3 downloads 51 Views 775KB Size Report
Edwin B. Parker professor of law at Columbia Law School, New York, professor at the Law Faculty of the ... WTO Member to the rest of the WTO Membership, as far as the opening of its ... suppliers or the requirements of an economic needs test ..... shall be fined under this title or imprisoned not more than two years, or both.



1 Highway XVI re-visited: the road from non discrimination to

market access in GATS P E T R O S C. M A V R O I D I S 25 Psychology and opposition to free trade

S IMON K EMP 45 The European Union and WTO law: a nexus of reactive, coactive,

and proactive approaches A NTONIS A NTONIADIS Dispute settlement corner 89 United States – Definitive Safeguard Measures on Imports of Certain

Steel Products G E N E M. G R O S S M A N


A L A N O. S Y K E S

Review essays 123 Economic justice in an unfair world: towards a level playing field

M ATHIAS R ISSE 135 Trade negotiations and developing countries

J. M I C H A E L F I N G E R Book reviews 149 The WTO and Sustainable Development by Gary P. Sampson

D AVID R OBERTSON 151 Economic Development and Multilateral Trade Cooperation by Simon J. Evenett

and Bernard M. Hoekman K E V I N P. G A L L A G H E R

World Trade Review (2007), 6 : 1, 1–23 Printed in the United Kingdom f Petros C. Mavroidis doi:10.1017/S1474745606003077

Highway XVI re-visited: the road from non-discrimination to market access in GATS P E T R O S C. M A V R O I D I S* Edwin B. Parker professor of law at Columbia Law School, New York, professor at the Law Faculty of the Un. of Neuchatel and CEPR

Abstract : The GATS (General Agreement on Trade in Services) is a highly complicated legal instrument and prone to various interpretations because of the embedded ambiguity. It should not come as a surprise that in the context of all disputes concerning services, WTO panels and the Appellate Body reached diametrically opposite conclusions on the same issues. Un-appealed panel reports, on the other hand, have not been welcome either. One of the major, if not the major, issue is the legal relationship between Arts. XVI and XVII GATS, which regulate market access. The manner in which this relationship has been interpreted inescapably leads to constructing the GATS as a move beyond negative integration. This is at odds with the intent of the founding fathers, the letter and the spirit of the GATS itself. Subjecting Art. XVI to Art. XVII GATS guarantees respect of the negative integration character of GATS. It thus avoids internationalizing issues that WTO Members want to keep in their domaine reserve´. Counter-intuitively probably, it allows for more trade liberalization. Using the unfortunate US–Gambling report as an example, this report suggests an approach to understanding Art. XVI as a sub-set of Art. XVII GATS.

1. The weight of the non-discrimination obligation in scheduling GATS commitments This paper deals with an issue which, at first glance, might seem quite narrow to the wider public but which has very important repercussions for the understanding and the functioning of the WTO General Agreement on Trade in Services * Numerous discussions with Bernard Hoekman, Henrik Horn, Patrick Low, Aaditya Mattoo, Alan Sykes, Bob Staiger have helped me shape my thinking on this score. I am also thankful to anonymous referees for their useful comments. Juan-Alberto Marchetti has very generously spent time and effort discussing all issues reflected in this paper. His very valuable comments helped me address many of the shortcomings in previous drafts and develop my understanding of and about GATS in general. He is in many ways the virtual co-author of this paper, our disagreements on some of the issues discussed notwithstanding.



P E T R O S C. M A V R O I D I S

(GATS) :1 what is – and what should be – the relationship between Art. XVI and Art. XVII GATS ? The first provision identifies six types of measures, which, absent contrary explicit indication to this effect in a Member’s schedule of commitments, cannot be enforced ; the second is the GATS re-incarnation of the national treatment-obligation. The question is whether the former reflects measures that apply to both foreign and domestic suppliers. The legal response to this question is judgemade law, and is affirmative. It is submitted that the current approach is wrong, and, for good policy but also legal reasons, the response should be negative : market access restrictions coming under the purview of Art. XVI GATS, to the extent included in a schedule of concessions, should be interpreted as concerning foreign suppliers only.2 The paper starts with a description of the two provisions and the manner in which their relationship has been understood in wider WTO practice, and in caselaw, more specifically (Section 2). With the positive understanding of this relationship established, I move to explain the reasons for my disagreement with current practice (Section 3). I then test my preferred approach on imaginary transactions (Section 4). Section 5 concludes.

2. The current understanding

2.1 What did negotiators have in mind?3 Even a brief perusal of negotiating documents suggests that there is a lot of disagreement among WTO Members as to the exact nature of Art. XVI GATS, and its relationship with Art. XVII GATS. So much is evidenced in the discussions held by WTO Members in the context of a technical review of the GATS carried out between 2000 and 2004, which focused primarily on Art. XX.2 GATS, and during which a substantial part of the negotiating history of the provisions re-emerged.4 Suffice to say that such discussions reveal that the negotiators were not on the same page even after they had agreed on the definitive legal texts of Arts. XVI and XVII GATS.

2.2 The relevant legal texts The schedule of concessions circumscribes the contractual promise given by each WTO Member to the rest of the WTO Membership, as far as the opening of its services market is concerned. WTO Members wishing to liberalize their market 1 For an excellent brief description of GATS, see Hoekman and Kostecki (2001). Hoekman (1996) offers an exemplary appraisal of the GATS’ impact. 2 The GATS of course concerns both services and services suppliers. When using the latter term only, we mean to cover the former as well. 3 We are of course not unaware of the limited legal significance of travaux pre´paratoires (Art. 32 of the Vienna Convention on the Law of Treaties, VCLT). 4 See ‘ Consideration of issues relating to Article XX:2 of the GATS’, Report by the Chairman of the Committee on Specific Commitments, WTO document S/C/W/237, dated 24 March 2004.

Road from non-discrimination to market access in GATS 3

will first indicate the sector where they will enter their commitments,5 and then will enter their commitments. Commitments can be entered under three different provisions : (a) Arts. XVI GATS (Market Access) ; (b) Art. XVII GATS (National Treatment) ; (c) and Art. XVIII GATS (Additional Commitments).6 Art. XVI.1 GATS reads : With respect to market access through the modes of supply identified in Article I, each Member shall accord services and service suppliers of any other Member treatment no less favourable than that provided for under the terms, limitations and conditions agreed and specified in its Schedule.

Art. XVI.2 GATS prohibits, in principle, recourse to six types of measures:7 (a) limitations on the number of suppliers ; (b) limitations on the total value of service transactions or assets ; (c) limitations on the total number of service operations or on the total quantity of service output ; (d) limitations on the total number of natural persons that may be employed ; (e) measures which restrict or require specific types of legal entity or joint venture ; (f) limitations on the participation of foreign capital. The wording of Art. XVI.1 GATS on its face suggests that Art. XVI GATS as a whole, and the list of measures included in the second paragraph, concern foreign suppliers only, and the wording of Art. XVI.2 GATS leaves no doubt that the prohibition of recourse to one of the measures included in its body is only in principle : the measures which a Members shall not maintain or adopt _ unless otherwise specified in its Schedule.

Consequently, WTO Members wishing to impose any of the six measures embedded in Art. XVI.2 GATS will have to indicate their wish in their schedule of commitments. Absent such indication, recourse to such measures is impermissible.8

5 Although there is not a mandatory sectoral classification, in most cases WTO Members have resorted to a classification prepared during the Uruguay Round by the then GATT Secretariat, and known as the ‘Services Sectoral Classification List’ (document MTN.GNS/W/120). This classification uses as a crossreference a product classification developed by the UN, and known as the ‘ Provisional Central Product Classification’ (Provisional CPC). 6 For the purposes of our discussion in this paper, this provision is immaterial. 7 See the excellent analysis of Art. XVI GATS in Ortino (2006). 8 At this stage, we do not want to prejudge the issue whether indication of both discriminatory as well as non-discriminatory measures is the condition for opposing such measures to foreign suppliers. This is


P E T R O S C. M A V R O I D I S

Table 1. Article XVI: limitations on market access Market access limitations (a) Limitations on the number of service suppliers whether in the form of numerical quotas, monopolies, exclusive service suppliers or the requirements of an economic needs test (b) Limitations on the total value of service transactions or assets in the form of numerical quotas or the requirement of an economic needs test (c) Limitations on the total number of service operations or on the total quantity of service output expressed in terms of designated numerical units in the form of quotas or the requirement of an economic needs test1 (d) Limitations on the total number of natural persons that may be employed in a particular service sector or that a service supplier may employ and who are necessary for, and directly related to, the supply of a specific service in the form of numerical quotas or the requirement of an economic needs test (e) Measures which restrict or require specific types of legal entity or joint venture through which a service supplier may supply a service (f) Limitations on the participation of foreign capital in terms of maximum percentage limit on foreign share-holding or the total value of individual or aggregate foreign investment

Example Licences for new restaurants subject to economic needs test based on population density Foreign bank subsidiaries limited to x% of total domestic assets of all banks Restrictions on the broadcasting time available for foreign films

Foreign labour should not exceed x% of the work force and/or not account for more than y% of total wages Commercial presence excludes representative offices Foreign equity participation in domestic insurance companies should not exceed x %

Note: 1 Subparagraph 2(c) does not cover measures of a Member which limit inputs for the supply of services.

Limitations and restrictions mentioned in Art. XVI.2 GATS can be entered with respect to one or more mode(s) of supply of the service concerned.9 Table 1, drawn from the Scheduling Guidelines,10 provides an illustration of limitations on market access. Art. XVII GATS is a classic reproduction of the national treatment principle, which WTO Members can, but are not obliged to give to their trading partners, even if they have undertaken specific commitments in a given sector. The relationship between Art. XVI and Art. XVII GATS is not explicitly clarified in the GATS except for an oblique reference in Art. XX.2 GATS which reads : Measures inconsistent with both Articles XVI and XVII shall be inscribed in the column relating to Article XVI. In this case the inscription will be considered to provide a condition or qualification to Article XVII as well.

indeed one of the questions we will be dealing with at a later stage and very much the issue discussed in the US–Gambling litigation. Suffice to say that at any rate discriminatory measures will have to be enlisted anyway, otherwise they cannot be opposed to those seeking market access. 9 According to the Scheduling Guidelines (document S/L/92). 10 Their legal import is discussed infra.

Road from non-discrimination to market access in GATS 5

This issue has been discussed in a series of documents surrounding the GATS negotiations as well as in subsequent case law. We turn to a study of negotiated documents in what immediately follows (Section 2.3), before we move to examine the relevant case law (Section 2.4).

2.3 The Scheduling Guidelines The Scheduling Guidelines is a document aimed at assisting Uruguay Round participants to schedule their commitments in a GATS-consistent manner. Critical issues for understanding both the schedules and the relationship between Art. XVI GATS and Art. XVII GATS are spelled out in the Scheduling Guidelines : absent understanding of expressions such as ‘ none ’, ‘ unbound ’, it is quite hard to understand what exactly has been committed. These terms have been used for the first time and interpreted in the Scheduling Guidelines.11 The original document (the 1993 Scheduling Guidelines) – a WTO Secretariat Note that had been circulated to the negotiators, its purpose being to facilitate the ongoing negotiations – was used for scheduling commitments during the Uruguay round negotiations.12 On 23 March 2001, the WTO Members adopted formally in the context of the CTS (Council for Trade in Services) a new, revised document (2001 Scheduling Guidelines).13 This document also reflects guidelines for the scheduling of specific commitments to be used for the purposes of scheduling commitments as of the date of its adoption. The 2001 Scheduling Guidelines contains no substantive change or deviation from the 1993 Scheduling Guidelines ; it re-states the 1993 Scheduling Guidelines and adds a few Annexes and an illustrative list of limitations to National Treatment. Its main thrust is described in its first paragraph in the following terms : This note is intended to assist in the preparation of offers, requests and national schedules of specific commitments. Its objective is to explain, in a concise manner, how specific commitments should be set out in schedules in order to achieve precision and clarity. It is based on the view that a common format for schedules as well as standardization of the terms used in schedules are necessary to ensure comparable and unambiguous commitments. The note cannot answer every question that might occur to persons responsible for scheduling specific commitments ; it does attempt to answer those questions which are most likely to arise. The answers should not be considered as a legal interpretation of the GATS. (emphasis added).14

11 This is not to assert anything on the legal significance of the Scheduling Guidelines, an issue to which we return in what follows. 12 See WTO Doc. (MTN.GNS/W/164 and Add. 1). Note that whereas the corresponding instrument to the CPC GATT-legal instrument is the HS, there is no corresponding (in the GATT-context) instrument to the Scheduling Guidelines. 13 WTO Doc. S/L/92. 14 As the AB put it in a nutshell in its US–Gambling report (·173), the 1993 Scheduling Guidelines address two questions: what items should be included, and how should included items be scheduled.


P E T R O S C. M A V R O I D I S

The relationship between Art. XVI and Art. XVII GATS in the Scheduling Guidelines When discussing the list of measures that are reflected in Art. XVI GATS as well as the addressees of the measures, the document states that The list is exhaustive and includes measures which may also be discriminatory according to the national treatment standard (Article XVII). In other words, all measures falling under any of the categories listed in Article XVI:2 must be scheduled, whether or not such measures are discriminatory according to the national treatment standard of Article XVII.15

Hence, this document takes the view that even measures which apply on a nondiscriminatory basis should be scheduled under Art. XVI GATS. This paragraph at the very least adds to the content of Art. XX.2 GATS : whereas the GATS provision refers to measures which are inconsistent with both Art. XVII and XVI GATS (hence, discriminatory), this paragraph requests from WTO Members to schedule in their Art. XVI GATS column non-discriminatory measures as well. Before, however, we pronounce on the well-foundedness of this approach, we should first delve into the legal significance of the Scheduling Guidelines. The legal significance of the Scheduling Guidelines The legal relevance of the (1993 and 2001) Scheduling Guidelines has already been addressed in case law. Before moving to that discussion, however, let us first see how this issue is addressed in the text of the Scheduling Guidelines. The already quoted first paragraph of the 2001 Scheduling Guidelines reads in pertinent part : The note cannot answer every question that might occur to persons responsible for scheduling commitments ; it does attempt to answer those questions which are most likely to arise. The answers should not be considered as an authoritative legal interpretation of the GATS.

By the same token, the decision (WTO Doc. S/L/91) adopting the 2001 Scheduling Guidelines reads 1. To adopt the Guidelines for the Scheduling of Specific Commitments under the General Agreement on Trade in Services contained in document S/CSC/W/30 as a non-binding set of guidelines. 2. Members are invited to follow these guidelines on a voluntary basis in the future scheduling of their specific commitments, in order to promote their precision and clarity. 3. These guidelines shall not modify any rights or obligations of the Members under the GATS. 15 See ·8 of the Scheduling Guidelines. It is not our purpose to address in a comprehensive manner the question whether the six measures mentioned in Art. XVI.2 GATS are the only measures a WTO Members can introduce in its schedule of concessions. We believe, however, that this should not be the case. Both the text of Art. XVI GATS (referring to limitations, conditions etc.) as well as the text of Art. XX.1 GATS posit the opposite conclusion.

Road from non-discrimination to market access in GATS 7

A WTO Secretariat Note underscores that the 2001 Scheduling Guidelines were not enacted with the aim to affect the balance of rights and obligations as struck by the founding fathers of the GATS.16 Two WTO panels have dealt with the issue of the legal value of the (1993 and 2001) Scheduling Guidelines. Both panels held for the proposition that the Scheduling Guidelines have important ramifications in the interpretation of the GATS. While doing that however, the two panels have reached irreconcilable conclusions on this issue. First, the report on Mexico–Telecoms reflects the view that the Scheduling Guidelines are an integral part of the Travaux Pre´paratoires (··7.66–7.67). Then, the panel report on US–Gambling held for the proposition that the Scheduling Guidelines form an integral part of the context in the sense of Art. 31 VCLT (·6.82). The different attitudes can have important legal repercussions : if the former view were to be correct, the relevance of the Scheduling Guidelines becomes a second-order concern since WTO adjudicating bodies will have recourse to them only under the conditions laid down in Art. 32 VCLT. Conversely, if the latter panel’s view prevails, WTO adjudicating bodies will have to have recourse to Scheduling Guidelines any time a scheduling issue is brought before them. The AB settled the score in its report on US–Gambling. Stating first that the Scheduling Guidelines constitute no agreement between the parties, it rejected the panel’s view that they should be considered as context in the VCLT-sense of the term (··174–175). It then went on to express its agreement with the view of the parties to the dispute that the Scheduling Guidelines should be understood as a supplementary means of interpretation, in the sense of Art. 32 VCLT.17 From a functional perspective, the consequence of this taxonomy is that recourse to the Scheduling Guidelines is a matter of discretion of the WTO judge.18 16 See ·4 in WTO Doc. S/CSC/W/12 of 10 October 1997. 17 As noted, however, the Scheduling Guidelines for trade in goods (the Harmonized system) and the Scheduling Guidelines for trade in services have different legal value. 18 The explicit rejection of the panel’s thesis that the Scheduling Guidelines should be considered in context, coupled with its affirmative statement that the Scheduling Guidelines should be viewed as supplementary means of interpretation, makes it clear that the only appropriate legal conclusion is that recourse to the Scheduling Guidelines from now on should be a matter of discretion for the WTO judge. This does not mean that recourse will not happen in future cases. It only means that there is no legal compulsion for the judge to do that. The opposite would have been true had the AB confirmed the panel’s approach. Krajewski (2005) believes that recourse might be necessary more often than not. I suppose it all depends on the clarity of the terms used in any given schedule. The twin question is of course, assuming recourse, is the WTO judge bound to follow the Scheduling Guidelines in a given matter? Once again, it is odd that this should be the case across transactions. There are instances where indeed the Scheduling Guidelines are of decisive help: for example, the meaning of the terms NONE and UNBOUND featuring in schedules. Such are cases where the Scheduling Guidelines were intended to be used as a gap-filling instrument (the terms NONE, UNBOUND being nowhere else defined in GATS). This is not necessarily the case with respect to other issues, such as the one treated in this paper where the Scheduling Guidelines reflect an opinion which (at the very least) could be contradicting the letter of the law. In such cases, were one to prefer the interpretation provided in the Scheduling Guidelines over that included in the GATS, it would be privileging the use of a negotiating document over that of the law itself. This cannot be the case


P E T R O S C. M A V R O I D I S

2.4 The case law (along came US–Gambling) The panel in its report on US–Gambling19 dealt with a series of US state laws which prohibited the supply of internet gambling in a non-discriminatory manner: neither US nor foreign services suppliers were allowed to supply such services.20 At the same time, however, the United States had not imposed in their schedule of commitments any limits on cross-border consumption (Mode 1) of internet gambling services. The question before the panel was to what extent the fact that the United States imposed no restrictions on consumption abroad of internet services, outweighed the fact that, under US laws, the supply of internet gambling was prohibited anyway for all suppliers, irrespective of their origin. To put it bluntly, should non-discriminatory restrictions be scheduled anyway ? The panel without formally pronouncing on the relationship between Art. XVI and XVII GATS, accepted the argument advanced by Antigua and Barbuda that, in the absence of explicit statements to this effect in its schedule of concessions, the United States was effectively in violation of their GATS commitments. The panel found that a series of US federal and state measures, which regulate the supply of services by foreign and domestic suppliers alike, violated Art. XVI GATS : (a) in ·6.36five, the Panel outlaws the Federal Wire Act, which reads in pertinent part : Whoever being engaged in the business of betting or wagering knowingly uses a wire communication facility for the transmission in interstate or foreign commerce of bets and wagers or information assisting in the placing of bets or wagers or any sporting event or contest, or for the transmission of a wire communication which entitles the recipient to receive money or credit as a result of bets or wagers shall be fined under this title or imprisoned not more than two years, or both.

(b) in ·6.373, the Panel finds that the Federal Travel Act is WTO inconsistent : this legislation reads in a quasi-identical manner with the Wire Act, and punishes, inter alia, those who use the mail with intent to distribute the proceeds of, or promote, establish, carry on, facilitate unlawful activities ; (c) in ·6.380, the Panel reaches the conclusion that the Illegal Gambling Act (IGBA) is WTO-inconsistent. The IGBA reads in pertinent part : Whoever conducts, finances, manages, supervises, directs or owns all or part of an illegal gambling business shall be fined under this title or imprisoned not more than five years, or both. for the only hierarchy established in public international law (with the exception of jus cogens which is irrelevant here) is that of lex posterior (subsequent law). The Scheduling Guidelines are neither lex, nor posterior. 19 See United States – Measures Affecting the Cross-Border Supply of Gambling and Betting Services, WTO doc. WT/DS 285. 20 With the exception of the Inter-state Horse Act (IHA).

Road from non-discrimination to market access in GATS 9

The Appellate Body (AB), explicitly, condoned these findings, by holding the view that with respect to the three statutes mentioned above, the United States had violated Art. XVI GATS (·265).21

2.5 Intermediary conclusion on the status of positive law There is a discrepancy between some negotiating documents and the explicit wording of Art. XVI GATS : whereas the former (Scheduling Guidelines) accept the view that non-discriminatory measures can come under the purview of Art. XVI GATS, the wording of Art. XVI.1 GATS suggests the opposite. In the AB’s view, a non-discriminatory statute can be found to be in violation of Art. XVI GATS.22 For this to be the case, however, a necessary condition is that the AB must have (at least implicitly) accepted that Art. XVI GATS covers market access for domestic services and services suppliers as well. This is what I find fault with.23

3. The suggested approach: XVI is a list of in principle impermissible exceptions to Art. XVII GATS

3.1 The approach in a nutshell The approach could be summarized as follows: Art. XVI GATS covers only foreign suppliers and is nothing but a list of violations of Art. XVII GATS. It is probably the case that the measures included therein are the most frequent violations of national treatment. Consequently, the negotiators, by including these measures in Art. XVI GATS, wanted to signal their desire to, in principle, abolish them. The suggested approach requires sequencing Art. XVI GATS to Art. XVII GATS : assuming a WTO Member has not granted national treatment to foreign suppliers, it cannot impose on them any of the measures featured in Art. XVI GATS, unless it has indicated so in its schedule of concessions.24 21 Whereas the panel dealt with both state and federal laws, the AB report concerns federal laws only. 22 Note, however, that the AB did not reach this conclusion by deferring to the Scheduling Guidelines. 23 Other commentators as well have expressed their views on US–Gambling see Krajweski (2005), Ortino (2006), Pauwelyn (2005), and Delimatsis (2006). None of them however, advances an understanding of Art. XVI GATS along the lines that we do in this paper. To them, as is the case with Mattoo (2000), Art. XVI GATS concerns not only foreign but domestic suppliers as well. There are differences in their analysis, however. While Pauwelyn (2005) and Delimatsis (2006) seem to privilege a narrow understanding of Art. XVI GATS, Mattoo (2000) and Krajewski (2005) seem to sequence Art. XVII to Art. XVI GATS in the sense that once a measure is covered by the latter, it is saved from a test of legality under Art. XVII. Ortino (2006) is the closest to the suggested approach. Even he, however, sees an overlap between Arts. XVI and XVII GATS: in his view, non-discriminatory measures can come under the purview of Art. XVI GATS. For the reasons explained in Sub-section 3.2, the approach suggested in this paper takes distance from all these writings. 24 Of course, we suppose that a WTO Member has made specific commitments. Otherwise, recourse to Art. XVII GATS is not an issue; that is, sequencing Art. XVI to Art. XVII GATS means that (a) a country has made specific commitments in a given sector, and (b) it has decided not to make a full national treatment commitment. In this case, the scheduling of commitments under Art. XVI GATS will respond to


P E T R O S C. M A V R O I D I S

3.2 Justifying the suggested approach There are good, both purely legal and policy, arguments justifying our approach. First, for pure textual reasons our interpretation is robust. Art. XVI.1 GATS reads : With respect to market access through the modes of supply identified in Article I, each Member shall accord services and service suppliers of any other Member treatment no less favourable than that provided for under the terms, limitations and conditions agreed and specified in its Schedule. (emphasis added).

Art. XVI.1 GATS is the chapeau governing all measures reflected in the second paragraph of this provision. Consequently, all these measures are measures to be applied to services suppliers of other Members.25 By the same token, it is clear that not all national treatment limitations are meant to be addressed by Art. XVII GATS, and that some of the so-called market access limitations listed in Art. XVI :2 GATS can only concern foreigners. Otherwise, how would one explain the inclusion of measures such as foreign equity limitations or the requirement of constituting a joint-venture within the so-called market access limitations? Besides, a closer look at the measures included in paragraph 2 of Art. XVI GATS raises doubts – at least in the eyes of the author – as to whether any of those measures was thought of as restricting access from national suppliers. For instance, what would be the regulatory rationale for limiting the total value of transactions, operations, or assets that can be performed or held by national suppliers ?26 By the same token, what would be the regulatory rationale for limiting the total number of natural persons that might be employed by a national supplier ?27 Certainly, those kinds of limitations seem to be intended to actually limit the scale of the supplier, and it would be odd to think that a regulator would like to limit the growth of national companies.28 Second, to conceive Art. XVI GATS as an instrument to regulate market access for domestic services and services suppliers goes against the very nature of the GATS. The GATS is the agreement to liberalize trade in services. Through its many provisions, foreign services and services suppliers will be granted market access to markets that they could not, in principle, access before the entry into force of the

the question how much less favourably will foreign service suppliers be treated (than domestic service suppliers)? 25 WTO courts that have a tendency to over-emphasize the importance of the textual element should be persuaded only by this observation that Art. XVI GATS covers only measures regulating market access for foreign services suppliers. 26 Assuming it is concentration of power that the regulator is after, this objective is typically achieved through domestic antitrust statutes. 27 It is typically developed nations with (relatively speaking) high unemployment rates that have made the most of the commitments. 28 Incidentally, note that the examples of market access limitations falling under paragraphs 2(b), (c), (d), and (e) given in paragraph 12 of the Scheduling Guidelines always refer to limitations applicable only to foreign service and service suppliers.

