP: AU: AVE:

African, Caribbean and Pacific Group of States Antidumping Agreement ASEAN Free Trade Area Attorney General (US) African Growth and Opportunity Act Acquired Immuno Deficiency Syndrome alii Aggregate Measurement of Support Agreement on Agriculture Asia-Pacific Economic Cooperation Article Association of South East Asian Nations Agreement on Textiles and Clothing Andean Trade Preference Agreement (‘Andean Pact’) African Union Ad valorem tariff equivalent

BI: BISD:

Bahnzhaf power index Basic Instruments and Selected Documents (of the GATT) Balance of payments IMF Balance of payments manual

BoP: BPM:

CACM: Central American Common Market CAFOD: Catholic Agency for Overseas Development CAP: Common Agricultural Policy (of the EU) CARICOM: Caribbean Common Market CCCT: Commonwealth Caribbean Countries Tariff CEEC: Central and Eastern European Countries CEFTA: Central European Free Trade Agreement CEMAC: Economic and Monetary Community of Central African States (cf UDEAC) cf: confer

402

Abbreviations

CGE: Computable general equilibrium CHF: Swiss Francs COMECON: Council for Mutual Economic Cooperation COMESA: Common Market for Eastern and Southern Africa CVD: Countervailing Duty DDA: Doc.: DOTS: DSB: DSR: DSU: DVD:

Doha Development Agenda Document IMF Direction of Trade Statistics Dispute Settlement Body Dispute Settlement Reports Dispute Settlement Understanding Digital Video Disk

E: EBA: EC: ECCAS: ECJ: ECOWAS: ECR: ECU: EEC: EFTA: e.g.: EMU: etc.: EU: Eurostat:

East Everything but Arms European Community Economic Community of Central African States European Court of Justice Economic Community of West African States European Court Reports (Reports of Cases before the Court of Justice and the Court of First Instance) European Currency Unit European Economic Community European Free Trade Area Exempli gratia Economic and Monetary Union (of the EU) et cetera European Union Statistical Office of the European Communities

FAO: fn.: FSC: FTA: FTAA:

Food and Agriculture Organization of the UN footnote Foreign sales corporation Free trade area Free Trade Area of the Americas

G20(+): G22: G33: G90:

Group of twenty(-plus) (countries in the WTO) Group of twenty-two (countries in the WTO) Group of thirty-three (countries in the WTO) Group of ninety (countries in the WTO)

Abbreviations

403

GATS: GATT: GDP: GI: GNP: GSP: GTAP:

General Agreement on Trade in Services General Agreement on Tariffs and Trade Gross Domestic Product Geographical indication Gross National Product Generalized System of Tariff Preferences Global Trade Analysis Project

HFCS: HIV: H.R.

High-Fructose Corn Syrup Human Immuno-deficiency Virus House Reports (of the US Congress)

i.a.: ICJ: ICJ REP: ICSID:

IFS: ILM: ILO: IMF: IP: IPR: ITA: ITC: ITLOS:

inter alia International Court of Justice ICJ Reports International Centre for Settlement of Investment Disputes International Center for Trade and Sustainable Development id est International Finance Corporation of the World Bank Group IMF’s International Financial Statistics International Legal Materials International Labour Organization International Monetary Fund Intellectual property Intellectual property right Information Technology Agreement International Trade Centre (of UNCTAD and WTO) International Tribunal for the Law of the Sea

JSEPA:

Japan-Singapore Economic Partnership Agreement

lit.: LDC:

litera Least developed country

ICTSD: i.e.: IFC:

MAS: Mutually agreed solution MEA: Multilateral Environmental Agreement MERCOSUR: Mercado Común del Sur (‘Southern Common Market’)

404

Abbreviations

MDG: MFA: MFN:

Millenium Development Goal Multifibre Arrangement Most favourite nation (treatment)

N: NAFTA: NGO: NIC: NTB: NTM:

North North American Free Trade Area Non-governmental organization Newly industrializing country Non-tariff barrier Non-tariff measure

ODA: OECD:

Official Development Assistance Organisation for Economic Co-operation and Development Official Journal of the European Communities One stage least-square OSlo PARis Convention / Commission for the Protection of the Marine Environment of the NorthEast Atlantic

OJ: OLS: OSPAR:

Para.: PCIJ: PPP:

Paragraph Permanent Court of International Justice Purchasing power parity

QR:

Quota rent

R&D: RTA:

Research and development Regional trade agreement

S: S&D: SAARC: SADC: SAL: SAR: SCM: SDR: SDT: Ser.: SM: SP:

South Special and differential treatment (SDT) South Asian Association for Regional Cooperation Southern African Development Community Structural Adjustment Loans Special Administrative Region Subsidies and Countervailing Measures Special Drawing Right (of the IMF) Special and differential treatment (S&D) Series Simple majority Special product

Abbreviations

SPS: SSG: SSI: SSM:

Sanitary and Phytosanitary Measures Special safeguard Shapley-Shubik index Special Safeguard Mechanism

T&C: TBT: TEC: TFP: TRAINS: TRIMs: TRIPS: TRQ:

Textiles and clothing Technical barriers to trade Treaty establishing the European Community Total factor productivity Trade Analysis and Information System Trade Related Investment Measures Trade-related aspects of intellectual property rights Tariff rate quota

UDEAC:

405

Communauté économique et monétaire de l’Afrique central (cf CEMAC) UEMOA: Union Économique et Monétaire Ouest Africaine UK: United Kingdom UN(O): United Nations (Organization) UNCLOS: United Nations Convention on the Law of the Sea UNCTAD: United Nations Conference on Trade and Development UR: Uruguay Round URAA: Uruguay Round Agreement on Agriculture US(A): United States (of America) USITC: United States International Trade Commission v.: VCLT:

versus Vienna Convention on the Law of Treaties

W: WDI: WHO: WIPO: WITS: WTO:

West World Bank World Development Indicators World Health Organization World Intellectual Property Organization World Integrated Trade Solution World Trade Organization

The Authors of this Volume Lorand Bartels, Lawyer, holds degrees in English Literature and Law from the University of New South Wales, Australia, and a doctorate in Law from the European University Institute in Florence, Italy. Before joining the University of Cambridge / UK, he served as a Lecturer in International Economic Law at the University of Edinburgh. Currently he is Humboldt Fellow at Max Planck Institute for International Law in Heidelberg. Jacques Bourgeois, Lawyer, professor at the College of Europe in Bruges. Before, he worked with the EC Commission’s legal service, with the main work focuses trade and antitrust policies. Fritz Breuss, Economist, is Jean Monnet-Professor for the economy of European integration at the Europainstitut of Wirtschaftsuniversität, Vienna, and collaborator of the Austrian Institute of Economic Research – WiFo. Since 2002 he is President of ECSA Austria. Thomas Cottier, Lawyer, is Managing Director of the World Trade Institute and Professor for European and International Economic Law at University of Berne, Switzerland. Since the 1980s he has been member of the Swiss GATT-delegation and Member and chairman of various GATT- and WTO-panels. William J. Davey, Lawyer, is Guy Raymond Jones-Professor of Law at the University of Illinois College of Law in UrbanaChampaign, teaching in international trade law, European Union law, and international business transactions. From 1995-99 he served as the Director of the Legal Affairs Division of the WTO. Piet Eeckhout, Lawyer, is Professor of European Law King’s College London, Director of its Centre of European Law, and Visiting Professor at College of Europe in Bruges, Belgium. Before he worked with Advocate General Francis Jacobs at the European Court of Justice in Luxembourg.

408

The Authors of this Volume

Stefan Griller, Lawyer, is Jean Monnet-Professor for Public Law with special regard to European Law at the Europainstitut of Wirtschaftsuniversität, Vienna. He was the founding President of ECSA Austria in 1996 and since 2002 serves as the organization’s Secretary General. Markus F. Hofreither, Economist, is Professor at the Institute for Sustainable Economic Development of the University of Natural Resources and Applied Life Sciences in Vienna. Apart from many other respective assignments, he served as an expert in agricultural issues for the European Commission. Gabrielle Marceau, Lawyer, is Member of the Cabinet of Pascal Lamy, the Director General of the WTO, and Visiting Lecturer at the Geneva-based Graduate Institute of International Studies. Since May 2005, she is associate professor at the Institute for International Public Law of the University of Geneva. Alan Matthews, Economist, is Jean Monnet-Professor of European Agricultural Policy at the Trinity College Dublin, Ireland. Formerly, he worked as a practitioner in agriculture in Africa and Ireland. Since 2003 he is Member of the Executive Committee of the Agricultural Economics Society. Peter-Christian Müller-Graff, Lawyer, is Jean Monnet-Professor for German and European Company- and Economic Law at the University of Heidelberg, Germany. Since 2002 he serves as Chairman of the Executive Board of ECSA Austria’s German partner-organization Arbeitskreis Europäische Integration. Joost Pauwelyn, Lawyer, is Associate Professor at the Duke University School of Law in Durham / NC, USA. Since 2002, he served – among others – as a Legal Affairs Officer for the WTO, first, in the Legal Affairs Division, then in the Appellate Body Secretariat. Satoko Takenoshita, Economist, graduated in economics at Tokyo’s Hosei University and in international relations in Bologna, Italy. When she wrote her contribution to this book with Thomas Cottier she worked with the World Trade Institute in Berne, Switzerland. Anastasios Tomazos, Lawyer, at the time of writing the paper in the volume at hand was working with the Legal Affairs Division of the WTO Secretariat. At present he working with Skadden, Arps, Slate, Meagher & Flom LLP in New York.

The Authors of this Volume

409

Erich Vranes, Lawyer, is working in the Legal Department of the Europainstitut of Wirtschaftsuniversität, Vienna, since 2000. Wolfgang Weiß, Lawyer, Professor for Public Law at the University of Erlangen-Nuremberg, Germany, and Reader in International Law at the Oxford Brookes University. His volume ‘Welthandelsrecht’ (co-authored with Ch. Herrmann) is the standard textbook on world trade law in the German speaking area.