Road from non-discrimination to market access in GATS 11

GATS. Art. XVI GATS is one of the provisions regulating liberalization commitments. Domestic services and services suppliers will access their national market under the conditions specified in domestic legal instruments, such as, for example, the national constitution, state laws, Bar regulations etc. The very title of Part III (covering Arts. XVI, XVII, and XVIII GATS) is Specific Commitments. Commitments are made to foreign nations and not to domestic suppliers. It is logically unthinkable to talk of trade liberalization – that is, an international transaction par excellence – and, at the same time, understand this term to cover market access for nationals as well.29 This conclusion is further supported by a quick look into the counter-factual. Assume that by introducing the number 5 into its commitments on financial services (banks), Argentina has been limiting the number of bankers, domestic and foreign alike. As a result, Argentines will be in a position to access their domestic banking market under the conditions mentioned in GATS and not, for example, in a domestic statute. That is, Argentina, by signing the GATS, accepted a change at the level of regulation for its domestic citizens as well : it cannot unilaterally decide the number of Argentine bankers, their total value of transactions, their number of operations, the total number of natural persons they may employ, the legal entity under which they will operate ; it cannot unilaterally decide any of the measures listed in Art. XVI GATS as far as Argentine suppliers are concerned. And if, in the future, it wants to modify its schedule (and reduce the number of Argentine bankers as well), it will, by virtue of Art. XXI GATS, have to pay compensation to its trading partners for damage inflicted (at least in part) on its own suppliers. One of the few issues where public international law is crystalclear is in the area of transfer of sovereignty: absent clear and unambiguous transfer, a regulatory exercise is presumed to rest with the sovereign state. This is what the maxim in dubio mitius accepted across international jurisdictions (and in the WTO as well) amounts to. The GATS founding fathers did not intend to transfer sovereignty in this matter. In fact, a look in the context (the rest of the GATS) clearly supports this conclusion. Take Art. VI GATS, for example : this provision pre-supposes that each WTO Member can unilaterally decide on the conditions for acceding into a particular service market. Regulatory diversity is the working hypothesis of this provision and, whenever gains from cooperation exist, regulatory cooperation will indeed take place under the conditions established in Art. VI GATS. This reading of the GATS is very much in line with the function of international trade agreements : they are there to internalize externalities stemming from the unilateral definition of trade and trade-related policies.30 The GATS is a negative integration contract : policies will be unilaterally defined and trade liberalization will take place against the background of a pre-existing regulatory diversity. Positive integration will occur under the 29 Matsushita et al. (2006) tangentially touch upon this issue. 30 An elegant account of the ‘received theory’ is offered in Bagwell and Staiger (2002).


P E T R O S C. M A V R O I D I S

limited conditions established in Art. VI GATS. This reading is consonant with the construction of the GATS as an agreement leading to the elimination of discrimination in international trade, and not as an agreement to de-regulate.31 The latter might occur when work undertaken under Art. VI GATS is proved successful. Until that day, the maximum promise to trading partners is national treatment (non-discrimination). National treatment pre-supposes unilateral definition of policies. Consequently, the number of Argentine bankers operating in Argentina is not a matter of commitments under the GATS, but simply a matter of domestic laws. Third, from a policy-perspective, it seems that trade liberalization will indeed be served were one to restrict the applicability of Art. XVI GATS to foreign services and services suppliers only : the limitations indicated in a schedule of concessions will apply to foreign services suppliers only. So in case Argentine limits the number of banks to five, all five will be foreign banks. This approach has a welcome positive externality. Mattoo (2000) in his insightful paper, points to a lacuna in the current GATS : assume that Argentina has limited the number of banks operating in its sovereignty to five. Assume further that they accept a sixth bank. Should this sixth bank be up for grabs for both foreign and domestic suppliers ? Assuming the US–Gambling approach is followed, the response should be in the affirmative. In the suggested approach, the sixth bank can be offered to foreign bankers only, and it becomes the benchmark for any additional bank that can request market access making a non-discrimination claim to this effect. Consequently, the suggested approach both at a static and at a dynamic setting, induces more trade liberalization, indeed the very purpose of the WTO edifice. There are, of course, some legitimate counter-arguments to the advanced thesis. The very existence of Art. XX.2 GATS could be perceived as signaling a potential overlap between Art. XVI and Art. XVII GATS. This argument, it is submitted, is ill-founded. Art. XX.2 GATS reads : Measures inconsistent with both Articles XVI and XVII shall be inscribed in the column relating to Article XVI. In this case the inscription will be considered to provide a condition or qualification to Article XVII as well.

Art. XX.2 GATS is nothing but a scheduling device. Its rationale is provided by its (proper) counter-factual : absent this provision, any time a WTO Member had recourse to one of the measures embedded in Art. XVI GATS, it would have to include them explicitly both under the Art. XVI, and the Art. XVII GATS columns. Art. XX.2 GATS makes the second inclusion redundant. It does only that. The other institutional argument against the suggested approach is ·8 of the Scheduling Guidelines, a point I briefly discussed supra. Recall that from a pure

31 Judging from the work accomplished under Art. VI GATS so far on accountants, this looks a very safe conclusion, see Delimatsis (2006). Recall that under Art. VI GATS, one could potentially move towards establishing common qualifications/standards.

Road from non-discrimination to market access in GATS 13

formal (legal) perspective, this is not much of an issue : case-law has clarified that such documents are supplementary means of interpretation. Moreover, this particular document itself reflects the view that it is not an authentic interpretation of the GATS. Its main value lies in the fact that it provides some information additional to the GATS without which a serious gap-filling exercise would have become, sooner or later, necessary. But, even if we were to set aside these arguments, a look into the content of the Scheduling Guidelines does not necessarily lead to the conclusion that what the negotiators had in mind when drafting Art. XVI GATS was measures applied to both domestic and foreign suppliers : one should not be forgetful of the fact that most of the examples that negotiators used while drafting the Scheduling Guidelines are examples of transactions concerning foreign suppliers only. In this regard, we read in ·39 : (a) Limitations on the number of service suppliers : – Licence for a new restaurant based on an economic needs test. – Annually established quotas for foreign medical practitioners. – Government or privately owned monopoly for labour exchange agency services. – Nationality requirements for suppliers of services (equivalent to zero quota). (b) Limitations on the total value of transaction or assets: – Foreign bank subsidiaries limited to x percent of total domestic assets of all banks. (c) Limitations on the total number of service operations or quantity of service output : – Restrictions on broadcasting time available for foreign films. (d) Limitations on the total number of natural persons : – Foreign labour should not exceed x percent and/or wages xy percent of total. (e) Restrictions or requirements regarding type of legal entity or joint venture: – Commercial presence excludes representative offices. – Foreign companies required to establish subsidiaries. – In sector x, commercial presence must take the form of a partnership. (f) Limitations on the participation of foreign capital : – Foreign equity ceiling of x percent for a particular form of commercial presence. (emphasis added).

The examples mentioned above show clearly that, what the founding fathers had in mind when drafting Art. XVI GATS, was a class of measures that concerned explicitly foreign service suppliers only. But, even the examples used, which do not fall into this category (that is, which explicitly concern foreign service suppliers), do not necessarily contradict that the working hypothesis of Art. XVI GATT was


P E T R O S C. M A V R O I D I S

to cover measures aimed at regulating market access for foreign service suppliers only. Take the economic needs test, for example : WTO Members could use it to block foreign entry of say a new gas station (or restaurant), arguing that there is already sufficient (domestic) supply in a given region. All this leads me to conclude, even assuming that the legal relevance of the Scheduling Guidelines is higher than it actually is, that it is a matter of positive law and that, as a result, recourse to it is compulsory, quod non, they do not clearly point to the opposite (of my preferred) direction. They are at best confusing : the examples used are examples that concern foreign service suppliers, whereas at the same time they reflect a statement to the effect that such measures could be nondiscriminatory. Crucially, all examples used can be understood as concerning foreign service suppliers only. One could further advance the argument that Art. XVI.1 GATS, although addressed to foreign service suppliers only, does not also clearly state whether the measures included in Art. XVI.2 GATS are discriminatory or not.32 This argument is not persuasive : first, if this were indeed the will of the founding fathers, they could very easily add the term ‘ domestic suppliers ’ in Art. XVI.1 GATS. They did not ; second, this reading is at odds with the construction of the GATS as a negative integration contract already dismissed (see supra). In the same vein, one could argue that Art. XVI GATS is a legislative response to the awkward de facto discrimination analysis : instead of having to argue whether an origin neutral measure de facto hits foreigners harder, the GATS founding fathers, for reasons of administrative ease, simply inserted a list of measures that should be, in principle, declared illegal anyway, irrespective of who (domestic/ foreign supplier) is hit harder. This is definitely an appealing argument, especially so because the WTO case law on de facto discrimination leaves a lot to be desired.33 This argument runs of course afoul the construction of GATS as an instrument eventually prohibiting discrimination and, for the reasons mentioned above, should be dismissed. And of course, they could have stated that such measures cover both domestic and service foreign suppliers alike. They did not. Moreover, the gains, were such policies to be declared outright illegal, should not be exaggerated. Assume for example, that Argentina requests that banks do not employ more than 200 employees, and it also requests by law that 50 % of the banking personnel speaks one of the idioms used in every Argentine region. Such a measure would have an impact on the nationality of employees hired, since de facto it confers an advantage on local hires. Should it be outlawed by Art. XVI GATS ? In other words, should measures which, not de jure but de facto, amount to a measure such as those proscribed by Art. XVI GATS, be also prohibited ? If the 32 This argument was advanced by an anonymous referee. 33 See on this score, Horn and Mavroidis (2004).

Road from non-discrimination to market access in GATS 15

founding fathers cared about administrative ease, they could have been more careful in their expression. One might finally legitimately ask the question whether there is a downside to the suggested approach? Indeed, one could argue that absence of scheduling of non-discriminatory regulatory interventions under Art. XVI GATS necessarily leads to absence of transparency, and, thus, to increased transaction costs. It would indeed be regrettable to incite opaque protectionism by restricting the list of items that should come under Art. XVI GATS (and, consequently, be scheduled). For scheduling, a measure contains by definition a transparency externality. Such fears are, however, grossly exaggerated : Art. III GATS obliges WTO Members to publish promptly _ all relevant measures of general application which pertain to or affect the operation of this Agreement.

This provision arguably leaves some discretion to the regulating state to decide whether a measure is of general application (this should be most of the time a mundane exercise) and, among this set of measures, which pertains to or affects the GATS. In the worst case scenario, assuming the WTO Member concerned commits a type II error (false negative),34 this provision, like any other provision in the WTO agreement is justifiable. Hence, the exercise of discretion is justifiable as well. Moreover, one should not neglect that the possibility for a cross-notification (under Art. III.5 GATS) is embedded in the GATS system. Looking for evidence in this context (cross-notification) should not be that burdensome: it suffices that a request be submitted to the regulatory regime governing market access for a particular service.

3.3 Expressing the suggested approach in four steps The approach advanced here suggests hence, a different role for Art. XVI GATS, than that envisaged by the AB. It is by now clear that, in our view, it seems plausible to view Art. XVI GATS as a sub-set of Art. XVII GATS. That is, WTO Members : (a) will decide whether or not to accord national treatment to foreign services and services suppliers by indicating their wish in the schedule ; (b) if yes, then Art. XVI GATS is irrelevant. With respect to measures affecting trade in services for which national treatment has been granted, WTO Members incur anyway the transparency obligation embedded in Art. III GATS ;

34 A Type-II error (false negative) in this context would lead to a situation where the agreed law has not been applied (respected). This would be the case if a WTO Member does not notify a GATS-covered law. False positives are not issue here: over-notification is rather welcome by the WTO Membership. More generally, on GATS transparency, see Iida and Nielsen (2003).


P E T R O S C. M A V R O I D I S

(c) if no, they will so indicate in their schedule of concessions under the column ‘ Art. XVII GATS ’ ;35 (d) if no, and assuming they choose an instrument reflected in Art. XVI.2 GATS, then they will have to indicate their concession (which should be termed limitation or restriction) under the column ‘ Art. XVI GATS ’ of their schedule, and not under Art. XVII GATS (by virtue of Art. XX.2 GATS). Points (b) and (c) need some further clarification. Let us start from (b), where a WTO Member has granted national treatment to foreign suppliers (the inscription in the schedule is none). In this case, Art. XVI GATS is irrelevant because national treatment has been afforded. However, this does not mean that a WTO Member granting national treatment can never have recourse to one of the measures included in Art. XVI GATS ; it only means that the WTO Member cannot apply such a measure against foreign suppliers only. It can of course impose a complete ban on the supply of services (a` la US–Gambling). In this case, it will have to notify its relevant laws (the ban) in accordance with Art. III GATS, and it will incur no further obligation to also include it in its Art. XVI GATS column.36 Let us move to point (c) now, irrespective of whether the inscription is unbound or other, Art. XVI GATS could be relevant. If the inscription is unbound, the WTO Member is free to deviate as much as it deems appropriate from national treatment. Importantly, it does not, with the exception of the measures reflected in Art. XVI GATS, prejudge the extent of the deviation ex ante. It can of course also deviate from Art. XVI.2 GATS, if it has indicated so in its schedule. If the inscription is other,37 the WTO Member will indicate whether it wishes to avail itself of the possibilities to refuse market access offered in Art. XVI.2 GATS. This is so since, under other, it can reflect measures going beyond the coverage of Art. XVI.2 GATS. In other words, WTO Members should, through the measures coming under the purview of Art. XVI GATS, be explaining the terms of discriminatory treatment afforded to foreign services and services suppliers. Such measures, however, should not be construed to be as the only permissible discriminatory measures. Hence, 35 In this case as well, WTO Members will have to abide by the transparency requirement embedded in Art. III GATS. However, in this case, knowledge of the regulatory regime in place will be of a mere academic interest, since no market access commitments will have been entered. 36 Indeed, requiring inscription in the Art. XVI GATS column looks like an inappropriate double emploi. If WTO Members have not respected their obligations under Art. III GATS, one cannot cure such defect by over-extending the ambit of Art. XVI GATS: the former covers laws irrespective of whether trade commitments have been entered into, and has, consequently, a wider coverage. However, nothing can stop WTO Members from reflecting in their Art. XVI GATS column measures notified under Art. III GATS. What we suggest here is that they should not be required to do so. 37 The term here is used as in the Scheduling Guidelines. That is, it covers cases where a WTO Member has not granted national treatment to foreign service suppliers, but it has ex ante bound its discretion as to the treatment to be provided.

Road from non-discrimination to market access in GATS 17

Art. XVI GATS should be understood as a provision that clarifies some of the agreed deviations from national treatment.

4. Testing the approach

4.1 Limitations on the number of service suppliers Let us suppose that Argentina made a commitment to allow only five banks into its market. Is this a guarantee that five banks will enter the market or just an opportunity given to five banks to enter the market ? In the absence of a National Treatment commitment (NT unbound),38 nothing is in fact guaranteed. Foreign banks can be subject to conditions that may be much more onerous than those imposed on national suppliers. If there is a National Treatment commitment (NT none), there is still no guarantee that the quota will be fulfilled since, again, the conditions imposed may be highly restrictive. Following this understanding – a quota only for foreign banks – foreigners only would be competing for the five banks, the host state having no obligation to grant five licences to this effect. In a NT none-scenario, the foreign banks will be competing for five licences. The licensing requirements will be equivalent to those applied to domestic incumbents. In a NT unbound-scenario, the foreign banks will be competing for five licences under conditions not specified ex ante in the schedule of concessions. However, the conditions under which the first licence is granted become ipso facto the benchmark for any subsequent licence (by virtue of the MFN, most favoured nation-obligation embedded in Art. II GATS). In a NT otherscenario, market access for foreign banks will be explained in the schedule. Here we distinguish between two alternatives : if the measure coming under the purview of other is part and parcel of Art. XVI GATS, then the relevant entry will be under Art. XVI GATS (by virtue of Art. XX.2 GATS) ; if the measure is a measure other than the measures included in Art. XVI GATS, it will figure under Art. XVII GATS.

4.2 Limitations on the total value of service transactions or assets This type of restrictions does not abound in schedules, and the provision itself (Art. XVI.2b GATS) is not absolutely clear. Most examples apply only to foreigners, which reinforces the idea that Art. XVI GATS seems to have been conceived as capturing all restrictions affecting foreign services and services suppliers, and only in a subsidiary manner domestic services and services suppliers.39 In addition, this type of restriction may apply at the market level (e.g. foreign bank subsidiaries 38 We use the terminology used in GATS schedules: unbound means that the WTO Member making the commitment does not at all limit its sovereignty; none, that no restrictions have been placed; and, finally, other, that there is specific language in the schedule of concessions specifying the extent of the commitment. NT stands for national treatment. 39 The only example provided by the Scheduling Guidelines is the following : ‘ Foreign bank subsidiaries limited to x percent of total domestic assets of all banks.’ See S/L/92, paragraph 12.


P E T R O S C. M A V R O I D I S

limited to x % of total domestic assets of all banks), or at the individual institution level (e.g. foreign banks are subject to per client foreign currency credit extension limit of 25 % of the net worth of their head office40). Let us analyze both options. A limitation on the total value of assets applying at the market level It is difficult, if not impossible, to think how such a limitation would apply to domestic banks. Clearly, all these limitations found by the authors in real schedules of commitments, as well as the example given by the Scheduling Guidelines, apply only to foreign suppliers. Check the following example : Foreign bank subsidiaries in Argentina are limited to 25% of total domestic assets of all banks.

By definition, this is a discriminatory measure, hence it is contrary to National Treatment. Following Art. XX.2 GATS, such a measure should be scheduled under the Market Access (Art. XVI GATS) column, and it would be considered as qualifying both market access and national treatment. Is this a guarantee or an opportunity to trade ? The only thing that is ‘guaranteed ’ is that foreign bank subsidiaries would eventually (i.e. assuming that they fulfil all other requirements) be allowed, provided the new addition to the market does not take foreign-banks’ market share beyond 25 %. A limitation on the total value of assets applying at the individual institution level Let us suppose that Argentina schedules the following measure : Banks are subject to a per client foreign currency credit extension limit of 25% of the net worth of their head office.

We are aware that such a measure might eventually be justified as a ‘ measure for prudential reasons’ as per ·2 of the GATS Annex on Financial Services, and, therefore, is not subject to scheduling. Following current practice, however, let us suppose that it has been scheduled anyway, as any other market access limitation. What is guaranteed by this commitment ? Very little indeed. The commitment only ‘ guarantees’ that banks will not be requested to cap foreign currency credits at a level equivalent to less than 25 % of the head office’s net worth. That’s all. The most important question in this case, however, is whether this measure should be understood as a limitation applying to all banks or only to foreign banks. If the measure is applied as such to all banks, irrespective of whether they are foreign owned, it would not constitute a national treatment restriction,41 and Art. XX.2 GATS would be irrelevant. The measure would, in principle, affect

40 Example taken from a real schedule of commitments. 41 We leave aside consideration of whether a limitation on foreign currency loans might constitute a de facto national treatment violation.

Road from non-discrimination to market access in GATS 19

everybody’s conditions of competition in that segment of the market. What if the measure applied only to foreign banks ? In that case, the measure – a Market Access limitation – would also constitute a National Treatment qualification since domestic banks would not be subjected to such a limit. If that is the case, Argentina would only need to schedule the measure in the Market Access column, since it would be also understood as a qualification on national treatment. In conclusion, the guarantee provided by this type of market access limitation is very little indeed. Considered in its entirety, with other limitations and qualifications that may apply, the commitment still provides an opportunity to access that market. Whether the limitation is meant to be applied to all banks or, in the case of dual regimes, to foreign banks only would not change the legal rights and obligations of Argentina because of Art. XX.2 GATS. However, it may change conditions of competition if the limitation is applied only to foreign banks. Besides, the legitimate question may arise : would it be possible to justify a discriminatory application of the measure post scheduling? This is a paradoxical, although logical, consequence of the protection granted by Art. XX.2 GATS.

4.3 Limitations on the total value of service operations Before proceeding with this example, it is worth noting that, at least in the case of banking, the distinction made in Art. XVI GATS between services operation/assets on the one hand (Art. XVI.2b GATS) and service operations/output on the other (Art. XVI.2c GATS) is far from clear. In practical terms, the distinction might not pose problems to Members, who are not required to provide a taxonomy of their limitations according to Art. XVI GATS. The lack of distinction may pose, however, problems for a hypothetical exercise such as the one undertaken here. With that proviso in mind, let us suppose that Argentina applies the following limitation : Banks can establish only up to ten off-premise Automated-Teller-Machines (ATMs).

What type of market access guarantee is this ? The only thing that is guaranteed is that banks will be told to restrict their off-premise ATMs to less than ten units. That is all. Apart from that, on its own, this measure does not constitute a real market access guarantee, even if no other Art. XVI GATS-type of measure applies, because access would again depend on other terms and conditions that may apply to banks wishing to establish in this country. Let us suppose that the measure only applies to foreign banks’ subsidiaries or branches, as per our preferred understanding of Art. XVI GATS. Since the measure modifies the conditions of competition against foreign entities, the measure is not only inconsistent with market access but also with national treatment. However, by virtue of Art. XX.2 GATS, the measure would only appear in the market access column and would qualify for national treatment as well.


P E T R O S C. M A V R O I D I S

4.4 Limitations on the total number of natural persons that may be employed in a particular services sector or that a service supplier may employ It is hard to think why domestic banks would be limited in their ability to hire natural persons. Again, most examples one can think of would concern either the ability of foreign service suppliers to employ natural persons in general, or the ability of companies to employ foreign natural persons. Both cases would constitute in fact national treatment restrictions, and, if scheduled, would have to be seen in light of Art. XX.2 GATS. The conclusions reached in Section 4.2 supra holds also for this example.

4.5 Measures which restrict or require specific types of legal entity or joint venture These measures abound in schedules. The requirement to form a joint venture could be a national treatment qualification, unless we think of regimes in which domestic investors could only supply services if and only if they are associated with a foreign partner, or even of regimes where a WTO Member imposes heavier requirements on foreign suppliers.42 Since requirements imposed on joint ventures (between domestic and foreign suppliers) are by definition both a market access and a national treatment restriction, as per our understanding of Art. XVI GATS, they should be scheduled only in the market access column and the entry would qualify both obligations. Is this a guarantee or an opportunity to trade ? Again, the answer to this question would necessarily depend on the interplay with other limitations that may have been scheduled, or with other conditions that may applied even if not scheduled. Requirements to constitute a specific type of legal entity might eventually be applicable to both domestic and foreign suppliers. In this case, assuming adherence to NT, Art. XVI GATS is obsolete : foreign like domestic banks will have to choose from a limited selection of company types. The Art. XVI GATS entry is important only if foreign banks are treated, in this respect, different from their domestic counterparts. In the case of banking and other financial services, the provision is particularly relevant for foreign entities, since the most important concern for regulators is whether foreign banks can access the market through a branch or through a subsidiary (i.e. a locally incorporated company). Restrictions to incorporate locally may be accompanied by the requirement to adopt a specific type of juridical person and this restriction may apply to both domestic and foreign

42 It could be, for example, that a WTO Member allows domestic suppliers to supply a service through a simple societal form, whereas imposes heavier requirements (limited responsibility) on foreign suppliers of the same service.

Road from non-discrimination to market access in GATS 21

entities. So, let us suppose that Argentina imposes the following requirement in its schedule of commitments : Only stock-companies can be licensed as banks, cooperatives and mutual companies are not allowed to take deposits.

If the measure applies both to foreign and domestic banks, there is no national treatment violation, and therefore no need to have recourse to Art. XX.2 GATS. If the measure concerns only foreign banks, then it is both a market access restriction and a national treatment violation, because deposit-taking activities might be performed by domestic entities organized as cooperatives. If this is the case, in order to safeguard say Argentina’s rights under the Agreement, the measure would have to be considered as qualifying also National Treatment, by virtue of Art. XX.2 GATS. Again, the conclusions reached in Section 4.2. supra applies mutatis mutandis to this case.

4.6 Limitations on the participation of foreign capital in terms of maximum percentage limits on foreign share holding or the total value of individual or aggregate foreign investment Limitations on foreign ownership (e.g. foreign ownership limited to 49 %) are by definition national treatment restrictions. Therefore, Art. XX.2 GATS applies almost automatically in interpreting these entries. However, some ownership limitations may apply to both domestic and foreign investors, for example direct or indirect ownership or voting rights in a credit institution of a single shareholder other than credit institution, insurance company or investment firm cannot exceed 25 %. If that is the case, the measure is non-discriminatory, but would need to be scheduled anyway because it applies to foreigners as well. What type of guarantee is this ? Well, the only thing that is guaranteed here is that, if allowed to enter, (i) a foreign bank would be required to hold less than 49 % of the institution’s equity or (ii) an individual shareholder would not be obliged to hold less than 25 % of voting rights. The limitation is per se a qualification on national treatment, and therefore Art. XX.2 GATS would apply automatically to this provision in gauging rights and obligations of Argentina under the Agreement.

5. A coherent interpretation is possible, but leaves much to be desired _ The suggested interpretation does not require a re-drafting of any of the GATS provisions. It might entail that some commitments are more meaningful than anticipated : a country for example which accepts five new banks and has not specified anything more in its Art. XVI GATS column, will be forced to, in principle, accept that the five new banks must be foreign. Keep in mind, however, that a five


P E T R O S C. M A V R O I D I S

bank commitment is an opportunity to trade and not an obligation to accept five banks anyway. In defence of this understanding, I will state that the suggested proposal is in line with the very essence of the GATS, which is an overwhelmingly negative integration type of contract: WTO Members remain largely uninhibited to unilaterally define their policies. To the extent that there is international spill-over stemming from the exercise of their regulatory sovereignty, it will be dealt with within the four corners of the existing legislative framework. National treatment is the maximum value of any national promise.43 If we could ex ante foresee the policies that will affect trade, then, in all likelihood, negotiators would have simply spelled out all these policies in the contract : the associated negotiating costs make such an exercise an impossibility. The GATS like the GATT,44 is an obligationally incomplete contract whereby national treatment operates as punishment in case unilateral policies deviate from the agreed objective not to provide non-negotiated protection.45 In the GATS, the situation is even more favourable to our argument since, national treatment is not even an obligation across the board, but merely a specific commitment. It follows that the negative integration character of the contract is even more prevalent.46 From an operational perspective, the suggested understanding is tantamount to sequencing Art. XVI GATS to Art. XVII GATS : the former is understood to reflect the types of deviations from national treatment that WTO Members should abolish, unless they have clearly indicated their wish not to do so. Even in this latter scenario, however, one step forward has been taken since ; they become the negotiating material for the next round.

References Bagwell, Kyel and Robert W. Staiger (2002), The Economics of the World Trading System, Cambridge, MA: MIT Press. Delimatsis, Panagiotis (2006), ‘ Don’t Gamble with GATS – The Interaction between Articles VI, XVI, XVII and XVIII GATS in the Light of the US–Gambling Case’, Journal of World Trade, forthcoming. Hoekman, Bernard (1996), ‘Assessing the General Agreement on Trade in Services ’, in Will Martin and L. Alan Winters (eds.), The Uruguay Round and Developing Economies, Cambridge: Cambridge University Press. Hoekman, Bernard M. and Michel Kostecki (2001), The Economics of the World Trading System, Oxford: Oxford University Press. 43 In theory, a WTO Member can treat foreign suppliers better than it treats its own nationals. To my knowledge, there is no national schedule including such concessions. 44 See, in this respect, the analysis in Horn and Mavroidis (2004). 45 Assuming of course, that a WTO Member has promised to grant national treatment to foreign suppliers. 46 Probably much of this discussion would have been taken care of if Art. XVI GATS were to follow Art. XVII GATS and be entitled as ‘Impermissible, in principle, deviations from national treatment.’

Road from non-discrimination to market access in GATS 23 Horn, Henrik and Petros C. Mavroidis (2004), ‘ Still hazy after all these years: the interpretation of GATT/ WTO case-law on tax discrimination ’, European Journal of International Law, 15: 39–69. Iida, Keiya and Julia Nielson (2003), ‘ Transparency in Domestic Regulation: Practices and Possibilities’, in Aaditya Mattoo and Pierre Sauve´ (eds.), Domestic Regulation and Service Trade Liberalization, Washington, DC : The World Bank and Oxford University Press, pp. 7–20. Krajewski, Markus (2005), ‘ Playing by the Rules? ’, Legal Issues of Economic Integration, 32: 417–447. Matsushita, Mitsuo, Thomas J. Schonbaum, and Petros C. Mavroidis (2006), The World Trade Organization Law, Practice and Policy, 2nd edition, Oxford: Oxford University Press. Mattoo, Aaditya (2000), ‘MFN and the GATS’, in Thomas Cottier and Petros C. Mavroidis (eds.), Regulatory Barriers and the Principle of Non-Discrimination, Ann Arbor: University of Michigan Press, pp. 51–99. Ortino, Federico (2006), ‘Treaty Interpretation and the WTO Appellate Body Report on US – Gambling: A Critique’, Journal of International Economic Law, 9 : 117–148. Pauwelyn, Joost (2005), ‘ Rien ne va plus? Distinguishing Domestic Regulation from Market Access in GATT and GATS’, World Trade Review, 4 : 131–170.

World Trade Review (2007), 6 : 1, 25–44 Printed in the United Kingdom f Simon Kemp doi:10.1017/S1474745606003089

Psychology and opposition to free trade SIMON KEMP* University of Canterbury

Abstract : This paper reviews psychological reasons why the enthusiasm of the general public for free international trade might be less than that of the economist. Six specific reasons are advanced : (1) lay views of utility emphasize employment over consumption; (2) status quo bias results from loss aversion; (3) people think altruistically but parochially; (4) people often consider fairness in bargaining situations; (5) people may hold inappropriate fixed pie beliefs ; and (6) people may misunderstand Ricardo’s principle of comparative advantage. The reasons vary in their apparent rationality and appear to operate in concert rather than independently.

1. Introduction Ever since Adam Smith’s attack on the mercantilist system in The Wealth of Nations, economists have generally favoured free trade. Moreover, many economists hold that the benefits of free trade derive not only from improved access to foreign markets but also from the increased availability of a wide range of imported goods in the home market (Corden, 1974 ; Bhagwati, 1991 ; Irwin, 2005).1 The existence of such a professional consensus is supported by survey data. Alston et al. (1992) reported the results of a large questionnaire survey of US economists. Over 71 % of them ‘generally agreed ’ that ‘ tariffs and import quotas usually reduce general economic welfare ’, a further 21 % ‘ agreed with provisos ’, and less than 7 % ‘generally disagreed ’. Frey et al. (1984) found 57 % of an international sample of economists generally agreeing, and just over 10 % generally * Address for correspondence : Simon Kemp, Department of Psychology, University of Canterbury, Christchurch, New Zealand. Fax: +64 3 364 2181. Email: [email protected] I am very happy to acknowledge helpful suggestions and encouragement from Paul Brown, Linda Cameron, Ananish Chaudhuri, Marion Jansen, Patrick Low, Philip Meguire, John Tisdell and Steven Tucker. Three anonymous reviewers and the editor, Douglas Irwin, also offered constructive and very helpful feedback. Finally, I am particularly grateful for the frequent and stimulating interactions I have had with Jonathan Baron on the topic. 1 Roberts (2001: 106) remarks: ‘Even those [economists] who make theoretical arguments against free trade are loath to advocate tariffs and quotas in practice. So at the next cocktail party when someone tells a joke about the one-armed economist being unemployed because he can’t say ‘ on the other hand’, or how if you laid all the economists end to end, they still wouldn’t reach a conclusions, don’t encourage them with the false laughter such dull-wittedness provokes. Smile a knowing smile and tell them you have heard otherwise on the question of international trade.’




disagreeing with a similar question. (The lower percentage of agreement might reflect either the different samples, or the omission of ‘ usually ’ from Frey et al.’s question.) Such professional enthusiasm broadly conflicts with the views of many people. Gomes (2003) has charted a long history of resistance to free trade, and this resistance continues today. Krueger (2004 : 490) comments that ‘ protectionists enjoy remarkable success around the world ’. Consider, for example, recent popular protests against the further liberalization of trade suggested by the World Trade Organization, widespread hostility within both the USA and the European Union to the lowering of tariff and non-tariff barriers, and much discussion of the evils of the global economy (e.g. Mander and Goldsmith, 1996). Social surveys show that such resistance has a firm foundation in the opinions of ordinary people in developed economies. The 1992 US National Election Survey found 67 % of the respondents favoured ‘ placing new limits on foreign imports ’ (Scheve and Slaughter, 2001), and a similar survey carried out in 1996 found 52 % agreeing with the same proposal (Hainmueller and Hiscox, 2006). Thus, a majority of both US samples preferred increased protectionism to either increased liberalization or maintenance of the status quo. The 1995 International Social Survey Program surveyed 23 countries and 55 % of all the respondents agreed or strongly agreed that their country ‘ should limit the import of foreign products to protect its national economy ’, while 22 % disagreed or strongly disagreed. In only two countries (Japan and the Netherlands) were there more pro-trade (disagree) than anti-trade respondents (Mayda and Rodrik, 2005). Economists have long been aware that their views are not always shared by the wider community, but tend to attribute this to economic interests. They recognize that, while free trade may benefit the wider community, it is not usually in the interests of everyone. Those within a community who produce goods that compete with imports are likely to be worse off as a result of such trade (Feenstra, 2004). Such producers are likely to be a minority within their community. Why then, within democratic communities, do their views sometimes prevail ? Answers to this question have been given by students of political processes (e.g. Mayer, 1984 ; Olson, 1982 ; Rowley, 2001). For example, dairy producers might persuade a larger political grouping to block the importation of cheaper or better dairy products in return for the dairy producers’ support for issues of greater concern to the larger group. This article focuses on something that economists tend to overlook – the psychological reasons why the views of ordinary people, who are neither economists nor particularly threatened producers, might be hostile to free trade. This does not mean that other reasons are unimportant, but the non-psychological reasons for hostility to free trade are well known. For example, one current source of opposition to trade liberalization is people’s fear that the environment is endangered by freeing up trade (e.g. Mander and Goldsmith, 1996). Environmental damage would obviously be an important non-psychological reason, and current

Psychology and opposition to free trade 27

discussion of the issue (e.g., Irwin, 2005 : ch. 2) focuses on the relationship between environmental damage and trade, and under what circumstances freeing up trade might harm or preserve the environment. Another important concern is that globalization might lead to the swamping of minority cultures (e.g. Broude, 2005). To date, much psychological research into economic decision-making has considered conditions where market forces apply, and there has been less interest in applying psychology to economic decisions made by governments. Yet psychology may be more rather than less useful in understanding such decisions. Consider, first, decision processes that are subject to some kind of bias. As McCaffery and Baron (forthcoming: 2) point out : ‘ In private markets, arbitrage mechanisms, which allow some to profit from the biases of others, with overall prices showing little effect, can be expected to reduce the effects of bias _ In the public sector, however, the absence of any simple, general arbitrage mechanism, such as the market itself _ gives reason to believe that the adverse effects of cognitive biases can persist for long periods of time.’ Government controls on trade reduce the effect of arbitrage mechanisms, and, in turn, lower the chances of reducing any effects of cognitive biases or misperceptions. Even if a decision process is not subject to cognitive biases, there is an important role for psychology in helping to determine the utility produced by government decisions. A reasonable, if imperfect, measure of the value of market-supplied goods and services is given by people’s purchase behaviour, but this measure is not readily applicable to goods and services supplied by governments. Similarly, government control of trade changes, and in some cases eliminates, the market, and behavioural measures of the value of imported products may thus be either distorted or unavailable. If a government decides to ban the importation of bananas, then the obvious behavioural measure of their utility – people buying and eating them – will no longer work very effectively. Finally, the general activity of trading is an extremely widespread human behaviour. Ridley (1997) argues that trade may be as old as our species. Horan et al. (2005) suggest that the trading propensity of our early ancestors may have given them a decisive evolutionary edge over Neanderthals. Trivers (1971) reasoned that trade might have its origins in reciprocal altruism. Given this long and extensive context, we would expect a human propensity to develop beliefs, attitudes, and customs about exchange. In particular, we would expect people to be concerned with fair exchange, and with the fair distribution of resources produced by collective but specialized activity. Indeed, this is what we see. For example, people will often reject outcomes that would benefit them if they perceive another party will derive a greater, ‘ unfair’ benefit. In ultimatum games, two players are offered a resource provided they can agree on its distribution. Player A is first given the choice of how the resource should be divided, and player B can then only accept or reject the suggested division. If B accepts, the resource is divided according to A’s suggestion ; if B declines, neither A nor B gets anything.



The game theoretic solution suggests B should accept any positive offer, and thus that A should offer as little as possible. In practice, no culture has yet been found in which many players follow the game theoretic solution. B players frequently reject the offer of small proportions of the resource, and A players often offer half of it (e.g. Gu¨th et al., 1982 ; Gu¨th and Tietz, 1990 ; Henrich et al., 2005). As a second example, there has also been considerable research into the way that children believe resources should be distributed, and a number of experiments have investigated whether and in what circumstances children choose to distribute goods equally, according to the input involved in producing them, or according to other criteria (e.g. McGillicuddy-DeLisi et al., 1994). Overall, it would be utterly unsurprising to find such beliefs, attitudes, and customs operating in the wider if more impersonal arena of international trade. Indeed, there has been at least one previous suggestion that fairness might be a crucial factor in people’s views on trade (Davidson et al., 2006). Hence, the following section presents six psychological factors that might be influential on people’s attitude to international trade.

2. Psychological factors in trade values

2.1 Production versus consumption in human utility Taken broadly, utility in economics is mostly considered to derive from consumption. Adam Smith (1776/1904, Vol. 2 : 159) put it directly : ‘ Consumption is the sole end and purpose of all production ; and the interest of the producer ought to be attended to, only so far as it may be necessary for promoting that of the consumer.’ When David Ricardo (1817/1971 : ch. 7), in demonstrating the principle of comparative advantage, discussed English and Portuguese specialization in the production of cloth and wine, he presumed that people derived most utility from maximizing their consumption of the two commodities. Similarly, recent expositions of the arguments for free trade concentrate on the expansion of the production possibility frontier and hence consumption possibilities. Economic models define utility in terms of consumption and show how people’s utility is affected by changes in income that in turn depend on changes in trade patterns (e.g. Dixit and Norman, 1986 ; Krebs et al., 2005). In general, much of the argument for free trade in economics rests on utility gains from improved consumption possibilities. By contrast, many objections to trade liberalization stress the reduction in utility from the loss of employment or, more rarely, deteriorating work conditions. Restrictions on trade are often advocated by workers, for example, US steel workers or European farmers, who are likely to be affected. It has been widely recognized by both psychologists and economists ever since pioneering research during the Great Depression (Eisenberg and Lazarsfeld, 1938 ; Jahoha et al., 1971) that employment has value over and above the purchasing power provided by the income people earn.

Psychology and opposition to free trade 29

Recent quantitative evidence on the effects of unemployment comes from largescale studies (e.g. Frey and Stutzer, 2000 ; Lucas et al., 2004 ; Winkelmann and Winkelmann, 1998) in which unemployment has been shown by regression analyses to produce substantial adverse effects on well-being, even when the effects of income have been statistically accounted for. Lucas et al.’s (2004) longitudinal study of a large German sample showed that these effects could persist for years, and do not always vanish when people re-enter the workforce (suggesting that long-lasting disutility also results from being forced to change one’s job or occupation). Research into happiness and well-being shows that people’s happiness is heavily influenced by their work satisfaction but little by their income (e.g. Diener et al., 1999). People’s work is believed to provide a structure to their lives, an outlet for their creativity, a source of social contacts, and a sense of identity (e.g. Darity et al., 1999). A striking demonstration of this phenomenon in recent years is in the continuation of work in some East European concerns even when the workers have effectively not been paid at all (Bojilova, 2006). In sum, people derive benefits from their work which go beyond income or the consumption obtainable from this income, and which are not easily compensated for by income changes and adjustments. Thus, there is some distance between the findings of research into unemployment and the viewpoint taken by much economic theory. Of course, economists are concerned with unemployment and the non-income benefits of work, as well as with the effects of changing trade policy on them. Both economists in general and those who advocate free trade (e.g. Irwin, 2005) have long been aware both that unemployment is a real and serious problem for those affected and that it can follow trade liberalization. An important line of research has been into the actual changes in employment following trade liberalization, with the results suggesting that employment losses from this cause, although sometimes real and enduring (e.g. Kletzer, 1998), are not invariable and often less than the losses from economic recession or the introduction of new technology (e.g. Papageorgiou et al., 1990 ; Trefler, 2001). There has been a good deal of attention paid to the issue of how the losers from trade changes might be compensated for their loss (e.g. Dixit and Norman, 1986 ; M. Kemp and Wan, 1986). There have also been economic models that predict unemployment effects from trade changes (e.g. Jansen and Turrini, 2004). However, many of these models do not address changes to utility over and above income effects on consumption, and in this respect they do not take account of broader research into the utility of work.

2.2 Status quo bias, predicting future utility and loss aversion This section deals principally with two somewhat different phenomena: the difficulty people appear to have in predicting utility, and loss aversion. The two are among the host of more-or-less interrelated phenomena that have been recently



studied by psychological decision-making theorists and behavioural economists. Both suggest a status quo bias in the way people conceive the gains and losses from trade. Recent years have seen an influx into economics of psychological thinking and results based on psychological experimentation. The work of Kahneman and Tversky (e.g. 1982, 2000) has been particularly influential, and both psychologists and economists have now carried out a great deal of empirical work that has focussed on the decision-making capabilities and performance of individual people. For example, there have been a large number of experiments showing that people may make inconsistent decisions depending on the wording used to inform them of the conditions of the experiment. (e.g. Tversky and Kahneman, 1981 ; for a recent review, see Kahneman and Tversky, 2000). The effects of these wording changes are often described as framing effects. Such effects surface in real-world situations, as well as in the laboratory. For example, there is good evidence from the USA that people are most likely to sign up for whatever pension plan is framed as the default option (Madrian and Shea, 2001). Framing effects probably do have an important influence on the way people think of trade (e.g. Davidson et al., 2006). Hiscox (in press) asked US respondents whether they favoured or opposed increasing trade with other nations. Significantly more respondents were opposed to trade if they were first informed about the possibility of job losses than if they were given no information or if they were informed about the possibility of lowered prices. Furthermore, the influence of the framing was increased for less-educated respondents. In general, framing effects could bias people either towards or against any particular trade policy, depending on the frame provided. More specific biases are indicated by research on utility prediction and loss aversion. In the previous section, we considered the utility people gain from employment as opposed to consumption. However, when people make judgements about a future trade (or any other) policy, often they must judge on the basis of predicted future utility changes rather than any changes they have actually experienced. This raises the question of how good people are at predicting the utility they gain from future situations and activities. The research done to date indicates that people are generally poor at predicting their future utility and enjoyment, even when the activity, for example, listening to music or eating their favourite ice cream, is quite familiar to them (Kahneman and Snell, 1992). An endowment effect, in which people will demand more to give up some good than they would spend to acquire it, has commonly been studied by behavioural decision theorists and economists (e.g. Kahneman et al., 1991). In one variation of the endowment effect experiments, Loewenstein and Adler (1995) showed subjects a mug (coffee mugs have been a frequently used good in endowment effect research) and asked them to imagine they would be given the mug and what they would be prepared to sell if for. Subsequently, they were actually given the mug and a real opportunity to sell it. The prices asked then rose significantly,

Psychology and opposition to free trade 31

indicating that the subjects now valued the mug somewhat more highly than they imagined they would. In general, research to date indicates that people tend to ignore or misperceive the way their tastes and preferences will change in the future (Kahneman, 1994). Moreover, people often seem to underestimate their future powers of adaptation, and tend to make judgements that favour maintenance of the status quo. The straightforward implication for trade policy is that people might underestimate their ability to adapt to the changes produced by a more liberal (or more protectionist) trade regime. However, some caution is needed here. The extent of adaptation to major life changes varies greatly from individual to individual, and perhaps from situation to situation. For example, Lucas et al.’s (2004) study into long-term effects of unemployment found great variation in the ability of individuals to adapt to unemployment. A status quo bias also results from the phenomenon of loss aversion (Samuelson and Zeckhauser, 1988). Loss aversion has been frequently studied and it has often been found that the gain in utility from receiving, say, an extra $100 is not generally perceived to be as great as the loss in utility from losing $100 (e.g. Tversky and Kahneman, 1991). The implication is that, while there might be a net financial gain from liberalization of trade, there need not be a net utility gain. This is especially true if, as we saw in the previous section, the utility loss from unemployment cannot be easily compensated for by making up for the lost income. Even if there is a net utility gain, it is not clear how this should or could be redistributed over the various people affected, because, for example, two different people who are made redundant from the same job may suffer greatly different psychological consequences. Fernandez and Rodrik (1991) identified yet another possible source of status quo bias. They show that the bias can result from uncertainty, and in particular from uncertainty about who the individual winners and losers from a policy change might be. At least one economics study, Krebs et al. (2005), has investigated the theoretical effect of risk aversion on people’s attitudes to trade policy. The essential result is that risk aversion, coupled with uncertainty about the effects of trade liberalization, is likely to reduce one’s estimates of future utility. Tovar (2006) has produced and tested a mathematical model of how loss aversion might produce an anti-trade bias in government policy. A basic insight of her model is that industries which experience or face losses will be more motivated to lobby for protection than those who gain from a more open economy, because losses have a greater impact on utility than gains. The more energetic lobbying then results in more protectionism. Loss aversion does not simply imply aversion to free trade. Certainly, the reasoning above indicates that liberalizing trade may not produce overall gains in utility that parallel the gains to be made in the provision of goods and services. But the phenomenon of loss aversion applies to change in either direction. Indeed,




Figure 1. As the extent of the actual loss increases, the loss of utility does not increase proportionally



A. Large utility loss from small cost

B. Proportionately smaller utility loss from larger cost

there is some reason to expect restriction of trade might be even more vehemently opposed than liberalization. One reason for this comes simply from the classic economic argument. Although utility is not completely derived from the provision and consumption of goods and services, the relationship is certainly expected to be positive. Ceteris paribus, restrictions of trade should produce lower overall utility than its liberalization. A more complex argument arises from the shape of the utility function. Figure 1 depicts its likely form, as based mostly on the results of experiments where people have been asked to choose between different gambles (e.g. Tversky and Kahneman, 1991). Note, first, that the function graphically depicts loss aversion in that it is more steeply sloping for gains than losses. More subtly, however, the slope of the function changes as the extent of the loss (or gain) increases. In particular, the function is more steeply sloping for small losses (and gains) than for large losses. Now consider a restriction of trade that is carried out to protect workers and owners of a particular industry. Such changes will produce moderate financial gains to the relatively few affected workers and owners, and small financial losses to the relatively numerous rest of the community. However, when we consider utilities, the small financial losses (to the community) should

Psychology and opposition to free trade 33

be magnified to more substantial utility losses, and there is relatively more magnification for the small losers than for the moderate winners (the workers). Thus, the model appears to predict, because of the way that the utility losses change evermore gradually as the financial loss increases, that the overall utility losses when trade is restricted should be greater than when it is liberalized. Overall, loss aversion suggests people will resist changes in the status quo with respect to trade regulation. Loss aversion provides an account of why trade liberalization should be opposed, but it suggests trade restriction should also be opposed. (One additional complication could also be noted. If a worker’s job in an existing industry is threatened by new competition from outside under an unchanging tariff and importation regime, is the status quo the retention of his job or the retention of the existing trade policy ? Clearly, this depends on which might be framed as the status quo, as well as on who is asked.) Finally, although the recent empirical research carried out by those interested in behavioural economics and decision making produces both solid evidence for status quo bias and, more importantly, insight into the conditions that produce it, the existence of the bias has long been acknowledged. In the field of trade, Corden’s (1974 : 5) conservative social welfare function recognized that an increase in income is likely to produce a smaller positive effect on welfare than the negative effect produced by a similar income decrease. Corden, too, saw this function as an explanation of reluctance to reduce tariffs.

2.3 Altruism and parochialism It has long been noticed that those who produce goods at a price greater than they could be imported for tend to oppose the importation of such goods, and they band together so as to lobby governments and persuade the general public that the importation should be restricted. There has also been attention paid to the occasional ability of such groups to lobby successfully, even though the benefits of restricting imports may only be enjoyed by a small minority of the electorate (e.g. Olsen, 1982). From a psychological point of view, the question is not so much why the producers or the politicians who represent them behave as they do, but why the public might be happy to go along with them. Answers to this question include the frequent cohesion and organization of the producers (Olson, 1982), the probability that individual consumers suffer a small rather than a large financial loss from the import restrictions (although, as we have just seen, research on loss aversion only supports this reasoning for maintenance of existing trade barriers), and frequent uncertainty about the benefits that arise from lifting them (Fernandez and Rodrik, 1991). Another possibility is that the rest of the community might be altruistic. The small material or financial loss is outweighed by utility derived from the belief that they are helping their compatriots.



There is good evidence that altruism is important in people’s stances on political and economic questions. For example, many voters in a referendum on a river channelization project in Roanoke, Virginia supported the project even though they would receive no direct benefit from it (Shabman and Stephenson, 1994). A key determinant of election outcomes appears to be how the nation as a whole is faring rather than how people’s own individual economic interest might be affected (Funk and Garcia-Monet, 1997). Rotemberg (2003) has provided an economic analysis of how altruism might lead to protectionism in both direct and indirect democracies. Thus, we would anticipate the ordinary citizen’s views on trade issues to have an altruistic component. But who should benefit from this altruism ? Consider someone who is thinking about the removal of import restrictions on imported clothing. She has no personal connection with the clothing industry, and would personally benefit from the removal of restrictions. If she decides altruistically, however, she could consider only the interests of her compatriot clothing workers, who would be harmed by the change, or she might also consider the interests of clothing workers in foreign countries, who would benefit from it. This kind of choice was discussed by Olson (1965, 1982) and has been modelled experimentally in the ‘layered prisoner’s dilemma ’ (Bornstein and Ben-Yossef, 1994 ; Schwartz-Shea and Simmons, 1990). In this set-up, participants are given the choice of ‘cooperating ’ or ‘ defecting’. Cooperation disadvantages the participant and an ‘out-group ’ but benefits members of the ‘ in-group ’ ; defecting does the opposite. The behaviour varies with the parameters of the experiment (for example, with regard to whether the participants can discuss the situation), but it is quite common for participants to choose a course of action which benefits the ingroup at the expense both of the participant him- or herself and the ‘wider community ’ (ingroup+outgroup). Such behaviour may be termed parochialism (Schwartz-Shea and Simmons, 1991). Baron (2001) investigated whether some parochialism might be the consequence of a ‘ self-interest ’ illusion in which people see themselves as benefiting from a sacrifice they make for their group, even when in fact they do not. He found the self-interest illusion increased from a simple individual and one-group situation to the individual and two-group situation of the layered prisoner’s dilemma, and that it decreased when the participants were made to calculate the actual outcomes mathematically. Extrapolation to trade suggests people might favour import restrictions that benefit their compatriots, even though these restrictions are neither in the interests of the individual nor of overall benefit to the world. There is some empirical support for this extrapolation. Baron and Kemp (2004) had respondents complete a scale measuring their tendency to think of the global versus the nation good (by including such questions as ‘ National governments should put the interests of the world as a whole ahead of their own national interests ’). Attitudes towards the restriction of imports were also measured. As expected, there was a moderate

Psychology and opposition to free trade 35

positive correlation between the tendency to favour the national over the global good and the tendency to favour import restrictions. To summarize, individuals often think beyond their own narrow self-interest when considering government economic policies in areas such as trade. However, this altruism may extend only to co-nationals, rather than to everyone affected by these government decisions. The circumstances under which this extension may occur are not well understood at present.

2.4 Trade and bargaining Suppose two countries, A and B, have trade barriers against each other. B is unwilling to lower its own trade barriers. What should A do ? Economic theory is not completely clear-cut on this issue, but many economists believe A would gain by lowering its own trade barriers anyway (e.g. Bhagwati, 2002). Similarly, if B were to unilaterally increase its trade barriers, economic theorists often suggest that A would gain little by raising its own in retaliation. In practice, it seems uncommon (although not unknown) for countries to lower their trade barriers unilaterally, and it is quite common for countries to retaliate against the trade restrictions of other countries by imposing their own. Both reluctance to liberalize trade unilaterally and retaliation to barriers imposed by another country are understandable in terms of research into how individuals perceive fairness and how individuals react to apparently unfair treatment. Earlier, a little of the research on fairness, and in particular research on the ultimatum game, was outlined. Very briefly, this work indicates that people are often reluctant to accept deals in which it appears that another party gains considerably more than they do. It is completely plausible, especially in light of the research discussed in Section 2.3, that an individual – whether as a trade negotiator or simply a citizen – would also be unwilling to accept a trade deal that is offered their country (or perhaps other group) if another country is perceived to gain more from the deal. So, for example, suppose an individual perceives her own country, A, as gaining 2 units of perceived utility from opening its borders to B’s imports and 4 units of perceived utility from B opening its borders to A’s exports. B is perceived to have similar expected gains. Clearly, A would benefit from unilaterally opening its borders regardless of what B does. However, the ultimatum game results suggest that a citizen of A might be hostile to this action because B is perceived to ‘unfairly ’ gain more from it. (For further discussion of the possible impact of considerations of fairness on trade policy, see Davidson et al., 2006.) There is some evidence indicating ordinary people do think along these lines. Baron and Kemp (2004) found people’s willingness to restrict imports from other countries was heavily dependent on the other country’s behaviour. Indeed, perceived reciprocity in this study produced larger effects on willingness to restrict than either the nature of the present status quo or whether one’s own compatriots were likely to suffer unemployment.



2.5 Fixed pie beliefs The fixed pie belief occurs when two or more concerned parties believe that a gain to one must be a loss to the others. There are real-life situations where the belief is accurate ; most obviously, when an actual pie must be divided between two or more people. But there are also situations where a gain to one party may be in the interests of other, apparently opposing parties. For example, a union’s desire to protect workers at an industrial plant from injury may also benefit management. It is well documented that people often do hold and act on inappropriate fixed pie beliefs, especially in negotiation situations (see, e.g., Bazerman et al., 1999, 2001 ; Pinkley et al., 1995). Thus, management might oppose a union proposal simply because it wrongly believes that what is good for the union must be bad for them. Economic theory has, of course, long held that the benefits of trade come from enlargement of the pie. However, it is quite likely that many people do not see it this way and instead believe that countries that benefit from trade only do so by exploiting other countries, not by benefiting them as well. Note here, incidentally, that, as for bargaining, although it is the individual’s views that are important, her views concern the benefit to her country rather than to herself directly.

2.6 Failure to understand comparative advantage Part of the argument for free trade derives from Ricardo’s principle of comparative advantage. Ricardo argued that one need not be the most efficient producer of any commodity at all in order to have something to offer in the trade process. His original example (Ricardo, 1817/1971 : 7) considered an economy of two countries, England and Portugal, both of which can produce two commodities, cloth, and wine. Ricardo posited that England can produce cloth using 100 units of labour and wine using 120 units of labour ; Portugal can produce cloth with 90 units of labour and wine with 80 units of labour. Although Portugal can produce both commodities more efficiently than England, the production and consumption of both commodities is maximized if Portugal specializes in wine and England in cloth. One way of summarizing the principle is to say that the determinant of where the goods should be produced is the ratio of the costs of production (comparative advantage) rather than the absolute costs (absolute advantage). Ricardo’s principle is not obvious. Samuelson, in response to the challenge of finding a theory in the social sciences that was both true and nontrivial, famously chose this principle, remarking ‘ that it is not trivial is attested by the thousands of important and intelligent men who have never been able to grasp the doctrine for themselves or to believe it after it was explained to them. ’ (Samuelson, 1972 : 683). Krugman (1994) gives a number of examples of eminent people apparently misunderstanding it, and goes on to suggest that resistance arises because it is not new, because understanding it requires knowledge of a number of other economic concepts, and because it is grounded in a mathematical model.

Psychology and opposition to free trade 37

If ‘ important and intelligent men’ have difficulty understanding and applying the principle, it would not be surprising to find the average person also had a limited understanding of it, or to find those who misunderstand more likely to advocate restrictions on international trade. Baron and Kemp (2004) administered various tests of the principle to members of the general public, presenting them with scenarios such as the following and then asking how production should be allocated so as to maximize production : Two branches of a firm, both within one country, make parts for a motorbike. Branch A can make engines at a cost of $100 each and frames at a cost of $80 each. Branch B can make engines at a cost of $80 [for an easy, $180 for a hard condition] each and frames at a cost of $100 each _ Think about how you would allocate work to get the most production.

People’s answers, especially where they had to consider comparative rather than absolute advantage, generally did not maximize production, indicating imperfect understanding of the principle. More importantly, those individuals with lesser understanding of the principle tended to favour restricting imports. A corollary to the principle of comparative advantage may also be unappreciated by lay people. Suppose for the moment that in Ricardo’s example only labour costs were important in determining the cost of cloth and wine (i.e. the advantages come about because of differing worker skills). In this case, maximizing production would come about through wage inequalities : the more efficient Portuguese workers would be paid more than their English counterparts, driving production of cloth out of Portugal and into England. Opposition to trading with poorer countries on the grounds either that their workers are exploited (because they are less well paid than one’s own) or that their workers have an unfair advantage (because they are willing to work for less than one’s own) also constitutes resistance to applying the principle of comparative advantage. It is thus an interesting question whether people who oppose trade because of differing wage levels do so because they misunderstand how the principle of comparative advantage might work.

3. Putting together and assessing the psychological factors

3.1 Combining the factors The effects of the factors just reviewed on people’s thinking about trade appear to be cumulative rather than independent. We would also anticipate the psychological factors to combine with the non-psychological factors. For example, special interest groups lobbying for protection of their particular industry are likely to appeal to the altruism (or parochialism) of their compatriots. Two further examples illustrate possible combinations of the psychological factors: Jane may oppose her country importing clothing from a poorer country because clothing workers in her own may become unemployed. Her opposition would thus



depend on an altruistic but parochial concern for other people in her own country, and on her belief that people derive utility from the work they do as well as from their consumption. Bruce opposes his country importing from Ruritania because Ruritania places heavy import restrictions on goods produced in his own country. Bruce’s opposition depends on both his perceptions of fairness, and on his applying his judgements of fairness to his country (i.e. parochially). His views may also depend partly on a fixed-pie belief.

3.2 Rationality and irrationality Section 2.2 briefly introduced some of the empirical and theoretical research work carried out by decision theorists and behavioural economists. The work carried out by these researchers is too diverse and innovative to categorize simply, but much has concentrated on decisions that appear not to be rational. So, for example, the ‘ Asian flu ’ experiments show that people make different decisions depending on whether the framing of the problem focuses on survivors or victims of the outbreak of a fatal disease (Tversky and Kahneman, 1981). People appear to act irrationally in that they decide inconsistently or not in their own best interest. We can ask whether the six factors outlined above also feature irrational processes. There is a wealth of evidence indicating that people’s happiness and well-being strongly relate to work satisfaction. Individual utility functions depend on the nature of work as well as on the money people earn that they can then use for consumption. Hence, it must be rational for people to consider effects on their work as well as on their consumption when they think about trade. At the other extreme, a misplaced belief that trade gains constitute a fixed pie or failure to understand the principle of comparative advantage can probably be considered irrational, no matter how common these errors might be. The other factors are more difficult to categorize. Tversky and Kahneman (2000 : 157) ask whether loss aversion is irrational, and suggest that at least some components of it appear to be rational. A value function that is more steeply sloping on the losses side ‘ appropriately reflects three basic facts : organisms habituate to steady states, the marginal response to change is diminishing, and pain is more urgent than pleasure. The asymmetry of pain and pleasure is the ultimate justification of loss aversion in choice.’ Thus, ‘ a decision maker who seeks to maximize the experienced utility of outcomes is well advised to assign greater weights to negative than positive consequences ’ (Tversky and Kahneman, 2000 : 157). On the other hand, an individual’s loss aversion is influenced by how he or she brackets the experience with others. Thaler (1999) considers the case of people who might reject a single bet (for example, a 50 % chance or either winning $200 or losing $100) but accept a series of such bets. The difficulty for a rational explanation here arises with why an individual with a reasonable total wealth would reject the single bet. A status quo bias arising from an individual’s inability to

Psychology and opposition to free trade 39

predict his future tastes also may not be rational, especially if (as in Kahneman and Snell, 1992) the situation is one in which the individual does appear to have enough information to predict a taste change. It is similarly not easy to know whether altruism and parochialism are rationally considered in the context of trade or not. People are in practice more likely to be sympathetic to the fate of their countrymen (even if they do not know them personally) than of those who live in other countries and about whom they are likely to be less well informed. On the other hand, people who take part in layered prisoner’s dilemma games are sometimes subject to a self-interest illusion in which they see themselves as benefiting more from the general good fortune of their own group than they actually do. When this is brought to their attention, their cooperation with the group decreases (Baron, 2001). Furthermore, the apparent worthiness of the out-group can be affected by discussion (Schwartz-Shea and Simmon, 1990). Extension of these experimental findings to trade suggests, for example, that the relative weight people might give to the welfare of clothing workers in their own country (and why not in their own city ?) versus those in a foreign country could be affected by attitudinal manipulations and framings of different sorts. At this point, it becomes more difficult to consider fellow feelings as strictly rational. Finally, with respect to bargaining, is it rational to reject a good outcome (the consequence of ridding oneself of import restrictions) in favour of a better one (ridding oneself of import restrictions and having other countries do so as well) ? What should an individual think of this ? Trade negotiations are not a single ultimatum game ; it is quite possible that country A might make a larger gain in utility by holding out until country B reciprocates. Indeed, it is conceivable that A might gain more utility from trade negotiations with B (and C and D) by not accepting short-term gains and refusing to accept deals that give other countries more utility than it receives itself (see Frank, 1988). One consideration in this case may be the relative power of the countries. Seen from the national point of view, if A is much smaller and less powerful than B, A might not be able to influence B’s policy at all (Rabin, 1997). Of course, whether an individual citizen of A would take this or similar considerations into account is quite another matter. In short, the question is not simple to answer. More could obviously be said about the rationality of the different psychological factors. One might even debate what it means to be rational (e.g. Singer, 1999). However, it should at least be clear that the psychological factors differ in the apparent extent of their rationality, and that taking account of psychological factors in trade is not the same thing as simply considering factors that are completely irrational.

3.3 Psychology and politics Broadly speaking, there are two different ways for lay opinions to affect trade policy. In the first place, trade negotiators and politicians are themselves human,



and often have neither specialized economic training nor a substantial personal interest in any particular industry. Thus, if there are general psychological reasons to resist free trade, we would expect decision makers or negotiators to share at least some of them with ordinary citizens. Note, too, that when ordinary citizens are questioned about trade in the research reviewed above, they are often effectively asked to put themselves in the place of a decision maker. For example, Scheve and Slaughter (2001) asked people whether they favoured placing limits on imports ; Baron and Kemp (2004) asked people to indicate their preferences for importing goods from a foreign country under different conditions. Secondly, in a democratic society the views of ordinary citizens influence those responsible for negotiating (or not negotiating) trade deals, either because the negotiators are elected and thus constrained by voters, or because the negotiators are to a greater or less extent controlled by the elected representatives. The extent of voters’ influence in any area of government policy is debatable. For example, the views of the majority, or those of the median voter (the voter who is right in the middle on a particular issue), are not decisive in determining the level of provision of public goods (e.g. Kemp, 2002). The social survey results presented in the introduction indicate that trade policy is no exception and that governments, or at any rate many western governments in the 1990s, do not follow the generally more protectionist policy desired by the median voter. Possibly government trade policies sometimes reflect a compromise between those desired by the electorate, and those advocated by economic experts. As the latter are more likely to improve general welfare, and the electoral fate of governments seems to be determined mainly by the overall state of the economy (e.g. Alvarez et al., 2000), this seems a plausible way for governments to behave. Davidson et al. (2006) suggest that keeping the public away from trade negotiations serves to advance more liberal trade policies than would otherwise be achieved. Finally, Hiscox’s (in press) demonstration that the extent of people’s surveyed opposition to free trade is subject to framing effects indicates that governments might reasonably expect to be able to reduce the opposition by appropriate marketing.

4. Conclusions A full understanding of how psychological factors interact with economic and political factors to produce people’s attitudes to trade would have important implications. Some factors, for example, the representation in utility functions of employment values that are not easily compensated for by income, suggest changes to economic models of the benefits of trade. On the more practical side, knowing more about psychological factors might help counter negative aspects of changes to trade policy. Greater education of the public, with respect to the fixed pie fallacy or the principle of comparative advantage, might also be suggested. Our present knowledge of psychological factors in trade is incomplete, perhaps primitive. Little empirical work has been done, and the previous discussion has

Psychology and opposition to free trade 41

often been based on extrapolating principles derived from studying other areas. For example, it is plausible, but it has not been demonstrated, that many people believe that one country’s trade benefits arise only if other countries lose proportionally. However, a start has been made, and interesting results have already emerged from the analysis of surveys containing attitude to trade questions. For example, people with higher skill and education levels are generally less likely to favour new trade barriers, at least in more developed economies (Mayda and Rodrik, 2005 ; Scheve and Slaughter, 2001). This result could be interpreted as a rational response to the increased likelihood for the lower skilled to lose their jobs to cheap foreign competition, but it is also possible that higher skilled and educated people are more pro-trade because of their greater exposure to economic ideas. Hainmueller and Hiscox (2006) suggest the latter is more likely, because those who have retired from the work force retain the attitudes appropriate to those with the same skill and education level who are employed. However, this result could also be explained by retirees feeling disproportionately altruistic to workers similar to themselves. Moreover, Baron and Kemp (2004) found understanding of comparative advantage to be relatively poor even among the educated. Further empirical research is needed to establish both the psychological factors that do affect people’s trade attitudes and their relative importance. The psychological factors reviewed generally predict that ordinary people might often be less keen on trade liberalization measures than economists, and it is very unlikely that this list exhausts the potential factors. When we add non-psychological factors (e.g. capture of the political decision making by lobby groups, environmental factors), the brief conclusion must be that the opposition of many people to free trade is, if anything, over-determined. Which, then, are the crucial factors in this opposition ? Are these important factors to be countered by education or institutional change or do they suggest some fundamental rethinking of the optimal trade policy ? The task of determining the relative importance of the factors is more difficult because, as we have seen, they often operate together rather than independently. Moreover, we should not expect different factors necessarily to have the same importance in rich and poor countries, in large and small countries, or all communities within the same country. However, the apparent worldwide hardening of opposition to free international trade indicates that finding answers to these questions is important. References Alston, R. M., J. R. Kearl, and M. B. Vaughan (1992), ‘ Is There a Consensus among Economists in the 1990s? ’, American Economic Review, 82(2): 203–209. Alvarez, R. M., J. Nagler, and J. R. Willette (2000), ‘Measuring the Relative Impact of Issues and the Economy in Democratic Elections’, Electoral Studies, 19(2–3): 237–254. Baron, J. (2001), ‘ Confusion of Group Interest and Self-Interest in Parochial Cooperation on Behalf of a Group’, Journal of Conflict Resolution, 45(2): 283–296.



Baron, J. and S. Kemp (2004), ‘Support for Trade Restrictions, Attitudes, and Understanding of Comparative Advantage’, Journal of Economic Psychology, 25(5): 565–580. Bazerman, M. H., J. Baron, and K. Shonk (2001), ‘You Can’t Enlarge the Pie’ : Six Barriers to Effective Government, New York: Basic Books. Bazerman, M. H., D. A. Moore, and J. J. Gillespie (1999), ‘ The Human Mind as a Barrier to Wiser Environmental Agreements ’, American Behavioral Scientist, 42(8): 1277–1300. Bhagwati, J. (1991), The World Trading System at Risk, Princeton, NJ: Princeton University Press. Bhagwati, J. (ed.) (2002), Going Alone: The Case for Relaxed Reciprocity in Freeing Trade, Cambridge, MA: MIT Press. Bojilova, A. (2006), Organizational Attitudes and Behaviours in Post-Communist Europe, Christchurch, NZ: University of Canterbury Applied Psychology Project. Bornstein, G. and M. Ben-Yossef (1994), ‘ Cooperation in Intergroup and Single-Group Social Dilemmas’, Journal of Experimental Social Psychology, 30(1): 52–67. Broude, T. (2005), ‘ Taking ‘‘Trade and Culture’’ Seriously: Geographical Indications and Cultural Protection in WTO Law’, 1 May 2005, Available at SSRN: or DOI: 10.2139/ssrn.714981 Corden, W. M. (1974), Trade Policy and Economic Welfare, Oxford: Clarendon. Darity, W. A., A. H. Goldsmith, and J. R. Veum (1999), ‘ Unemployment and Well-Being ’, in P. Earl and S. Kemp (eds.), The Elgar Companion to Consumer Research and Economic Psychology, Cheltenham: Edward Elgar. Davidson, C., S. Matusz, and D. Nelson (2006), ‘Fairness and the Political Economy of Trade’, The World Economy, 29(8): 989–1004. Diener, E., E. M. Suh, R. E. Lucas, and H. L. Smith (1999), ‘Subjective Well-Being: Three Decades of Progress’, Psychological Bulletin, 125(2): 276–302. Dixit, A. K. and V. Norman (1986), ‘Gains from Trade without Lump-Sum Compensation ’, Journal of International Economics, 21(1–2): 111–122. Eisenberg, P. and P. F. Lazarsfeld (1938), ‘The Psychological Effects of Unemployment’, Psychological Bulletin, 35: 358–390. Feenstra, R. C. (2004), Advanced International Trade: Theory and Evidence, Princeton, NJ: Princeton University Press. Fernandez, R. and D. Rodrik (1991), ‘ Resistance to Reform: Status Quo Bias in the Presence of Individual-Specific Uncertainty’, American Economic Review, 81(5): 1146–1155. Frank, R. H. (1988), Passions within Reason: The Strategic Role of the Emotions, New York: Norton. Frey, B. S., W. W. Pommerehne, F. Schneider, and G. Gilbert (1984), ‘Consensus and Dissention among Economists: An Empirical Inquiry’, American Economic Review, 74(5): 986–994. Frey, B. S. and A. Stutzer (2000), ‘Happiness, Economy and Institutions’, The Economic Journal, 110(466): 295–307. Funk, C. L. and P. A. Garcia-Monet (1997), ‘ The Relationship between Personal and Rational Concerns in Public Perceptions about the Economy’, Political Research Quarterly, 50(2): 317–342. Gomes, L. (2003), The Economics and Ideology of Free Trade: A Historical Review, Cheltenham: Edward Elgar. Gu¨th, W., R. Schmittberger, and B. Schwartz (1982), ‘An Experimental Analysis of Ultimatum Bargaining’, Journal of Economic Behavior and Organization, 3(4): 367–388. Gu¨th, W. and R. Tietz (1990), ‘Ultimatum Bargaining Behavior – A Survey and Comparison of Experimental Results’, Journal of Economic Psychology, 11(3): 417–449. Hainmueller, J. and Hiscox, M. J. (2006), ‘ Learning to Love Globalization: Education and Individual Attitudes toward International Trade’, International Organization, 60: 469–498. Henrich, J., R. Boyd, S. Bowles, C. Camerer, E. Fehr, H. Gintis, R. McElreath, M. Alvard, A. Barr, J. Ensminger, H. N. Smith, K. Hill, F. Gil-White, M. Gurven, F. W. Marlowe, J. Q. Patton, and D. Tracer (2005), ‘ ‘‘Economic Man’’ in Cross-cultural Perspective : Behavioral Experiments in 15 Small-Scale Societies’, Behavioral and Brain Sciences, 28(6): 795–815. Hiscox, M. J. (in press), ‘Through a Glass and Darkly: Attitudes toward International Trade and the Curious Effects of Issue Framing’, International Organization.

Psychology and opposition to free trade 43 Horan, R. D., E. Bulte, and J. F. Shogren (2005), ‘ How Trade Saved Humanity from Biological Extinction: An Economic Theory of Neanderthal Extinction ’, Journal of Economic Behavior and Organization, 58(1): 1–29. Irwin, D. A. (2005), Free Trade under Fire, 2nd edn, Princeton: Princeton University Press. Jahoda, M., P. F. Lazarsfeld, and H. Zeisel (1971), Marienthal: The Sociography of an Unemployed Community, Chicago: Aldine. Jansen, M. and A. Turrini (2004), ‘ Job Creation, Job Destruction, and the International Division of Labor’, Review of International Economics, 12(3): 476–494. Kahneman, D. (1994), ‘ New Challenges to the Rationality Assumption’, Journal of Institutional and Theoretical Economics, 150(1): 18–36. Kahneman, D., J. L. Knetsch, and R. H. Thaler (1991), ‘The Endowment Effect, Loss Aversion, and the Status Quo Bias’, Journal of Economic Perspectives, 5(1): 193–206. Kahneman, D. and J. Snell (1992), ‘Predicting a Changing Taste: Do People Know what They Will Like?’, Journal of Behavioral Decision Making, 5 : 187–200. Kahneman, D. and A. Tversky (1982), ‘ The psychology of preferences’, Scientific American, 246: 160–173. Kahneman, D. and A. Tversky (eds.) (2000), Choices, Values, and Frames, Cambridge: Cambridge University Press. Kemp, M. C. and H. Y. Wan (1986), ‘ Gains from Trade with and without Lump-Sum Compensation’, Journal of International Economics, 21(1–2): 99–110. Kemp, S. (2002), Public Goods and Private Wants: The Psychology of Government Spending, Cheltenham: Edward Elgar. Kletzer, L. G. (1998), ‘Job Displacement’, Journal of Economic Perspectives, 12(1): 115–136. Krebs, T., P. Krishna, and W. Maloney (2005), ‘Trade Policy, Income Risk and Welfare’, World Bank Policy Research Working Paper No. 3622. Krueger, A. O. (2004), ‘Wilful Ignorance: The Struggle to Convince Free Trade Sceptics’, World Trade Review, 3(3): 483–493. Krugman, P. (1994), ‘Ricardo’s Difficult Idea’, Downloaded from ricardo.htm on 20 June 2006. Loewenstein, G. and D. Adler (1995), ‘A Bias in the Prediction of Tastes ’, Economic Journal, 105: 929–937. Lucas, R. E., A. E. Clark, Y. Georgellis, and E. Diener (2004), ‘Unemployment Alters the Set Point for Life Satisfaction ’, Psychological Science, 15(1): 8–13. Madrian, B. C. and D. F. Shea (2001), ‘ The Power of Suggestion: Inertia in 401(k) Participation and Savings Behaviour’, Quarterly Journal of Economics, 116(4): 1149–1187. Mander, J. and E. Goldsmith (1996), The Case against the Global Economy, San Francisco: Sierra Club Books. Mayda, A. M. and D. Rodrik (2005), ‘Why are Some People (and Countries) More Protectionist than Others? ’ European Economic Review, 49: 1393–1430. Mayer, W. (1984), ‘Endogenous Tariff Formation’, American Economic Review, 74(5): 970–985. McCaffery, E. J. and J. Baron (forthcoming), ‘ Heuristics and Biases in Thinking About Tax’ National Tax Association Annual Proceedings. McGillicuddy-DeLisi, A. V., C. Watkins, and A. J. Vinchur (1994), ‘The Effect of Relationship on Children’s Distributive Justice Reasoning’, Child Development, 65(6): 1694–1700. Olson, M. (1965), The Logic of Collective Action: Public Goods and the Theory of Groups, Cambridge, MA: Harvard University Press. Olson, M. (1982), The Rise and Decline of Nations : Economic Growth, Stagflation, and Social Rigidities, New Haven: Yale University Press. Papageorgiou, D., A. Choski, and M. Michaely (1990), Liberalizing Foreign Trade in Developing Countries : The Lessons of Experience, Washington, DC : World Bank. Pinkley, R. L., T. L. Griffith, and G. B. Northcraft (1995), ‘‘‘Fixed Pie’’ a la mode: Information Processing and the Negotiation of Suboptimal Agreements’, Organizational Behavior and Human Decision Processes, 62(1): 101–112.



Rabin, M. (1997), ‘ Bargaining Structure, Fairness and Efficiency’, Department of Economics, Paper E00’280, University of California, Berkeley. Ricardo, D. (1971), On the Principles of Political Economy, and Taxation, Harmondsworth: Penguin. (Originally published in 1817.) Ridley, M. (1997), The Origins of Virtue: Human Instincts and the Evolution of Cooperation, New York: Viking. Roberts, R. D. (2001), The Choice: A Fable of Free Trade and Protectionism, Upper Saddle River, NJ: Prentice Hall. Rotemberg, J. J. (2003), ‘Commercial Policy with Altruistic Voters’, Journal of Political Economy, 111(1): 174–201. Rowley, C. K. (2001), ‘ The International Economy in Public Choice Perspective ’, in W. F. Shughart and L. Razzolini (eds.), The Elgar Companion to Public Choice, Cheltenham: Edward Elgar. Samuelson, P. A. (1972), ‘The Way of an Economist’, reprinted in R. C. Merton (ed.), The Collected Papers of Paul A. Samuelson, Cambridge, MA: MIT Press. Samuelson, W. and R. Zeckhauser (1988), ‘ Status Quo Bias in Decision Making’, Journal of Risk and Uncertainty, 1(1): 7–59. Scheve, K. F. and M. J. Slaughter (2001), ‘What Determines Individual Trade-Policy Preferences? ’, Journal of International Economics, 54: 267–292. Schwartz-Shea, P. and R. T. Simmons (1990), ‘The Layered Prisoners’ Dilemma: Ingroup vs. MacroEfficiency’, Public Choice, 65(1): 61–83. Schwartz-Shea, P. and R. T. Simmons (1991), ‘Egoism, parochialism, and universalism’, Rationality and Society, 3(1): 106–132. Shabman, L. and K. Stephenson (1994), ‘ A Critique of the Self-Interested Voter Model: The Case of a Local Single Issue Referendum’, Journal of Economic Issues, 28(4): 1173–1186. Singer, A. (1999), ‘ General Theory of Rationality’, in P. E. Earl and S. Kemp (eds.), The Elgar Companion to Consumer Research and Economic Psychology, Cheltenham: Edward Elgar. Smith, A. (1904), An Enquiry into the Nature and Causes of the Wealth of Nations, London: Methuen. (Originally published in 1776.) Thaler, R. H. (1999), ‘ Mental Accounting Matters’, Journal of Behavioral Decision Making, 12(3): 183–206. Tovar, P. (2006), ‘ The Effect of Loss Aversion on Trade Policy: Theory and Evidence’, Paper presented at the Spring 2006 Midwest International Economics Conference. Downloaded from http://gemini. on 13 October 2006. Trefler, D. (2001), ‘The Long and Short of the Canada–US Free Trade Agreement ’, National Bureau of Economic Research Working Paper 8293, May. Trivers, R. L. (1971), ‘ The evolution of reciprocal altruism’, Quarterly Review of Biology, 46(1): 35–57. Tversky, A. and D. Kahneman (1981), ‘ The Framing of Decisions and the Psychology of Choice’, Science, 211(4481): 453–458. Tversky, A. and D. Kahneman (1991), ‘Loss Aversion in Riskless Choice: A Reference-Dependent Model’, Quarterly Journal of Economics, 106(4): 1039–1061. Tversky, A. and D. Kahneman (2000), ‘ Loss Aversion in Riskless Choice: A Reference-Dependent Model’, in D. Kahneman and A. Tversky (eds.), Choices, Values and Frames, Cambridge: Cambridge University Press. Winkelmann, L. and R. Winkelmann (1998), ‘Why are the Unemployed so Unhappy? Evidence from Panel Data’, Economica, 65(257): 1–15.

World Trade Review (2007), 6 : 1, 45–87 Printed in the United Kingdom f Antonis Antoniadis doi:10.1017/S1474745606003090

The European Union and WTO law: a nexus of reactive, coactive, and proactive approaches ANTONIS ANTONIADIS* Durham University and Durham European Law Institute

Abstract : Ranging from the denial of direct effect to WTO law by the Court of Justice to a WTO-friendly legislative culture currently booming in the EU’s political institutions, different approaches towards WTO law have been adopted within the EU. This article classifies the different approaches into reactive, coactive, and proactive by drawing on their common characteristics. The principal aim is to explore the considerations shaping the development of the different approaches and to argue that these stem from the interaction between the judiciary and the legislature. In doing so, this article purports to provide a comprehensive view of the application of WTO law within the Community legal order.

1. Introduction The discussion concerning the application of WTO law in the Community legal order is not new. While the importance of the issue is undoubted, academic comment has largely focused on the judicial treatment afforded by the Court of Justice towards WTO law, and neglected the multifarious aspects of its application by the Union institutions.1 This article aims to fill the vacuum by engaging the Community’s political institutions alongside the Court of Justice and scrutinizing the interaction between the judiciary and the legislature. It thereby aims to better comprehend the broader ramifications of WTO law in the Community legal order. In order to achieve its aims, the present analysis will cut across institutional practice, draw the commonalities between the different approaches of the * I am grateful to Tony Arnull, Roy Davis, Robert Schu¨tze, Colin Warbrick, Marc Be´nitah, Stijn Billiet, Rafael Leal-Arcas, the anonymous referees and the members of the editorial board of World Trade Review for their comments. I am responsible for all remaining shortcomings. 1 Allan Rosas, ‘ Annotation of Case C-149/96 Portugal v. Council’ (2000) 37 Common Market Law Review 797 ; Francis Snyder, ‘ The Gatekeepers: The European Courts and WTO Law’ (2003) 40 Common Market Law Review 313. A notable exception can be found in Grainne de Bu´rca and Joanne Scott, ‘The Impact of the WTO on EU Decision-Making’ in G. de Bu´rca and J. Scott (eds.), The EU and the WTO: Legal and Constitutional Issues (Hart Publishing, 2003), p. 1.




institutions towards WTO law, attempt a classification on the basis of these common characteristics, and explore their interdependence. The Court of Justice’s approach will remain the focus of attention and will form the starting point of the discussion. It is well established that the Court of Justice has, in principle, denied the direct effect of WTO law in the Community and Member States’ legal orders. This approach is best described by the term reactive. The Court, however, has not always maintained such an approach and, in certain circumstances, has submitted the Community to the normative control of WTO law. This alternative approach, which is complemented by legislative initiatives taken by the Community’s political institutions, shall be called coactive. The main locus of the third, or proactive, approach can be found in the activity of the political institutions and the fact that WTO law has acquired a central role in the shaping of the Union’s internal and external policies. The critical examination of the political and legal considerations shaping this categorization will follow, while the impact of the interaction between the legislature and the judiciary will be reviewed.

2. Reactive approach

The Court of Justice Introducing the concept of direct effect The question of direct effect of WTO law forms part of the more general debate of the reception of general international law in the Community legal order.2 From an international law point of view, it is clear that international law will prevail in the case of conflict with domestic law.3 However, the determination of whether a certain provision is directly effective is a matter for domestic law, unless, of course, the parties to an international agreement have agreed to make its provisions directly effective.4 Domestic law will also determine the conditions under which a provision of international law can be directly effective. Clearly not an invention of the Court of Justice, the definition of direct effect is not without controversy. As a preliminary issue, it should be mentioned that direct effect has appeared in international law under several headings ranging from direct applicability and self-executing provisions to direct effect.5 In some of the original Community Member States, belonging predominantly to the monist legal tradition, international treaties were intended to confer rights on

2 J. M. Prinssen and A. Schrauwen (eds.), Direct Effect: Rethinking a Classic of EC Legal Doctrine (Europa Law, 2002). 3 Article 27 of the 1969 Vienna Convention on the Law of Treaties. 4 Jan Klabbers, ‘ International Law in Community Law: The Law and Politics of Direct Effect’ (2002) 21 Yearbook of European Law 263; Thomas Cottier, ‘A Theory of Direct Effect in Global Law’ in A. Von Bogdandy, P. C. Mavroidis and Y. Me´ny (eds.), European Integration and International Co-ordination : Studies in Transnational Economic Law in Honour of Claus Dieter Ehlermann (Kluwer Law International, 2002), p. 99 at p. 104. 5 Klabbers, supra note 4 at p. 272.

The European Union and WTO law 47

individuals.6 In fact, within the Community context, the ‘ objective ’ or ‘ classic’ definition of direct effect refers to a legal provision granting rights to individuals which must be upheld by national courts.7 It has been argued that direct effect not only provides the norm that governs a given case, it provides, in addition, the standard for legal review.8 The generic use of the concept of direct effect to include the standard of review has been particularly popular in the GATT/WTO context, owing to the participation of Member States in proceedings before the Court for which the classic definition of direct effect would clearly have been inadequate.9 Completing the demarcation of the concept of direct effect, one must note the development of the principle by the Court of Justice in so far as Community law is concerned. By holding in Van Gend en Loos10 that the Community is a special case and that the determination of whether Community law can be directly effective derives from Community law itself,11 the Court recognized the right-conferring qualities of Community norms in the ‘ new legal order ’ as a result of which rights were conferred upon individuals that national courts were bound to protect. The reference to Van Gend en Loos, while seemingly of limited input to the understanding of the concept of direct effect of international/WTO law in the Community and Member States legal orders, is, it is argued here, of cardinal importance to understand the indoctrination of academic comment by a communitarized understanding of the concept.12 Generally speaking, several arguments have been brought forward in an attempt to classify the Community legal order as monist or dualist following the traditional distinction in international law.13 The Court stated in Haegeman that an international agreement is an act of the institutions and that the provisions of such an agreement form an integral part of the Community legal order.14 Immediately received as a confession of monism, it was explained in Kupferberg15 that the meaning attached to the ‘integral part of the Community legal order’ proviso was that the Member States had not only assumed a responsibility for the fulfilment of 6 E. Denza, The Intergovernmental Pillars of the European Union (Oxford University Press, 2002), at p. 14. 7 P. Craig and G. de Bu´rca, EU Law : Texts, Cases and Materials, 3rd edition (Oxford University Press, 2002), at p. 180; T. Hartley, The Foundations of European Community Law, 4th edition (Oxford University Press, 1998), at p. 187. 8 S. Prechal, Directives in European Community Law (Oxford University Press, 1995), at p. 148. 9 The Court was criticized for connecting legality with direct effect. See, Ulrich Everling, ‘ Will Europe slip on Bananas? The Bananas judgment of the Court of Justice and national courts’ (1996) 33 Common Market Law Review 401 at p. 421; Fernando Castillo de la Torre, ‘The Status of GATT in EC Law Revisited’ (1995) 29 Journal of World Trade 53 at p. 58. 10 Case 26/62 NV Algemene Transport – en Expeditie Onderneming Van Gend and Loos v. Nederlandse Administratie der Belastingen [1963] ECR 1. 11 Klabbers, supra note 4 at p. 273. 12 For reasons, it is hoped, that will become clearer towards the end of this analysis. 13 For a general discussion see, I. Brownlie, Principles of Public International Law, 6th edition (Oxford University Press, 2003), at pp. 33–34 and 40–45. 14 Case 181/73 Haegeman v. Belgium [1973] ECR 449 at paras. 4 and 5. 15 Case 104/81 Hauptzollamt Mainz v. Kupferberg [1982] ECR 3641 at para. 13.



the agreement towards non-Member States but also towards the Community. Kupferberg has also been helpful for the elaboration of the criteria laid down by the Court for a provision of international law to develop direct effect. The Court added that the nature and structure of an international agreement may prevent an individual from invoking its provisions before a court in the Community.16 If this hurdle is overcome, international law provisions are required to be unconditional and sufficiently precise in the context of the agreement they form part of.17 Despite the criticism that the criterion concerning the nature and structure of the agreement received from commentators in the early stages,18 the Court has been consistent in this requirement.19 The two fundamental cases of International Fruit (GATT era) and Portuguese textiles (WTO era) will serve to illustrate the Court’s understanding of the WTO Agreement’s nature and purpose. The GATT crops: from International Fruit to Bananas The Court was faced with the GATT for a first time in International Fruit in the course of a preliminary ruling on validity.20 It opined that the Community, not a GATT Contracting Party, was bound by the GATT Agreement ;21 the Court’s own jurisdiction, however, depended on whether the GATT provisions were capable of conferring rights on individuals.22 The Court then found that owing to the great flexibility of the GATT provisions, and, in particular, those conferring the possibility of derogation, including taking safeguard measures when confronted with exceptional difficulties, and the inadequacy of the provisions for the settlement of the disputes between the Contracting Parties, individuals could not invoke GATT provisions before national courts.23 Accordingly, because of lack of direct effect, the Court was unable to examine the validity of the regulations. The Court maintained its position in subsequent rulings and, despite criticism,24 extended its findings to preliminary rulings on interpretation.25

16 Kupferberg at paras. 10–22. 17 Kupferberg at para. 23. 18 Henry G. Schermers, ‘Community Law and International Law’ (1975) 12 Common Market Law Review 77 at p. 80. 19 Case 12/86 Meryem Demirel v. Stadt Swa¨bisch Gmu¨nd [1987] ECR 3719 at para. 14. 20 Joined Cases 21/72 and 24/72 International Fruit Company NV and others v. Produktschap voor Groenten en Fruit [1972] ECR 1219. 21 The Court refrained from saying that the GATT forms an integral part of the Community legal order. See, Bourgeois, infra note 153 at p. 103. 22 International Fruit Company at paras. 4–9. 23 Ibid., at para. 27. 24 Ernst-Ulrich Petersmann, ‘ Application of the GATT by the Court of Justice of the European Communities ’ (1983) 20 Common Market Law Review 397 ; idem, ‘ The EEC as a GATT Member – Legal Conflicts between GATT Law and European Community Law’ in M. Hilf, F. Jacobs and E.-U. Petersmann, The European Community and the GATT (Kluwer, 1989), pp. 53–59; Kuilwijk, The European Court of Justice and the GATT Dilemma: Public Interest versus Individual Rights? (Nexed Editions, 1996). 25 Case 266/81 SIOT [1983] ECR 731; Joined Cases 267/81 and 269/81 SAMI [1983] ECR 801; Case C-469/93 Chiquita Italia [1995] ECR I-4533.

The European Union and WTO law 49

Member States had a vested interest in the direct applicability of the GATT in annulment proceedings before the Court of Justice because they could, in principle, challenge measures taken by the Council as GATT inconsistent. Viewed against the backdrop of qualified majority voting in the Council in the field of the common commercial policy and taking into account that no individual rights were involved, this appeared to raise a strong claim before the Court. The Court of Justice did not entertain this claim. In the first Bananas judgment, it held that : those features of the GATT, from which the Court concluded that an individual within the Community cannot invoke it in a court to challenge the lawfulness of a Community act, also preclude the Court from taking provisions of GATT into consideration to assess the lawfulness of a regulation in an action brought by a Member State under the first paragraph of Article 173 of the Treaty.26

In sum, individuals may not invoke GATT law in national courts, even more so when such legislation is invoked in order to challenge Community law. Neither can Member States and Community institutions invoke GATT law to challenge Community law.27 The Court’s thesis, as interpreted at the time in the light of International Fruit, meant that because the provisions of the GATT do not have direct effect, they cannot serve as a criterion for legality. With the benefit of hindsight, it can be argued here28 that the Court meant that the same defining features of the GATT preclude its provisions from both being invoked by individuals in national or Community courts and serving as a standard for the review of legality of secondary Community law. The WTO advent with Portuguese Textiles After the conclusion of the Uruguay Round of Multilateral Trade Negotiations, which led to the transformation of the GATT into the WTO, questions concerning the direct effect of WTO law in the Community legal order started inundating the Court in the form of requests for preliminary rulings made by national courts.29 In

26 Case C-280/93 Germany v. Council [1994] ECR I-4973 at para. 109. Emphasis added. 27 Piet Eeckhout, ‘The Domestic Legal Status of the WTO Agreement : Interconnecting Legal Systems’ (1997) 34 Common Market Law Review 11 at pp. 24–29. 28 Contra, Eeckhout who argues that the Court simply extended the principle of direct effect to cover also direct actions for annulment in P. Eeckhout, External Relations of the European Union: Legal and Constitutional Foundations (Oxford University Press, 2004), p. 249. 29 Case C-53/96 Herme`s International v. FHT Marketing Choice BV [1998] ECR I-3603; Case C-183/ 95 Affish v. Rijksdienst voor de keuring van Vee en Vlees [1997] ECR I-4315; Case C-147/96 Netherlands v. Commission [2000] ECR I-4723; Case C-106/97 Dutch Antillean Dairy Industry v. Douane-Agenten [1999] ECR I-5983; Case C-301/97 Netherlands v. Council [2001] ECR I-8853; Joined Cases C-300/98 and C-392/98 Parfums Christian Dior and Assco Geru¨ste GmbH [2000] ECR I-11037; Case C-377/98 Netherlands v. Council [2001] ECR I-7079; Case C-452/98 Dutch Antilles v. Council [2001] ECR I-8973; Case C-89/99 Schieving-Nijstad [2001] ECR I-5851; Case C-307/99 OGT Fruchthandelsgesellschaft [2001] ECR I-3159; Case T-52/99 T-Port v. Commission [2001] ECR II-981; Case T-1/99 T. Port v. Commission [2001] ECR II-465; Case T-18/99 Bocchi Food Trade v. Commission [2001] ECR II-913.



most of them, the Court refrained from grasping the nettle30 despite pressure from academics31 and the encouragement by Advocate Generals (AG)32 to revisit its case law on the matter. When Portugal brought an action for the annulment of Council Decision 96/386,33 alleging that it violated certain provisions of the GATT, the Agreement on Textiles and Clothing, the Agreement on Import Licensing and general principles of Community law,34 the Court was presented with a prime opportunity for a definitive resolution of the matter in the light of the new developments. AG Saggio suggested a change in policy. In his Opinion, after a brief overview of the WTO agreements relevant to the dispute, the Court’s case law regarding international agreements, generally, and the GATT, in particular, the AG held that only GATT rules must be found directly effective before the legality of Community acts can be tested against their provisions.35 Faced with the eleventh recital of the preamble to the Council Decision concluding the WTO Agreement, which declares that the Agreement and its annexes are not susceptible to direct invocation in Community or Member States’ courts, the AG, in the light of the relevant provisions of the Vienna Convention on the Law of Treaties and the Opinion of Advocate General Tesauro in Herme`s,36 dismissed its importance as nothing more than a policy statement which cannot affect the jurisdiction of either the Community or national courts to interpret and apply the rules contained in the WTO agreements.37 Consequently, he went on to analyse the novel characteristics of the WTO system, including the new system for the settlement of disputes, and argued that the inclusion of a system for the settlement of disputes within the WTO does not usurp the Court of Justice’s jurisdiction to interpret and apply WTO rules and to annul or sanction any internal measures which might be contrary to those rules.38 The final hurdle, the issue of reciprocity, raised by the

30 Case C-53/96 Herme`s International v. FHT Marketing Choice BV [1998] ECR I-3603 at para. 35; C-183/95 Affish v. Rijksdienst voor de keuring van Vee en Vlees [1997] ECR I-4315. 31 Pierre Pescatore, ‘ Free World Trade and the European Union: The reconciliation of interests and the revision of dispute resolution procedures in the framework of the WTO’ in Van Kappel and Heusel (eds.), Free World Trade and the European Union: The Reconciliation of Interests and the Review of the Understanding on Dispute Settlement in the Framework of the World Trade Organization (Vol 28, Series of Publications by the Academy of European Law in Trier, Bundesanzeiger, 2000), p. 9. 32 AG Tesauro in Case C-53/96 Herme`s International v. FHT Marketing Choice BV [1998] ECR I-3603; AG Cosmas in Case C-183/95 Affish v. Rijksdienst voor de keuring van Vee en Vlees [1997] ECR I-4315. 33 Council Decision of 26 February 1996 concerning the conclusion of Memoranda of Understanding between the European Community and the Islamic Republic of Pakistan and between the European Community and the Republic of India on arrangements in the area of market access for textile products O.J. L 153, 27/6/1996, p. 47. Notably, Portugal had voted against the conclusion of the Memoranda in the Council; see, Rosas, supra note 1 at p. 801. 34 Case C-149/96 Portugal v. Council [1999] ECR I-8395 at paras 53 et seq. 35 AG Saggio Opinion in Case C-149/96 Portugal v. Council [1999] ECR I-8395 at para. 18. 36 AG Tesauro in Case C-53/96 Herme`s International v. FHT Marketing Choice BV [1998] ECR I-3603 at para. 24. 37 AG Saggio Opinion in Case C-149/96 Portugal v. Council [1999] ECR I-8395 at para. 20. 38 Ibid., at para. 23.

The European Union and WTO law 51

fact that no other major trading partner in the WTO had granted direct effect to WTO law, was dismissed on the basis of the principle in adimplementi non est adimplendum.39 However, he then held that the obligation to apply WTO law does not extend to the violation of primary Community law. If WTO law is found to be in conflict with primary Community legislation, the latter should be upheld despite the risk of the Community suffering international responsibility.40 In its judgment, the Court held that : having regard to their nature and structure, the WTO agreements are not in principle among the rules in the light of which the Court is to review the legality of measures adopted by the Community institutions.41

This statement is foundational of the jurisprudence constante42 of the Court of Justice denying GATT/WTO law direct effect. Terminologically, it maintained the broad scope of the concept ‘ direct effect ’ to include the effect of WTO law as a standard of review.43 In order to arrive at this conclusion, the Court started with the statement from Kupferberg that the parties to an agreement have the power to determine the effect of the agreement and the means for its implementation within the parties ’ legal orders ; in the absence of an agreement thereon, it is up to the Court to rule on the matter.44 The Court then juxtaposed the system established under the WTO Agreement with the GATT and declared that, although the former differs significantly from the provisions of GATT 1947, it still accords considerable importance to the negotiation between the parties.45 This is proven by the fact that, while under Article 3.7 DSU measures found inconsistent with the agreement should be withdrawn, there is a possibility for compensation and, should this be declined, retaliation against the party whose legislation was found to be inconsistent with the WTO Agreement.46 The Court then laid down the two basic considerations for the denial of direct effect, namely the lack of 39 Enshrined in Article 60 of the Vienna Convention on the Law of Treaties and meaning that the failure of one party to observe its obligations under an agreement does not justify the other parties from applying the agreement among themselves. 40 AG Saggio Opinion in Case C-149/96 Portugal v. Council [1999] ECR I-8395 at para. 22. 41 Portugal v. Council at para. 47. 42 The case law should be considered settled for all purposes and this is exemplified by the fact that the Court, following Article 104(3) of its Rules of Procedure, responded to the Finanzgericht Hamburg by means of an order to the question raised in the context of a case relating to the Bananas litigation. Case C-307/99 OGT Fruchthandelsgesellschaft mbH v. Hauptzollamt Hamburg-St. Annen [2001] ECR I-3159. 43 See above. In Joined Cases C-300/98 and C392/98 Dior and Assco [2000] ECR I-11344 at para. 44 the Court held: ‘ For the same reasons as those set out by the Court in paragraphs 42 to 46 of the judgment in Portugal v. Council, the provisions of TRIPs, an annex to the WTO Agreement, are not such as to create rights upon which individuals may rely directly before the courts by virtue of Community law.’ 44 Portugal v. Council at paras. 34–35; cf. Kupferberg at paras. 17–18. 45 Portugal v. Council at para. 41. Kuijper mentions that a proposal granting direct effect to the WTO Agreement brought during the negotiations by the Swiss delegation was rejected. See, Peter Jan Kuijper, ‘The Conclusion and Implementation of the Uruguay Round Results by the European Community’ (1995) 6 European Journal of International Law 222. 46 Portugal v. Council supra note 34 at paras. 36–39.



reciprocity,47 and the freedom of the political institutions.48 The basic argument behind the principle of reciprocity is that the most important commercial partners of the Community do not allow their domestic courts to review the legality of their legislation against WTO law. It distinguished Kupferberg where the principle of reciprocity was dismissed on the premise that, while the WTO system is based on reciprocal and mutually advantageous arrangements, the EEC–Portugal FTA introduced a certain asymmetry of obligations.49 It further held that the grant of direct effect would lead to disuniform application of WTO rules in the different WTO Members.50 In addition, the freedom of political institutions would be usurped should the obligation of the Community to comply with the WTO rules be devolved to the judiciary.51 The Court finally held that such an interpretation is consonant with the Council’s view enshrined in the last recital of Council Decision 94/800.52 The principle of reciprocity. With regard to the principle of reciprocity, the Court essentially held that since the major trading partners of the Community do not grant direct effect to WTO law, in the interests of the principle of reciprocity, the Court of Justice was precluded from doing so.53 This was the first time that the Court had resort to the lack of reciprocity in order to deny granting direct effect to the provisions of an international agreement. Until that time, the Court referred to the principle in a dismissive manner.54 For instance, in Bresciani, a case concerning the interpretation of the Yaounde´ Convention, the Court introduced the principle only to explain that in the circumstances governing the Yaounde´ Convention strict reciprocity should not apply, because of the Community’s intention to assist the development of the associated countries by granting privileges to them.55 Accordingly, the rights that the nationals of the parties to the Yaounde´ Convention could invoke in Member States’ courts were not conditional on reciprocal treatment of Community citizens in these countries. By juxtaposing these cases with Portugal v. Council, it may be assumed that the principle of reciprocity constitutes a valid justification for the denial of direct effect in relations between equals and is not prone to being invoked in international agreements establishing asymmetric relations.

47 Portugal v. Council at paras. 42–45. 48 Portugal v. Council at paras. 40 and 46. 49 Portugal v. Council at para. 42. 50 Portugal v. Council at para. 45. 51 Portugal v. Council at para. 46. 52 Portugal v. Council at para. 48. 53 Portugal v. Council at para. 43. 54 Kupferberg at para. 18. 55 Case 87/75 Bresciani v. Amministrazione Italiana delle Finanze [1976] ECR 129 at para. 23. This was mainly a response to the AG Trabucchi’s argumentation on the lack of direct effect.

The European Union and WTO law 53

The central position of reciprocity in the Court’s denial of direct effect in Portugal v. Council has not been treated favourably by commentators.56 Eeckhout pointed out that there are certain inherent asymmetries within the WTO system as well, citing the examples of the Community as a Customs Union and the preferential treatment of developing countries within the WTO system.57 Given that it is not difficult to justify the Court’s position from a pragmatic perspective – the Community would find itself in a particularly disadvantageous position compared with the USA and Japan – the question is whether the Court’s position is defensible also from a doctrinal perspective. The asymmetry in the Community’s bilateral agreements features as an essential element of the design of those agreements. It expresses the Community’s decision to grant rights under the agreement to third, usually less developed, countries thereby encouraging trade and the integration of those countries into the multilateral trading system. In most cases, the substance of the provisions contained in these agreements replicated those of Community law.58 By contrast, the rights and obligations the Community and its Member States assumed when founding the WTO constitute a delicate balance, the result of mutually satisfactory concessions arrived at after eight years of negotiations. Asymmetric application thereof would disturb that balance and undermine the agreement struck. Accordingly, it was established by the Court in Portugal v. Council,59 and later clarified in Van Parys, 60 that it is in the interest of the appropriate interpretation and application of WTO law that the Court makes the grant of direct effect to WTO law subject to the principle of reciprocity.61 The freedom of political institutions. It was established in Portugal v. Council that the grant of direct effect would compromise the freedom of the Community’s political institutions within the WTO system. There are two aspects of the freedom of the political institutions : first, the external aspect, where the grant of direct effect is destined to weaken the negotiating strength of the institutions within the WTO in relation to the most important trading partners and, second, the internal aspect, the shift of the institutional balance in external trade matters from the Council and the Commission to the Court. The grant of direct effect would have the consequence that any Community legislative measure could be challenged before the Court of Justice as WTO incompatible.

56 Inter alia, Stefan Griller, ‘Judicial Enforceability of WTO Law in the European Union: Annotation to Case C-149/96, Portugal v. Council’ (2000) 3 Journal of International Economic Law 441. 57 Piet Eeckhout, ‘Judicial Enforcement of WTO Law in the European Union – Some Further Reflections ’ (2002) 5 Journal of International Economic Law 91 at p. 95. 58 Cottier, supra note 4 at p. 108. 59 Portugal v. Council at para. 45. 60 Case C-377/02 Le´on Van Parys NV v. Belgisch Interventie – en Restitutiebureau (BIRB) [2005] ECR I-1465 at para. 53. 61 Antonis Antoniadis, ‘The Chiquita and Van Parys Judgments: Rules, Exceptions and the Law’ (2005) 32 Legal Issues of Economic Integration 460 at p. 467.



In this respect, certain considerations relating to EU enlargement and the locus standi of private parties in annulment proceedings under Article 230 EC are raised. After UPA,62 the restrictive interpretation of ‘ individual concern ’ suggests that the number of private applicants who satisfy the standing requirements for the challenge of Community legislation under WTO law will remain small. The impact of preliminary references on validity under Article 234 EC will be more significant as it can be expected that national courts will seek guidance on questions of WTO law, the specialized and complex nature of which is unlikely to encourage them to tackle them themselves. Direct challenges brought by Member States, on the other hand, are likely to increase because of EU enlargement. Under the Treaty of Nice rules on qualified majority voting, as many as 12 Member States could form a nonblocking minority against WTO-related acts adopted ultimately by the Council.63 This makes increased litigation probable with national governments, under pressure from powerful economic and political lobbies and NGOs, forced to bring annulment proceedings against potentially WTO-inconsistent acts which they failed to block in the Council. The freedom of political institutions should also be examined within the WTO context. There is no better example to illustrate the considerations at issue than the EC–Bananas dispute. This dispute, concerning the inconsistency with WTO rules of the Common Market Organization (CMO) in bananas, ran in parallel in the WTO dispute settlement system and the Court of Justice.64 The CMO in bananas, as amended, contained a complex system of quotas and import licensing procedures favouring traditional and ACP importers of bananas and has been long viewed as an important development tool. It was unavoidable that, but for the lack of direct effect, one of the several challenges brought against the CMO in Community courts would have succeeded and the Community would have been divested of the possibility of applying for a preferential regime on imports from ACP countries, a practice which is central to the Community’s development policy. Assuming that the Community purports to apply WTO-consistent policies, this is not, in principle, undesirable. However, at the same time, the Community would have forfeited the facility to obtain a waiver, a route it actually followed in EC–Bananas.65 Giving away such an important WTO-legitimate option is like shooting oneself in the foot. Unlike the argument in Kupferberg where the Court dismissed the importance of the safeguard clauses in the FTA with Portugal as being too

62 Case C-50/00 P Union de Pequen˜os Agricultores v. Council [2002] ECR I-6677 at para. 44 63 As amended by Article 12 Treaty of Accession. The number will be reduced to ten under the Constitutional Treaty. See Article I-25(1) Treaty establishing a Constitution for Europe. 64 In the WTO, WT/DS27 European Communities – Regime for the Importation, Sale and Distribution of Bananas; in the Community, numerous annulment, preliminary rulings and Community liability cases, supra. 65 WT/MIN(01)/15 European Communities – The EC-ACP Partnership Agreement, Decision of 14 November 2001.

The European Union and WTO law 55

specific,66 Article IX :3 WTO Agreement represents a possibility for waiving an obligation under the covered agreements, broadly so conceived. Had the Court granted direct effect to WTO law, the political institutions’ arms would have been tied and their authority to negotiate with other WTO Members would have been frustrated. The side effect this development would have would be to turn the Court of Justice, rather than the WTO dispute settlement system, into the principal forum where Community legislation would be challenged. This represents a disruption not only of the institutional balance between the institutions of the Community, but also of the institutional framework established by the WTO.67 Direct effect of Panel and Appellate Body rulings The conclusion to the previous section indicates that the Court of Justice did not wish to usurp the function of the WTO bodies to interpret WTO law and review the consistency of Community acts against its provisions. Would the Court of Justice be prepared to change its position were it faced with a case where a Panel and/or the Appellate Body ruled that a Community measure violated WTO law and consider itself bound by such ruling ? The question accordingly is whether, from a WTO perspective, the Panel and Appellate Body rulings create an obligation to perform in a traditional international law sense and whether domestic courts are bound by the rulings.68 Before answering this question, it would be useful to analyse the nature and legal force of DSB recommendations from a WTO perspective. In Japan–Taxes on Alcoholic Beverages, at the examination of the bindingness of adopted GATT panel reports, the Appellate Body held : Adopted panel reports are an important part of the GATT acquis. They are often considered by subsequent panels. They create legitimate expectations among WTO Members, and, therefore, should be taken into account where they are relevant to any dispute. However, they are not binding, except with respect to resolving the particular dispute between the parties to that dispute.69

At this stage, it would be useful to briefly note how a dispute is resolved between WTO Members under the Dispute Settlement Understanding (DSU). At the finding of a violation, the Panel or the Appellate Body shall ‘ recommend that the Member concerned bring the measure into conformity with the agreement ’. In the proper interpretation of the DSU terms, compliance with the WTO obligation may be

66 Case 104/81 Hauptzollamt Meinz v. Kupferberg [1982] ECR 3641 at para. 21. 67 Confirmed by the Court in the recent Case C-377/02 Le´on Van Parys NV v. Belgisch Interventie- en Restitutiebureau (BIRB)[2005] ECR I-1465 at para. 53. See, Antoniadis, supra note 61. 68 John H. Jackson, ‘The WTO Dispute Settlement Understanding – Misunderstandings on the Nature of Legal Obligation ’ in J. Cameron and K. Campbell (eds.), Dispute Resolution in the World Trade Organisation (Cameron May, London, 1998), pp. 69–74 at p. 74. 69 Appellate Body Report in WT/DS8/AB/R Japan – Taxes on Alcoholic Beverages (complaint brought by the EC), at p. 13.



seen as unequivocal, compensation and suspension of concessions being only temporary alternatives.70 Article 22.1 DSU provides that ‘ neither compensation nor the suspension of concessions or other obligation is preferred to full implementation of a recommendation to bring a measure into conformity with the covered agreements ’. In addition to its recommendations, a Panel or the Appellate Body may suggest ways in which the Member concerned could implement the recommendations.71 Panels have been reluctant to recommend specific ways for implementation, thereby showing deference to the national margin of manoeuvre at the implementation of their recommendations72 and illustrating the hypothesis that compliance may have several variants.73 Commentators, however, support bindingness beyond the traditional sense and militate in favour of the full effect of those rulings in the Community legal order.74 Jackson argues that despite their linguistic shortcomings there are several provisions in the DSU that point towards the direction of bindingness.75 He does however acknowledge that the US Courts will not treat them as such, but indicates that these rulings may affect the US jurisprudence as well as those of other WTO Members. Regarding the Community legal order, the question of the effect of WTO Panel and Appellate Body rulings was presented to the Court on several occasions. In the course of the EC–Bananas litigation, several actions were brought seeking to annul the CMO in bananas or claim damages invoking the relevant DSB recommendations. The initial approach by the Court of First Instance (CFI) avoided the issue by narrowing the scope of the Panel and Appellate Body’s findings and declining the review of the entire tariff quota system established by the CMO in bananas.76 The Court of Justice later recognized the link between direct effect of WTO provisions and the DSB recommendations, by explaining that DSB decisions establishing the inconsistency of Community law with the GATT could only have been taken into consideration should the Court have found the GATT to have direct 70 Jackson, supra note 68. 71 Article 19.1 DSU. 72 Allan Rosas, ‘Implementation and Enforcement of WTO Dispute Settlement Findings: An EU Perspective ’ (2001) 4 Journal of International Economic Law 131 at p. 134. See also, Panel Report in WT/ DS152/R United States – Sections 301–310 of the Trade Act of 1974 at para. 7.102. 73 D. Palmeter and P. C. Mavroidis, Dispute Settlement in the World Trade Organization, Practice and Procedure, 2nd edition (CUP, 2004), at pp. 295–300. Eeckhout, supra note 57 at p. 93; Rosas supra note 72 at pp. 135–136. 74 Geert A. Zonnekeyn, ‘The Status of Adopted Panel and Appellate Body Reports in the European Court of Justice and the European Court of First Instance – The Banana Experience’ (2000) 34 Journal of World Trade 93; Geert A. Zonnekeyn, ‘The Bed Linen Case and its Aftermath: Some Comments on the European Community’s ‘‘World Trade Organization Enabling Regulation’’’ (2002) 36 Journal of World Trade 993 ; N. Lavranos, Decisions of International Organizations in the European and Domestic Legal Orders of Selected EU Member States (Europea Law Publishing, 2004). 75 Jackson, supra note 68. McNelis arrives at the same conclusion on the basis of the principle of good faith. See, Natalie McNelis, ‘ What Obligations Are Created by World Trade Organization Dispute Settlement Reports?’ (2003) 37 Journal of World Trade 647 at pp. 657–659. 76 Case T-254/97 Fruchthandelsgesellschaft mbH Chemnitz v. Commission [1999] ECR II-2743 at para. 26.

The European Union and WTO law 57

effect.77 In principle, therefore, the effect of DSB recommendations is inextricably linked with the more general question of direct effect of the WTO Agreement. As a result, direct effect of DSB recommendations should be excluded. The question which logically follows is whether, where the Community adopts a legislative measure in response to adverse DSB recommendations, it could be construed as doing so in order to implement a particular WTO obligation, thereby triggering the so-called Nakajima exception, according to which the said legislation can be reviewed against the WTO provisions it is intended to implement.78 The question was initially broached by the CFI in three more ‘ bananas ’ cases, Cordis, Bocchi, and T. Port, in which the applicants requested from the CFI to examine the provisions of Regulation 2362/98 in the light of the WTO Agreement and, in particular, those provisions that the previous Regulation had been found to violate. The CFI declined the application of the implementation exception, declaring that : neither the reports of the WTO Panel of 22 May 1997 nor the report of the WTO Standing Appellate Body of 9 September 1997 which was adopted by the Dispute Settlement Body on 25 September 1997 included any special obligations which the Commission ‘intended to implement ’, within the meaning of the case law, in Regulation No 2632/98.79

The judgments attracted criticism as a missed opportunity to extend the implementation principle to its proper scope.80 It should be counter-argued however that this represents the logical consequence of the denial of direct effect to WTO law. The Court revisited the issue in Biret,81 a case concerning the claim for noncontractual liability of the Community for damages suffered by Biret because of the Community’s import ban on hormone-treated beef. The DSB had established the inconsistency of the ban with Articles 3.3 and 5.1 of the SPS Agreement82 and the Court seemed to respond to the calls from academics and indicate its readiness to trigger the Nakajima exception. In this judgment, the Court confirmed that the WTO rules are not among those rules in the light of which the Court is to review the legality of measures adopted by the Community institutions subject to a temporal limitation.83 Since the Community had stated that it intended to comply 77 Case C-104/97P Atlanta v. European Community [1999] ECR I-6983 at paras. 19–20. See also, Case C-307/99 OGT Fruchthandelsgesellschaft v. Hauptzollamt Hamburg-St. Annen [2001] ECR I-3159. 78 See below, Section 3. 79 Case T-18/99 Cordis Obst und Gemu¨se Grosshandel v. Commission [2001] ECR II-913 at para. 59; Case T-30/99 Bocchi Food Trade International v. Commission [2001] ECR.II-943 at para 64; Case T-52/99 T. Port v. Commission [2001] ECR II-981 at para. 59. 80 Eeckhout, supra note 57 at p. 107. See also the criticism launched by Snyder supra note 1 at p. 338. 81 Cases C-93/02 and 94/02 Biret International and Etablissements Biret et Sie [2003] ECR I-10497. 82 WT/DS26 European Communities – Measures concerning meat and meat products (hormones) (complaint brought by US) ; WT/DS48 European Communities – Measures concerning meat and meat products (hormones) (complaint brought by Canada). 83 Biret at para. 52.



with its WTO obligations but that it needed reasonable time to do so and, in fact, was granted a period of 15 months for that purpose under Article 21.3 DSU, the Community Courts could not review the legality of the Community measures in question without rendering ineffective the grant of a reasonable period for compliance with the DSB recommendations, as provided by the WTO Agreements.84 This judgment was justifiably perceived as having the potential of introducing the direct effect of DSB recommendations subject to the deadline for their implementation granted under Article 21.3 DSU having elapsed.85 The implication was that, after the deadline, the legality of Community measures would be subject to challenge. In fact, the CFI in Chiquita,86 following the signal given by the Court in Biret, accepted, in principle, the assumption that, in amending the CMO in bananas the Community intended to implement the substantive GATT and GATS obligations it was found to violate.87 The Van Parys judgment, delivered by the Court of Justice one month after the CFI made its pronouncements in Chiquita, put an end to theoretical discussions on the scope of the Nakajima exception that would afford direct effect to DSB recommendations in the Community legal order.88 The question presented to the Court in Van Parys was whether the WTO agreements give Community nationals a right to rely on WTO agreements in legal proceedings challenging the validity of Community legislation, where the DSB has held that both that legislation and subsequent legislation adopted by the Community in order, inter alia, to comply with the relevant WTO rules, are incompatible with those rules.89 The Court answered the question in the negative. Faithful to its analysis in Portuguese Textiles, analyzed above, it laid down the three basic reasons denying direct effect : first, the freedom of the institutions to reach a mutually acceptable solution ;90 second, the principle of reciprocity ;91 third, the nature of the rules enshrined in the WTO dispute settlement system. The third reason is of particular importance for its selfstanding merit and because it informs the first two and provides an additional layer of understanding of the principles analysed above.92 Using the paradigm of the Community’s bananas regime and the forthcoming resolution of the dispute to the satisfaction of the Community, the Court undertook an overview of the dispute and implicitly congratulated the legislature for managing to reconcile the 84 Biret at paras. 61–62. 85 Jochem Wiers, ‘ One Day, You’re Gonna Pay: The European Court of Justice in Biret’ (2004) 31 Legal Issues of Economic Integration 143 at p. 147. 86 Case T-19/01 Chiquita Brands International, Inc., Chiquita Banana Co. BV and Chiquita Italia, SpA v. Commission of the European Communities [2005] ECR II-315. 87 Chiquita at para. 127. For a fuller critique see Antoniadis, supra note 61. 88 Case C-377/02 Le´on Van Parys NV v. Belgisch Interventie- en Restitutiebureau (BIRB) [2005] ECR I-1465. 89 Van Parys at para. 38. 90 Van Parys at paras. 48 and 51. 91 Van Parys at para. 53. 92 See above.

The European Union and WTO law 59

requirements of the common agricultural policy with the Community’s obligations towards the ACP states by taking advantage of the room for manoeuvre provided by the WTO legal system.93 The freedom of the political institutions has been exercised ‘ in conformity with those rules ’ according to the Court.94 At the same time, disregard to the principle of reciprocity ‘ would risk introducing an anomaly in the application of the WTO rules ’.95 Accordingly, the freedom of the political institutions and reciprocity are significant, not only from a Community viewpoint but also as principles leading to the correct application of WTO rules and, in particular, those enshrined in the DSU.96 The nature of the DSU also makes it necessary to overturn the Biret judgment in that, even after the lapse of the deadline for compliance, there is still room for negotiation in conformity with the DSU provisions.97 In Van Parys, the Court did not go into the examination of the Nakajima exception in detail. It simply held that the measures taken by the Community institutions cannot be interpreted as measures intended to ensure the enforcement of a particular obligation within the context of the WTO.98 Implementation/ transposition of the WTO Agreements as such should be resorted to after close examination of all legal, political, and economic considerations. As such, the implementation exception should remain narrow. The finding by a Panel or the Appellate Body that a given Community act is inconsistent with the WTO Agreement creates an obligation which is little different to the obligations of the WTO Members under the provisions of those agreements and, in particular, Article XVI :4 of the WTO Agreement.99 This general obligation, if taken literally, would mean that all measures taken by WTO Members falling within the scope of the WTO and covered agreements should be treated as aiming at the implementation of those agreements. While this idea might sound attractive to some, such a finding would overrule the Court’s general position concerning direct effect of the WTO Agreement, and, should the Court wish to maintain consistency in this respect, it must be resisted. In lieu of a coda to this analysis, it must be pointed out that the Panel and Appellate Body interpretations have developed an increasing significance at the resolution of disputes in Community Courts. The Court of Justice, as is habitually the case in relation to rulings by international courts or tribunals,100 resorts to the 93 Van Parys at paras. 49–50. 94 Van Parys at para. 51. 95 Van Parys at para. 53 96 Antoniadis, supra note 61 at p. 467. 97 Van Parys at para. 51. 98 Van Parys at para. 52. 99 ‘Each Member shall ensure the conformity of its laws, regulations, and administrative procedures with its obligations as provided in the annexed Agreements. ’ 100 Allan Rosas, ‘With a Little Help from My Friends: International Case law as a Source of Reference for the EU Courts’ (2005) 5 The Global Community Yearbook of International Law and Jurisprudence 203.



interpretations granted by the WTO bodies in the exercise of its general duty to interpret Community law as far as possible consistently with international law.101 Non-contractual liability of the Community Several cases have been brought to the Community Courts by affected banana traders requesting compensation from the Community for damages suffered. For the Community to incur liability, a number of conditions must be met. There must be an unlawful act imputable to the Community, damage to the applicant, and the existence of a causal link between the unlawful act and the damage suffered.102 It is settled case law that, for the unlawful conduct condition, a sufficiently serious breach of a rule of law intended to confer rights on individuals must be established.103 It is therefore obvious that the question of Community liability is inextricably linked with the issue of direct effect. In the absence of direct effect, the Community courts have consistently denied any right to damages,104 as for the Community to incur liability all conditions must be met.105 It has been suggested, making the analogy with Francovich, that direct effect is not necessary for the award of damages.106 This argument can be sourced to the definition of direct effect and the attempt to treat direct effect and the creation of rights for individuals as non-synonymous concepts,107 thereby overlooking the defining characteristic of directives which are, in principle, capable of conferring rights on individuals. Those rights can simply not be enforced against other individuals in national courts because of the absence of horizontal effect of directives.108 In Francovich, because the directives at issue were capable of granting rights to individuals and the remaining conditions were fulfilled, damages were awarded. By contrast, the Court’s decision to deny liability to the affected traders for damages suffered in violation of WTO law makes perfect sense. The

101 Case C-245/02 Anheuser-Busch Inc. v. Bude˘jovicky` Budvar, na´rodnı´ podnik [2004] ECR I-10989. For a fuller analysis see Section 3. 102 Case 5/71 Aktien-Zuckerfabrik Scho¨ppenstedt v. Council [1971] ECR 975. 103 Case C-352/98 P Bergaderm and Goupil v. Commission [2000] ECR I-5291 at para. 42. 104 Case T-18/99 Cordis Obst und Gemu¨se Grosshandel v. Commission [2001] ECR II-913; Case T-30/99 Bocchi Food Trade International v. Commission [2001] ECR.II-943; Case T-52/99 T.Port v. Commission [2001] ECR II-981; Case C-104/97P Atlanta AG and others v. Commission and Council [1999] ECR I-6983; Cases C-93/02 and 94/02 Biret International and Etablissements Biret et Sie [2003] ECR I-10497. 105 Case C-146/91 KYDEP v. Council and Commission [1994] ECR I-4199; Joined Cases T-198/95, T-171/96, T-230/97, T-174/98 and T-225/99 Comafrica and Dole Fresh Fruit Europe [2001] ECR II-1975 at para. 134. 106 Birgit Schoißwohl, ‘ The ECJ’s Atlanta Judgment: Establishing a Principle of Non-Liability? ’ in F. Breuss, S. Griller and E. Vranes (eds.), The Banana Dispute: An Economic and Legal Analysis (Research Institute for European Affairs, Springer, 2003), p. 309. 107 Jochem Wiers, ‘ One Day, You’re Gonna Pay: The European Court of Justice in Biret ’ (2004) 31 Legal Issues of Economic Integration 143 at p. 148. 108 See Case C-6/90 Francovich and others v. Italian Republic [1991] ECR I-5357 at para. 27, Case 152/84 Marshall v. Southampton and SW Hampshire Area Health Authority [1986] ECR 723 at para. 48 and their progeny.

The European Union and WTO law 61

Court’s analysis on direct effect of WTO law makes clear that the WTO Agreement is not, in principle, capable of granting any rights to individuals.109 Therefore, individuals cannot have an entitlement to a certain level of tariff, quota, or any specific treatment by any WTO Member110 and such treatment cannot cause damage to them. The established position was put in doubt after Biret. In this well-known judgment, the Court opined that a review of legality of Community law cannot be undertaken in the context of an action for damages under Article 235 EC while the deadline for compliance with the DSB recommendations under Article 21.3 had not yet expired.111 The unavailability of a damages claim against the Community in Community Courts was extended by the CFI in Chiquita until the end of the dispute in the WTO, particularly so, when the Community was subject to retaliation under Article 22 DSU.112 Linked with the preceding analysis on the effect of Panel and Appellate Body rulings, these two judgments read together meant – Chiquita more explicitly so – that when the Community adopted legislation so as to comply with adverse DSB recommendations, the affected private parties could claim compensation for damages suffered as a result of the measure at issue subject to the proceeding under WTO dispute settlement having been terminated. This proposition is based on the theoretical assumption that the implementation exception can apply when the WTO has ruled on the matter. Following Van Parys, which explained that when the Community amends its legislation in order to comply with adverse DSB recommendations it cannot be presumed to intend to implement any particular obligation under the WTO Agreements, the foundation of this reasoning is overturned.113 Indeed, since the implementation exception does not apply, the subsequent construction of the Community Courts in Biret and Chiquita cannot stand. The ruling in Van Parys lessened the anxiety over the long-awaited judgments by the CFI in FIAMM.114 These cases concerned traders who had suffered damages not as a direct result of WTO-incompatible Community legislation but because they were subject to retaliation by another WTO Member. Are those traders entitled to any compensation ? The CFI, based on the previous case law, held that the Community cannot, in principle, be held liable by reason of any infringement of 109 For a thoughtful analysis see Armin von Bogdandy, ‘Legal Effects of World Trade Organization Decisions Within European Union Law: A Contribution to the Theory of the Legal Acts of International Organizations and the Action for Damages Under Article 288(2) EC’ (2005) 39 Journal of World Trade 45. 110 Rosas, supra note 72 at p. 140. Pieter Jan Kuijper and Marco Bronckers, ‘ WTO Law in the European Court of Justice’ (2005) 42 Common Market Law Review 1313 at p. 1332. Also, Cordis at para. 51. 111 Biret at para. 62. 112 Chiquita at para. 166. 113 Van Parys at para. 41. 114 Joined Cases T-69/00 T-301/00, T-320/00, T-383/00 and T-135/01 FIAMM and FIAMM Technologies and others v. Council and Commission, judgment of 14 December 2005, not yet reported.



WTO rules by the Community institutions.115 This could only be the case if the Community intended to implement a particular obligation assumed in the context of the WTO or where the Community measure expressly refers to specific provisions of the WTO.116 Mirroring the substance of Van Parys, yet, not citing it, the CFI held that by amending legislation found to be incompatible with WTO rules, the Community did not intend to implement specific obligations arising from those rules.117 Neither, did the relevant Community legislation make express reference to specific WTO rules.118 The seemingly insurmountable insulation of Community law against any challenge under WTO in Community Courts would make the analysis of the other conditions of liability, namely the damage and the causal link between the act of the institutions and damage suffered, redundant. Yet, the CFI in FIAMM was faced with a claim to examine the non-contractual liability of the Community for the lawful conduct of the Community institutions drawing on national laws of the Member States. In fact, the CFI found that national laws enable individuals to obtain compensation for damages suffered even in the absence of unlawful conduct by the perpetrator.119 Following Dorsch Consult, the CFI held that the Community could incur liability in the absence of unlawful conduct if actual damage has been sustained, the causal link between that damage and the conduct of Community institutions has been established, and that damage was of unusual and special nature.120 The claim made by the applicants in FIAMM presented a prime opportunity to the CFI to examine the issue of damages and causal link, which, owing to the hurdle of direct effect, was hardly broached in the Court’s case law in the area of WTO law.121 Starting from the nature and extent of damages, the CFI simply stated that the applicants must have necessarily suffered commercial damage by reason of the incontestable rise in the price of their products, resulting from the imposition of an additional duty of 96.5 %.122 Straightforward as it is, this statement is not convincing. Clearly, damage in this context includes the damage actually suffered plus any lost profits.123 The emphasis on actually suffered runs counter to the presumption ‘ must have necessarily suffered ’.124 Accordingly, in this context, damages should not be calculated on the basis of the amount by which the tariffs at the imports of Community products have increased by virtue of the suspension of concessions applied by the complaining WTO Member, but the actual effect the 115 FIAMM at para. 113. 116 FIAMM at para. 114. 117 FIAMM at para. 137. 118 FIAMM at para. 144. 119 FIAMM at para. 159. 120 FIAMM at para. 160. 121 With the exception of the manifest lack of causality in Biret at paras. 63–64. 122 FIAMM at para. 168. 123 Cases 5, 7, 13-24/66 Kampffmeyer v. Commission [1967] ECR 245. 124 FIAMM at para. 168.

The European Union and WTO law 63

raise of tariffs has had upon the competitive position of the affected traders. In this respect, the CFI’s presumption will be difficult to apply in cases of traders like Louis Vuitton whose sales of leather handbags recorded a surge in the US market despite the sanctions.125 Going on to the issue of causal link, the CFI, referring to previous case law, established that causality exists when there is a sufficiently direct causal nexus between the conduct of the Community institutions and the damage suffered.126 While the CFI recognized the option available to the United States to settle the dispute, it nonetheless explained that ‘ the withdrawal of concessions in relation to the Community results objectively, in accordance with the normal foreseeable operation of the WTO dispute settlement system accepted by the Community’.127 The unilateral act by the United States to increase the customs duties on imports of batteries does not, in the CFI’s view, break the causal link. In the CFI’s words, the damage suffered by the applicants ‘ must be regarded as the immediate cause ’ of the Community conduct.128 Two objections must be raised against the CFI’s reasoning in this respect : first, it provided a very broad interpretation of the concept of the ‘ direct causal link’. In the light of the large number of steps between the Community breach of WTO law and the damage suffered by an individual trader as a result of a WTO Member’s suspension of concessions,129 it ought to be questioned how immediate and direct such a causal relationship is. Extrapolating the conditio sine qua non theoretical foundation of causality to other areas of Community law is likely to change the landscape of Community liability as at the moment it sits uncomfortably with it.130 In fact, it is questionable how the CFI considers the causal link broken in a case in which the Community institutions enjoy no discretion at the adoption of the harmful act,131 while the opposite is the case for discretionary acts of other WTO Members.132 Second, the CFI’s analysis seems to disregard its previously vigorously advocated thesis on the great flexibility of the WTO DSU provisions. Indeed, there is a logical error here as, if the flexibility of the DSU is as great as the CFI has repeatedly stated, the relationship between the conduct of the Community institutions and the damage suffered by the individual traders could not be

125 While this raise has been attributed to either smart management, see, for instance, Ashok Som, ‘Personal touch that built an empire of style and luxury’ available at 3-Personal-touch%20.pdf, at p. 8 or American consumers’ increasing demand for luxury products, see, for instance,, no study assesses the competitive position of Louis Vuitton in the absence of sanctions. 126 FIAMM at para. 178 and the case law mentioned there. 127 FIAMM at para. 183. 128 FIAMM at para. 185. 129 Article 3.7 DSU. Rosas, supra note 72 at p. 140. 130 See, inter alia T-184/95 Dorsch Consult Ingenieurgesellschaft mbH v. Council of the European Union and Commission of the European Communities [1998] ECR II-667 at paras. 70–74. 131 Dorsch Consult at para. 72. 132 FIAMM at para. 184. See also, Kuijper supra note 110 at p.1337 who anticipates the problem.



characterized as either ‘immediate ’ or ‘ direct ’. With regard to the suspension of concessions, it should be pointed out here that in the light of recent disputes, even in the presence of DSB authorization, WTO Members have been reluctant to apply retaliatory measures.133 Therefore, the imposition of additional duties on products originating in the EC can hardly qualify as ‘ the normal foreseeable operation of the WTO dispute settlement system ’.134 Damage will be of an unusual nature, held the CFI in FIAMM, when it exceeds the limits of the economic risks inherent in operating the sector concerned and of special nature when it affects a particular circle of economic operators in a disproportionate manner by comparison with other operators.135 In light of the nature of international trade, the CFI had little difficulty dismissing the claim brought by the applicants that they suffered damage of an unusual nature and exercised judicial economy on the question of special nature.136 Whilst a full critique of the FIAMM judgment escapes the confines of this contribution, it must be stated here that it sends mixed messages. The CFI correctly followed the established case law ruling out Community liability because of the lack of direct effect. It is submitted, however, that it erred in embarking on the analysis of Community liability for lawful conduct. Even if such a principle can find sufficient support in the cited judgments and national law,137 its operation could not be seen as an alternative remedy when the unlawfulness of the acts of the Community institutions cannot be established because of the absence of direct effect. In addition, ‘ actual damage’ and ‘direct causal link ’ do not warrant such broad interpretations. It could be assumed that the CFI’s analysis is revealing of the collective mindset of its members and potential disagreements regarding the continued Community Courts’ reactive approach. In sum, this approach denies liability to private traders for damages suffered as a result of unlawful or lawful Community acts in the WTO domain. As a precursor to the discussion on the coactive approach, it should be pointed out that the discussion concerning the conditions for Community liability maintains its significance. That is because Community liability may apply in these cases. As will be demonstrated in detail below, the Community applies a communitarized version of WTO law in the context of the coactive approach. It could be argued in this respect that communitarization of, say, the WTO AntiDumping Agreement vests this Agreement with the constitutional qualities of

133 WT/DS108 United States – Tax Treatment for Foreign Sales Corporations; WT/DS136 United States – Anti-Dumping Act of 1916; WT/DS222 Canada – Export Credits and Loan Guarantees for Regional Aircraft. 134 FIAMM at para. 183. 135 FIAMM at para. 202. 136 FIAMM at para. 212. 137 FIAMM at paras. 158–160. See also, T-184/95 Dorsch Consult Ingenieurgesellschaft mbH v. Council of the European Union and Commission of the European Communities [1998] ECR II-667 at para. 77.

The European Union and WTO law 65

Community law, including direct effect. By contrast, it could also be argued that since WTO law is not capable of conferring rights on individuals, the content of the exception created by the Court should only be construed to include the incidental review of legality of the Community measures under WTO law in annulment proceedings, but not to establish a claim for Community liability.138 It must be pointed out however that the Court is not likely to deal with the issue because of the dual avenue available to traders to enforce their rights in Community Courts and the WTO Dispute Settlement and the measures taken by the institutions to minimize such an eventuality.139

The political institutions In the reactive approach, the political institutions showed resistance towards the full effect of WTO law. The most characteristic example of a reactive approach by the Council is the very inclusion in the preamble to its decision concluding the WTO Agreement of a clause stating that ‘ Whereas, by its nature, the Agreement establishing the World Trade Organization, including the Annexes thereto, is not susceptible to being directly invoked in Community or Member State courts ’.140 The Council, in this instance, was acting within its powers as duly recognized by the Court141 and this statement should be considered of cardinal importance, despite the Opinion to the contrary by Advocate General Saggio in Portuguese Textiles.142 In addition, the political institutions have adopted a reactive approach in specific instances, primarily concerning those cases where they chose to maintain measures reflecting fundamental policy choices despite adverse Panel and Appellate Body rulings. This stance has been particularly important in the foundational years of the WTO and tested the limits of the system. At the same time, it tested the Community’s own limits as an idiosyncratic actor within this system.143

3. Coactive approach

The Court of Justice In the reactive approach towards WTO law, the Court of Justice denied the enforcement of WTO law in national and Community Courts and the review of 138 Contra, Wiers, supra note 85 at p. 148, who makes a distinction between the conferral of rights on individuals and direct effect. 139 See below, Section 3. Also, Antonis Antoniadis, ‘The Participation of the European Community in the World Trade Organisation: An External Look at European Union Constitution-Building’ in T. Tridimas and P. Nebbia (eds.), EU Law for the 21st Century: Rethinking the New Legal Order, Vol. I (Hart Publishing, 2004), p. 321 at pp. 340–343. 140 Council Decision 94/800 of 22 December 1994 concerning the conclusion on behalf of the European Community, as regards matters within its competence, of the agreements reached in the Uruguay Round multilateral negotiations (1986–1994), O.J. L 336, 23/12/1994, p. 1. 141 Kupferberg at para. 17. 142 AG Saggio Opinion in Case C-149/96 Portugal v. Council [1999] ECR I-8395 at para. 20. 143 Antoniadis, supra note 139 at pp. 343–344.



secondary Community law against its provisions. However, as already implied in the course of the analysis of the Portuguese Textiles, there are clearly defined exceptions recognized by the Court. These exceptions can be classified as : the legality standard, the transposition/implementation, the clear reference, and the consistent interpretation.144 The exceptions partly reflect the Court’s response to action by the Community’s political institutions showing inclination to include the application of or reference to WTO law in their activities, and partly the realization that, in some cases, judicial enforcement of WTO rules in the Community and national Courts serves the better application of these rules and furthers the WTO objectives. The term ‘ coactive ’ in this sense does not simply represent a notable deviation from the lack of direct effect doctrine, it also signifies the intention of the institutions to use WTO law in their activities alongside Community law. Legality standard Under Article 300(7) EC, the conclusion of the WTO Agreement by the Community bears the consequence that both the Community and its Member States must observe its provisions. The legality standard exception aims to enable the Community to hold the Member States to their commitments. As stated in Kupferberg : In ensuring respect for commitments arising from an agreement concluded by the Community institutions the Member States fulfil an obligation not only in relation to the non-member country concerned but also and above all in relation to the Community which has assumed responsibility for the due performance of the agreement. That is why the provisions of such an agreement, as the Court has already stated in its judgment of 30 April 1974 in case 181/73 Haegeman (1974) ECR 449, form an integral part of the Community legal system.145

The Court underlined here the functional considerations present in the fulfilment of the Community’s international obligations and the aim to avoid incurring international responsibility.146 Premised upon the need to establish a unified front, which should not be undermined by Member States’ breach of the Community’s international commitments, and supported by the principle of cooperation and Article 300(7) EC, the Community can coerce Member States to fulfil their obligations under international agreements. There is no issue of direct effect here, simply an unconditional obligation incumbent on Member States every time the Community concludes an international agreement. It flows from the unique position of the Community assuming its own obligations on the international plane, a consequence of the fact that it has been granted with legal personality to

144 I prefer the terms ‘transposition ’ and ‘clear reference’ adopted by Snyder, supra note 1 at p. 342 over the ‘indirect effect ’ terminology proposed by Eeckhout, supra note 27 at p. 40. 145 Kupferberg at para. 13. 146 Klabbers, supra note 4 at p. 281.

The European Union and WTO law 67

be able to invoke the mechanisms available at Community level in order to ensure that the Member States observe their obligations under the Treaties. Accordingly, when the Commission considers that Member States violate their obligations under the WTO Agreement, it may invoke enforcement proceedings in order to bring the recalcitrant Member States back to order. In the ordinary interpretation of Article 226 EC, Member States’ failure to fulfil their obligations under the Treaty includes also obligations assumed by the Community in the WTO. WTO rules become the Community legal standard by which Member States must abide. This obligation stems from the nature of the Community as a Customs Union under the GATT, which demands that each Contracting Party shall take such reasonable measures as may be available to it to ensure observance of the provisions of the GATT Agreement by the regional and local governments and authorities within its territories.147 Given that the Community was perceived as a single Contracting Party in the GATT,148 the whole meaning of regional integration arrangements within the international trading system would be frustrated should the Community be unable to use the means available to it in order to enforce compliance with the GATT provisions.149 The most important case under the legality standard exception is the International Dairy Agreement (IDA).150 In that case, the Commission requested the Court to declare that, by authorizing the importation of dairy products at a customs value lower than the minimum price provided by the IDA, an agreement annexed to the Tokyo Round of Multilateral trade negotiations conducted under the GATT, Germany failed to fulfil its obligations under the Treaty. The Court obviated the examination of whether the Commission had the right to bring proceedings against a Member State under the GATT, and, on the facts of the case, had little difficulty in concluding that Germany was in violation of the Annexes to the IDA. In order to arrive at this conclusion the Court had to overcome difficult hurdles. The most important was the claim by Germany, taken up by Advocate General Tesauro in his Opinion,151 that the Community legislation on inward processing relief was in violation of the IDA too. The Court asserted the duty of consistent interpretation and held that the relevant Council Regulation could be interpreted in conformity with the IDA.152 This judgment has attracted criticism as it seemingly promotes inconsistency and, possibly, also an imbalance between the rights and obligations of the Member States. It was argued that allowing the Commission to challenge a Member State measure for violation of the GATT, while at the same time denying a Member State the corollary right of challenge in annulment proceedings, ‘ is on balance not 147 Article XXIV:12 GATT. 148 Article XXIV:1 GATT. 149 In this case Article 226 EC Treaty. 150 Case C-61/94 Commission v. Germany (International Dairy Agreement) [1996] ECR I-3989. 151 AG Tesauro’s Opinion in International Dairy Agreement at para. 24. 152 International Dairy Agreement at para. 57.



satisfactory, if one considers that the two situations are comparable ’.153 Clearly, the situations are not comparable. Had the Commission or any other institution been permitted to apply for the annulment of Community legislation, while the Member States had not, it would have been a comparable yet untenable position. However, the different treatment of qualitatively different proceedings (annulment in Portuguese Textiles and enforcement in IDA) can be reconciled. First, what is at stake in annulment proceedings is the fulfilment of an EC obligation vis-a`-vis the GATT/WTO whilst in enforcement proceedings the fulfilment of an obligation of a Member State vis-a`-vis the EC. As a result, in both cases, the Community upholds EC law over international law or national law. This seems to remedy the alleged inconsistency154 and raises the question of the supremacy of the Community legal order, which will be dealt with below.155 In addition, the Court’s statement in Haegeman, regarding international law forming an integral part of the Community legal order,156 is vindicated in that WTO law will be applied in enforcement proceedings.157 More recently, the Court condemned Ireland for failing to adhere to the 1971 Berne Convention for the Protection of Literary and Artistic Works.158 Unlike IDA, which concerned compliance with an agreement concluded by the Community, it is not Community membership that provides the cause of action in this case, since the Community is not a party to the Berne Convention. The Berne Convention is, in principle, binding on the Community indirectly by virtue of Article 9 of the TRIPs Agreement in accordance with which the Community, as a WTO Member bound by all covered agreements, is obliged to afford the level of copyright protection enshrined in the Convention to all other WTO Members. Curiously, the TRIPS Agreement is not mentioned at all but, instead, Article 5 of Protocol 28 to the EEA Agreement, which obliges the parties to the EEA to adhere to the Berne Convention.159 Transposition/Implementation In the following aspect of the coactive approach, the Community is required to act in conformity with WTO law insofar as it has adopted measures intended to implement certain of its provisions. This is the so-called ‘transposition’/‘ implementation ’ exception in accordance with which individuals may invoke WTO law in order to challenge incompatible Community 153 Bourgeois, ‘ The European Court of Justice and the WTO’ in Weiler, J. H. H. (eds.), The EU, The WTO and the NAFTA (Oxford University Press, 2001), at p. 112. See also, AG Tesauro’s Opinion in International Dairy Agreement at paras. 23–24. More recently, Bronckers, supra note 110 at pp. 1348–1349. 154 Ibid. at pp. 112–113. 155 See Section 5. 156 Case 181/73 Haegeman v. Belgium [1973] ECR 449 at paras. 4 and 5. 157 See also, Rosas, supra note 100 at p. 218. 158 Case C-13/00 Commission v. Ireland [2002] ECR I-2943. 159 Ibid., at para. 20.

The European Union and WTO law 69

acts.160 While, in principle, the transposition exception may apply generally in all areas of Community legislation falling within the scope of WTO law, it must be pointed out that its application has been limited to the field of Anti-Dumping.161 The foundational case for this exception is Nakajima.162 In that case, Nakajima, a Japanese company importing serial-impact dot matrix printers into the Community, applied for the annulment of Regulation 3651/88 imposing definitive anti-dumping duties on its imported products. Nakajima claimed that the duties were in violation of certain provisions of the GATT Anti-Dumping Code and sought to rely on those in order to achieve the annulment of the Regulation. The Council argued that, as with the GATT, the Anti-Dumping Code does not confer on individuals rights which may be relied on before the Court and that the provisions of that Code are not directly applicable within the Community.163 The Court accepted that at the adoption of Council Regulation 2423/88 (the Basic Antidumping Regulation at the time) the Community intended to implement the international commitments stemming from the GATT Anti-Dumping Code. The Community must therefore ensure compliance with its international obligations at the adoption of implementing measures. Thus, the legality of Regulations imposing anti-dumping duties should be examined against the provisions of the GATT Anti-Dumping Code.164 The Court of Justice confirmed the validity of the Nakajima doctrine in the WTO context in Petrotub,165 a case that concerned the imposition of anti-dumping duties on imports of certain seamless pipes and tubes of iron or non-alloy steel originating in Romania. It explained, again in the context of anti-dumping, that since the Community, pursuant to the Basic Anti-Dumping Regulation, intended to transpose the WTO Anti-Dumping Agreement (ADA), the Court may review the legality of Community measures in the field of anti-dumping under the ADA. The Court went even further to explain that those rules being subsumed within the Community legal system attract the application of an additional layer of protection prescribed by this system, in that case Article 253 EC Treaty on the obligation to provide reasons. The implication which flows is that the ADA provisions in question shall be treated qua Community law. The Court then laid down an additional implication : the requirement to state reasons in that particular case should be interpreted in the context of anti-dumping, namely the procedure provided in Article 2.4.2 ADA and the Commission’s undertakings assumed within the WTO Committee on Anti-Dumping,166 thereby fully fusing the two legal systems in the context of anti-dumping. 160 Eeckhout supra note 27 at p. 56 uses the term ‘implementation’ instead of ‘ transposition’. 161 Chiquita at para. 120. 162 Case C-69/89 Nakajima All Precision Co. Ltd v. Council [1991] ECR I-2069. 163 Nakajima at para. 27. 164 Nakajima at paras. 31–32. 165 Case C-76/00P Petrotub SA and Republica SA v. Council [2003] ECR I-79 at paras. 54–60. 166 Petrotub at para. 59.



The Nakajima doctrine establishes that the Court may review the legality of Community legislation against WTO law. The Court’s jurisdiction to deal with such matters does not only equip affected parties with substantive arguments stemming from the ADA but also establishes the possibility that parallel proceedings may be instituted challenging the same Community legislation both before the Court of Justice and WTO dispute settlement. In the EC–Bed Linen case,167 a Panel and the Appellate Body had the opportunity to examine the conformity of Council Regulation 2398/97 imposing definitive anti-dumping duties on cotton bed linen originating in India with the ADA. In its Report, the Panel concluded that the said Regulation violated certain provisions of the ADA.168 The Community appealed the Panel findings, and, while the Appellate Body reversed some of the Panel’s conclusions, it maintained that the ‘ zeroing ’ methodology,169 applied by the EC, is inconsistent with Article 2.4.2. ADA.170 The EC–Bed Linen dispute raises important questions regarding the impact of WTO bodies’ interpretations in the Community legal order,171 also in connection with the preceding discussion of direct effect of Panel and Appellate Body rulings.172 Following the facts of the EC–Bed Linen case, what will the Court of Justice do when faced with a complaint against the ‘ zeroing ’ methodology inconsistency with Article 2.4.2 ADA ? In principle, the Court has jurisdiction to decide the issue anyway ; nonetheless, the timing of the parallel proceedings raises several possibilities. If the dispute is still pending in the WTO, the Court could, if it finds inconsistency with the ADA, deprive those proceedings of their subject matter by annulling the Regulation. Equally, it may decide that the ‘ zeroing ’ methodology is consistent with the ADA. In both cases, it can also stay its proceedings in anticipation of the Panel and Appellate Body rulings.173 Staying the proceedings should be the least preferred option ; since the Court has jurisdiction to assess the legality of Anti-Dumping Regulations against the ADA, it should exercise it. A finding of consistency by the Court of the ‘ zeroing ’ methodology with the ADA, followed by a contrasting finding by the WTO bodies shifts to the Community institutions the responsibility to comply with the DSB recommendations and remedy the situation as they would have done in the absence 167 WT/DS141 European Communities – Anti-Dumping Duties on Imports of Cotton-Type Bed Linen from India (complaint by India). 168 WT/DS141/R. 169 In the calculation of the dumping margin the Community applied the following methodology : First, it divided the Indian bedlinen into several categories. In some, the export price was lower than the normal price and in some it was higher, the later being called a ‘ negative dumping margin’. Then, it calculated the average dumping margin calculating the negative dumping margins as zero. Obviously, the ‘zeroing’ methodology resulted into a higher dumping margin. 170 WT/DS141/AB/R. 171 See, Natalie McNelis, ‘What Obligations Are Created by World Trade Organization Dispute Settlement Reports?’ (2003) 37 Journal of World Trade 647. 172 Section 2. 173 Bourgeois, supra note 153 at 121 citing K. P. E. Lasok, The European Court of Justice, Practice and Procedure, 2nd edition (Butterworth, London, 1994), at p. 72.

The European Union and WTO law 71

of a judgment by the Court. When the WTO has dealt with the first issue, it should be conceded that the Court of Justice should accept the legal interpretations found in the DSB recommendations. In this respect, they could be treated as directly effective. The ensuing inconsistency that the direct effect of only certain Panel and Appellate Body rulings is recognized is methodologically sound. The fact that only certain WTO provisions may be directly effective in the Community legal order once transposed, necessitates that the interpretations given by the bodies entrusted with the task of providing security and predictability to the system should follow. In the aftermath of EC–Bed Linen, the Council was faced with the possibility of actions for annulment before the Court of Justice by traders against whose imports the same ‘zeroing ’ methodology had been applied. As mentioned above, the Court would be expected to apply the interpretations contained in the Panel and Appellate Body Reports, annul the Regulations ex tunc, and require the reimbursement of the collected duties. In response, the Council adopted the Enabling Regulation174 whose provisions shall be analysed below.175 Suffice it to say here that it requires the Council to take the necessary measures to bring Community acts in conformity with the rulings provided in the DSB recommendations. The Enabling Regulation provides prospective remedies and expressly states the intention to avoid the reimbursement of the collected duties.176 The interest of the Council in pre-empting the Court of Justice’s jurisdiction offers strong evidence in favour of the argument that in the field of Anti-Dumping and Subsidies, the Panel and Appellate Body Reports are directly effective. The recognition of direct effect in areas falling within the transposition/ implementation exception means that the infringement of WTO law will readily establish a sufficiently serious breach that will trigger Community liability.177 However, in the circumstances of EC–Bed Linen, a claim for compensation, while legally possible, is unlikely to succeed. As mentioned above, traders have a dual avenue to enforce their rights under the ADA. In addition, it could be argued that WTO law falling within the scope of the Nakajima exception, whilst allowing incidental review of legality, does not grant rights to individuals. Finally, it would be difficult to argue that in, for instance, applying the ‘ zeroing’ methodology at the calculation of the dumping margin – a common practice internationally in the field

174 Council Regulation (EC) No 1515/2001 on the measures that maybe taken by the Community following a report adopted by the WTO Dispute Settlement Body concerning anti-dumping and antisubsidies matters O.J. L 201, 26/07/2001, p. 10. Hereinafter, the Enabling Regulation. McNelis, supra note 171 at p. 670. 175 See below. 176 Article 3 Enabling Regulation. 177 Case C-5/94 R v. Ministry of Agriculture, Fisheries and Food, ex parte Hedley Lomas [1996] ECR I-2553 at para. 28.



of anti-dumping – the Community has manifestly and gravely disregarded the limits of its discretion.178 Clear reference The Court confirmed in Portuguese Textiles and subsequent cases that, where a Community measure refers expressly to the precise provisions of the WTO agreements, the Court may review the legality of the Community measure in question in the light of those WTO rules.179 The Court was referring to Fediol, a judgment delivered during the GATT era, which established this exception and laid down the conditions for its application.180 In that case, the association of Seed Crushers and Oil Processors brought a complaint before the Commission requesting the latter to initiate, on the basis of the New Common Commercial Policy Instrument,181 proceedings against certain alleged illicit commercial practices employed by Argentina. Fediol claimed that those practices were in violation of certain provisions of the GATT. Following an investigation, the Commission concluded that there was no violation and Fediol applied to the Court for the annulment of the Commission’s decision. The Commission based its defence of inadmissibility on the argument that GATT has no direct effect. The Court however held that it cannot be inferred from the lack of direct effect that ‘citizens may not, in proceedings before the Court, rely on the provisions of GATT in order to obtain a ruling on whether conduct criticized in a complaint lodged under Article 3 of Regulation No 2641/84 constitutes an illicit commercial practice within the meaning of that Regulation.’182 The gist of the ruling is that, irrespective of the lack of direct effect, the Commission, at the exercise of its discretion on whether to pursue a complaint under the GATT Dispute Settlement, has to interpret the relevant GATT provisions. This should not preclude the private parties having an interest and being involved in the procedure from requesting the judicial review of the Commission’s decision. In essence, since the Commission possesses the prerogative of interpretation of the GATT for the purposes of initiating a GATT/WTO complaint it should not then hide behind the lack of direct effect of the GATT.183 It is preferable that review of its decision be available to interested parties.

178 Joined Cases C-46/93 and C-48/93 Brasserie du Peˆcheur and Factortame [1996] ECR I-1029 vat para. 55; Joined Cases C-178/94, C-179/94, C-188/94, C-189/94, C-190/94 Dillenkofer and Others v. Germany [1996] ECR I-4845 at para. 25. 179 Portugal v. Council at para. 49; Biret at para. 53. 180 Case 70/87 Fediol v. Commission [1989] ECR 1825. 181 Council Regulation 2641/84 of 17 September 1984 on the strengthening of the common commercial policy with regard in particular to protection against illicit commercial practices O.J. L252, 20/09/ 1984, p. 1. 182 Fediol at para. 19. 183 See the Commission’s arguments in paragraph 18 of the judgment.

The European Union and WTO law 73

At the time the Court delivered its ruling on Portuguese Textiles, no Commission interpretation taken under the Trade Barriers Regulation, the successor of the New Common Commercial Policy Instrument, which was adopted by the Council for the Community to exercise its rights under the WTO Agreement,184 had been challenged. The clear reference exception was implicitly confirmed by the CFI in FICF.185 The significance of the clear reference exception can hardly be overstated. Unlike the transposition exception, GATT/WTO law here is not interpreted qua Community law but on its own merits. What is more important is that, in the ‘ clear reference ’ cases – overlooking for a moment that the challenge involves a Community act – it is, in essence, not the conformity of Community legislation tested against WTO law but the legislation of another WTO Member. The consequent application of WTO law by the Court in these cases is not such as to have any further effects in the Community legal order. Conversely, the interpretation given is limited within the facts of ‘ the given case ’ and ‘ certain specific commercial practices ’.186 Consistent interpretation The previous three instances of the coactive approach concern the application of WTO law by the Court of Justice itself and not the Member States’ courts. The doctrine of consistent interpretation, however, applies in all instances when both the Court of Justice and Member States’ courts are called to interpret otherwise non-directly effective WTO law. The Court of Justice, as is characteristic of courts of several Member States, primarily those belonging to the dualist tradition,187 consistently made efforts to interpret EC law in conformity with the Community’s international obligations.188 Regarding WTO law, the case laying down the foundations of the doctrine of consistent interpretation is Herme`s.189 Herme`s was a French company whose trade mark ‘Herme`s ’ was infringed by FHT. The interpretation of Article 50(6) of the TRIPs Agreement, which provides for provisional measures for the

184 Council Regulation 3286/94 of 22 December 1994 laying down Community procedures in the field of the common commercial policy in order to ensure the exercise of the Community’s rights under international trade rules, in particular those established under the auspices of the World Trade Organization O.J. L349, 31/12/1994, p. 71. 185 Case T-317/02 Fe´de´ration des industries condimentaires de France (FICF) and Others v. Commission of the European Communities [2004] ECR II-4325. 186 Paraphrasing paragraph 20 of the Fediol judgment. Emphasis added. 187 Gerrit Betlem and Andre´ Nollkaemper, ‘ Giving Effect to Public International Law and European Community Law before Domestic Courts. A Comparative Analysis of the Practice of Consistent Interpretation ’ (2003) 14 European Journal of International Law 569 at pp. 574–575. The principle is well established in the United States cf. Charming Betsy. 188 Case C-61/94 Commission v. Germany (International Dairy Agreement) [1996] ECR I-3989 at para 52; Case C-90/92 Dr Tretter v. Hauptzollamt Stuttgart-Ost [1993] ECR I-3569 at para. 11; Case C-286/90 Poulsen and Diva Navigation [1992] ECR I-6019 at para. 9. 189 Case C-53/96 Herme`s International v. FHT Marketing Choice BV [1998] ECR I-3603.



protection of intellectual property rights, was raised in proceedings before Dutch courts, and the Dutch court referred the matter to the Court of Justice. In view of the fact that the Court had found in Opinion 1/94 that the competence under TRIPs Agreement was shared between the Community and the Member States, without any allocation between them,190 the extent of the Court’s jurisdiction was also raised. The Court held that, because Regulation 40/94 on the Community trademark provided for provisional measures and those measures should be taken in the light of the wording and purpose of Article 50(6) TRIPs,191 the Court had jurisdiction to interpret it, even if the facts of the case did not concern a Community trademark but one registered in the Benelux.192 This is because it is clearly in the Community interest that, in order to forestall future differences of interpretation, Article 50(6) TRIPs should be interpreted uniformly whatever the circumstances in which it is to apply.193 The Court’s assumption that national courts were to interpret Article 50(6) TRIPs, read together with Advocate General Tesauro’s Opinion favouring direct effect for Article 50(6) TRIPs, raised the question of direct effect. The Court avoided the question and simply extended to national courts the duty to interpret national provisions in the light of international agreements concluded by the Community.194 The essence of the doctrine of consistent interpretation was later made clear in Dior where the Court’s reasoning can be summarized, in a general manner, into the following proposition : in areas falling within the subject-matter of the WTO Agreement and the Community has already legislated, courts of the Member States must by virtue of Community law interpret the provisions of national law as far as possible in the light of the otherwise non-directly effective provisions of the WTO Agreement.195 Overall, the duty of consistent interpretation provides a satisfactory alternative to the full direct effect of WTO law. While acknowledging that, owing to their special nature, WTO rules are not capable of being enforced in the Community legal order, their undoubted importance at the construction of Community legislation in areas of substantive legislative overlap is thereby restored.196 The practical implication of these judgments is that WTO law will be interpreted and applied by Community and national Courts on a daily basis save for when it is in conflict with Community law.

190 Herme`s at para. 24. 191 Herme`s at para. 28. 192 Herme`s at para. 30. 193 Herme`s at para. 32. See also, Joined Cases C-300/98 and C-392/98 Dior and Assco [2000] ECR I-11344 at paras. 35–37. 194 Herme`s at para. 35. 195 Dior at para. 49. 196 P. Koutrakos, EU International Relations Law (Hart Publishing, 2006), at p. 288.

The European Union and WTO law 75

The political institutions The examination of the intention of the parties is of cardinal importance for the determination of direct effect of any international agreement.197 Whilst the Court referred to the preambular clause of the Council Decision concluding the WTO Agreement in order to deny, in principle, direct effect to the WTO Agreements, it is argued here that the case law analysed above under the coactive approach was intended to give effect to the intention of the Community institutions as demonstrated in the NCPI and TBR,198 and the Anti-Dumping199 and Anti-Subsidies Regulations.200 The political institutions demonstrated strong evidence of their intentions in the legislative acts under examination, and the Court responded to the signal that led to the Nakajima and Fediol judgments. The consistency of this strategy is demonstrated by the subsequent legislative activities of the Community, especially in the field of dumping and subsidies. In the example of the Enabling Regulation, the interaction between the Court and the political institutions of the Community in the application of WTO law becomes manifest. The Enabling Regulation As mentioned above, shortly after the adoption by the DSB of the Appellate Body Report in EC–Bed Linen, the Council adopted Regulation 1515/2001 (the Enabling Regulation) laying down the measures to be taken by the EC so as to comply with adverse Panel and Appellate Body Reports.201 In accordance with the Enabling Regulation, the Community should either amend or repeal the disputed measures or adopt any special measures deemed to be appropriate in the circumstances in order to follow the Panel and Appellate Body rulings.202 The Commission may request all parties to submit all necessary information and it may conduct a review insofar as this is appropriate.203 The Council may also suspend the application of the measure but only for a limited period of time.204 The Council and the Commission may also review measures, not the subject of the dispute, if 197 Case 12/86 Meryem Demirel v. Stadt Schwa¨bisch Gmu¨nd [1987] ECR 3719 and especially the Opinion of Advocate General Darmon. 198 Case 70/87 Fediol v. Commission [1989] ECR 1825. 199 Council Regulation 384/96 of 22 December 1995 on protection against dumped imports from countries not members of the European Community [1996] OJ L56, 06/03/1996, p. 1. 200 The Fediol doctrine should apply to Anti-subsidies and Countervailing duties mutatis mutandis. See, Council Regulation 2026/97 of 6 October 1997 on protection against subsidized imports from countries not members of the European Community [1997] OJ L288, 21/10/1007, p. 1. 201 Council Regulation (EC) No 1515/2001 on the measures that maybe taken by the Community following a report adopted by the WTO Dispute Settlement Body concerning anti-dumping and antisubsidies matters O.J. L 201, 26/07/2001, p. 10. See, Geert A. Zonnekeyn, ‘ The Bed Linen Case and its Aftermath: Some Comment of the European Community’s ‘‘World Trade Organization Enabling Regulation’’’ (2002) 36 Journal of World Trade 993 ; Dan Horowitz, ‘ A Regulated Scope for EU Compliance with WTO Rulings’ (2001) 7 International Trade Law and Regulation. 153; McNelis, supra note 171. 202 Article 1(1) Enabling Regulation. 203 Article 1(3) Enabling Regulation. 204 Article 1(4) Enabling Regulation.



affected by the legal interpretations made in the Report adopted by the DSB,205 although clearly, in WTO law, they are not obliged to.206 It should be recalled that in EC–Bed Linen207 the Appellate Body found that the practice of ‘ zeroing ’ applied by the EC at the determination of the dumping margin was incompatible with Article 2.4.2 ADA.208 In the first instance of application of the Enabling Regulation, the Commission issued a Notice inviting all importers against whom the ‘ zeroing’ methodology had been applied at the determination of the antidumping duties, to request a review on the basis of Article 2 of the Enabling Regulation and in the light of the WTO Panel and Appellate Body interpretations.209 From a Community law perspective, it must be argued that the Community is not acting as a ‘ good citizen ’ but in support of its own interests.210 After all, the Community could still adopt all necessary measures on the basis of Article 133 EC Treaty. However, as McNelis211 pointed out, the Council adopted the Enabling Regulation because it provides further options in addition to the interim review provided in the Basic Anti-dumping Regulation.212 Moreover, the procedure enshrined in the Enabling Regulation not only seeks to enhance the transparency, predictability and automaticity of the response to a ruling but, as mentioned above, to deter the Court from drawing inspiration from the interpretations contained in the Panel and Appellate Body rulings in line with the Nakajima doctrine, annulling Anti-Dumping Regulations and ordering the reimbursement of the collected duties to the affected traders. This mechanism of self-defence has the welcome, from a WTO perspective, repercussions of extending the legal interpretations given by the Panels and Appellate Body to sets of facts unrelated to the WTO dispute and beyond the res judicata created by the Reports. At the same time, it pre-empts the Community Courts from applying these interpretations and reserves this right to the Council and the Commission. Once in control, the political institutions can utilize the WTO Dispute Settlement System in order to coerce the Community’s trading partners to comply with the same legal interpretations and establish a level-playing field. For instance, on the issue of ‘zeroing ’ methodology, as soon as the dispute with India ended, the EC initiated a dispute against the 205 Article 2 Enabling Regulation. 206 McNelis, supra note 171 at pp. 659–661. 207 WT/DS141 European Communities – Anti-dumping measures on imports of cotton-type bedlinen from India. 208 Paragraph 6.119 of the Panel Report upheld by the Appellate Body in Paragraph 86(1) of its Report. 209 Notice regarding the anti-dumping measures in force following a ruling of the Dispute Settlement Body of the World Trade Organisation adopted on March 2001 O.J. C 111, 08/05/2002, p. 4. 210 McNelis, supra note 171 at p. 666. 211 Ibid., at p. 649. 212 Article 11(3) Council Regulation 384/96 of 22 December 1995 on protection against dumped imports from countries not members of the European Community, O.J. L56, 06/03/1996, p. 1.

The European Union and WTO law 77

United States challenging their use of the ‘ zeroing ’ methodology at the implementation of their anti-dumping legislation.213 Compliance with Panel and Appellate Body rulings The first few years of operation of the WTO and its Dispute Settlement System were marked by the two major Hormones and Bananas disputes and the perceived resistance of the Community institutions to comply with the Panel and Appellate Body rulings. It could be argued that these instances were seen as running counter to the Community’s stated policy of compliance with the WTO Agreement. Some years on, it seems that the Community is developing an excellent record of compliance with rulings while settling most disputes at the diplomatic stage of dispute settlement.214 In the Bananas dispute, the Community adopted legislation to bring its regime into conformity with the rulings,215 it requested a waiver to maintain its current regime for a transitional period,216 and, despite some difficulties,217 the Community seems to be bringing the matter to a satisfactory conclusion.218 Similarly, in Hormones, the Community not only did it take measures which, in its opinion, complied with the DSB recommendations,219 but also challenged the continuing retaliation by Canada and United States.220 In sum, one could describe the coactive approach as the Community playing by the rules.

213 WT/DS294 United States – Laws, Regulations and Methodology for Calculating Dumping Margins (‘ Zeroing’) complaint brought by the EC on 19 June 2003. 214 Elisa Baroncini, ‘ The European Community and the Diplomatic Phase of the WTO Dispute Settlement Understanding ’ (1998) 18 Yearbook of European Law 157. Recent examples include Council Regulation 510/2006 of 20 March 2006 on the protection of geographical indications and designations of origin for agricultural products and foodstuffs, O.J. L93, 31/03/2006, p. 12 adopted so as to comply with the DSB recommendations in WT/DS174 European Communities – Protection of Trademarks and Geographical Indications for Agricultural Products and Foodstuffs and Council Regulation 980/2005 of 27 June 2005 applying a scheme of generalised tariff preferences amending the Community’s GSP conditions so as to comply with the DSB recommendations in WT/DS246 European Communities – Conditions for the Granting of Tariff Preferences to Developing Countries. 215 Council Regulation 216/2001 of 29 January 2001 amending Regulation (EEC) No 404/93 on the common organisation of the market in bananas, O.J. L31, 02/02/2001, p. 2; Council Regulation 2587/ 2001 of 19 December 2001 amending Regulation (EEC) No 404/93 on the common organisation of the market in bananas, O.J. L345, 29/12/2001, p. 13. 216 WT/MIN(01)/16, European Communities – Transitional Regime for the EC autonomous tariff rate quotas on imports of bananas, Decision of 14 November 2001, Ministerial Conference, Fourth Session, Doha, 9–14 November 2001. 217 WT/L/616 European Communities – The ACP – EC Partnership Agreement – Recourse to Arbitration pursuant to the Decision of 14 November 2001, Award of the Arbitrator, 1 August 2005. 218 See, in particular, prembular clause (5) of Council Regulation 1964/2005 of 29 November 2005 on the tariff rates for bananas O.J. L316, 2/12/2005, p. 1. 219 Directive 2003/74 of the European Parliament and of the Council of 22 September 2003 amending Council Directive 96/22/EC concerning the prohibition on the use in stockfarming of certain substances having a hormonal or thyrostatic action and of beta-agonists, O.J. L262, 14/10/2003, p. 17. 220 WT/DS320 US – Continued suspension of obligations in the EC – Hormones dispute; WT/DS321 Canada – Continued suspension of obligations in the EC – Hormones dispute.



4. Proactive approach The term proactive defines the position of the Community institutions which is not limited to the observance of the obligations undertaken by the Community and the Member States under the WTO Agreements but is perceived to promote WTO law as the standard for the conduct of international trade externally and the benchmark for the adoption of internal legislation. In the context of global governance, proactivity has been demonstrated in the full support provided by the Community institutions to the WTO and the confidence entrusted to its dispute settlement system. Among the institutions, the Court’s approach towards WTO law cannot be immediately characterized as proactive, a position which is in contrast with the one taken by Advocates General in their Opinions. It is uncontested that the main locus of proactivity towards the WTO has been the practice of the Community’s political institutions. In the proactive approach, the Community actively encourages the application of WTO law. WTO law is set as the normative benchmark for the Community’s internal and external policies and international agreements. Regarding internal policies, the political institutions adopt legislation that purports to be in conformity with WTO law and is normally presumed to achieve this objective.221 Consequently, it is not uncommon for Community legislation to state that its provisions comply with the relevant provisions of the WTO covered agreements.222 At the formulation of Community policies, the Commission intends to make them WTO compliant and, in fact, goes to great lengths to develop a full WTO-compliance test at its interaction with the other institutions.223 The realization that WTO law is omnipresent in the everyday activities of the Commission DGs, as well as the services of the Council and the European Parliament, clearly indicates that there is an emerging WTO culture, which started to dominate the law-making process within the Community. The role of WTO law in the Community’s external relations raises several important considerations. The WTO and its legal system have developed into an essential instrument of the Community’s trade diplomacy. Politically, the Community assists and actively encourages the broadening of the WTO membership by sponsoring the accession process of its important trading partners.224 The Community fully engages in the long and cumbersome accession process developing countries and formerly centrally planned economies are faced with, it introduces WTO law by reference in its relations with non-WTO members, and

221 Snyder, supra note 1 at p. 316. 222 See, Sebastiaan Princen, ‘EC Compliance with WTO Law: the Interplay of Law and Politics’ (2004) 15 European Journal of International Law 555; Sebastiaan Princen, EU Regulation and Transatlantic Trade (Kluwer Law International, 2002). 223 de Bu´rca and Scott, supra note 1. 224 The EU acted as a catalyst to the Chinese accession and is doing the same with the accession of the Russian Federation to the WTO.

The European Union and WTO law 79

treats the WTO Agreements as the common vocabulary for the conduct of international trade. For example, international agreements concluded before the entry into force of the WTO Agreement made reference to the forthcoming conclusion of the Uruguay Round, rendering the agreements’ provisions subject to amendment in the light of the results of multilateral negotiations.225 In some cases, agreements concluded by the Community will stipulate the revision of their provisions when accession to the WTO is achieved by the other contracting party.226 All in all, it is hard to find any post-1995 Association Agreements, Partnership and Cooperation Agreements, Trade and Development Agreements, Stabilization and Association Agreements concluded by the Community without a detailed reference to WTO law.227 The relations thus established are instrumental towards the furthering of the objectives of the multilateral system as it becomes clear for WTO and non-WTO Members alike what standard they are expected to follow in their trade relations with the Community and its Member States. In so far as bilateral agreements of a general nature are concerned, turning WTO rules into the applicable standard will inevitably strengthen the culture of WTO compliance in the Community’s legislative practice. In addition, WTO law has been at the forefront of sectoral agreements concluded with developed trading partners, in particular the US. For instance, the preamble of Council Decision 98/258/EC on the conclusion of the Agreement between the European Community and the United States of America on sanitary measures to protect public and animal health in trade in live animals and animal products states, ‘ Whereas the Agreement between the European Community and the United States of America on sanitary measures to protect public and animal

225 Article 59 of the Europe Agreement establishing an Association between the European Communities and their Member States, of the one part, and the Republic of Bulgaria, of the other part O.J. L358, states: ‘The provisions of Chapters II, III, and IV of Title IV shall be adjusted by decision of the Association Council in the light of the result of the negotiations on services taking place in the Uruguay Round and in particular to ensure that under any provisions of this Agreement a Party grants to the other Party a treatment no less favourable than that accorded under the provisions of a future GATS Agreement. ’ 226 See, for instance, Articles 4, 5, 16 and Annex 2 of the Agreement on partnership and cooperation establishing a partnership between the European Communities and their Member States, of one part, and the Russian Federation, of the other part, O.J. L327, 28/11/1997, pp. 3–69. 227 Notably, Articles 20, 24, 61, 70, 78, 80, 83, 89, 92, and 103 of the Agreement establishing an association between the European Community and its Member States, of the one part, and the Republic of Chile, of the other part, O.J. L352, 30/12/2002, pp. 3–1439. The EC – Chile Agreement is particularly interesting in that it not only it refers to the WTO generally or the WTO Annexed Agreements specifically, but also to Decisions taken by bodies established under the WTO Agreement, in particular, the TBT Committee (Article 85). More importantly, it subordinates the dispute settlement system established under the Agreement to the WTO DSU should any of the contracting parties chooses to seek redress for a violation of obligations under the EC – Chile in the WTO (Article 189(4)(c)). Also, Articles 6, 15, 29, 65 of the Stabilization and Association Agreement between the European Communities and their Member States, of the one part, and the former Yugoslav Republic of Macedonia, of the other part, O.J. L84, 20/ 03/2004, pp. 13–81 and in particular Article 69(5) which for the purposes of states aids sets the relevant WTO rules and Community legislation as alternative normative frameworks.



health in trade in live animals and animal products provides an adequate means for putting into practice the provisions of the WTO Agreement on the application of sanitary and phytosanitary measures as regards public and animal health measures. ’228 In fact, the agreement establishes a Joint Committee, which is entrusted with the task to guide the activities carried out under the Agreement.229 At a first glance, such agreement must be interpreted as a bilateral instrument, which seeks to apply and consequentially promote the WTO rules. However, scratching below the surface, one must note that, although the development of mutually acceptable SPS standards must be seen as furthering the objectives and principles of the WTO Agreement, the SPS Agreement in this instance, it may also have the side effect of restricting market access to non-participants to such a bilateral arrangement.230 This remark notwithstanding, the Community’s policy to conclude bilateral mutual recognition agreements with reference to WTO law clearly forms part of its proactive approach. From the point of view of judicial enforcement, it could be argued that provisions of agreements concluded by the Community containing clear reference to the WTO Agreement could develop direct effect ; predominantly so, regarding agreements with states non-members of the WTO, and, therefore, incapable of availing themselves of the WTO DSU. This is reinforced by the requirement inserted in such agreements to be implemented ‘ in full conformity with the provisions of the WTO ’.231 Importantly, according to the Court’s case law, the nature and purpose of these agreements do not preclude individuals from invoking their provisions in national courts.232 When such provisions refer to WTO rules, the answer to the question of their judicial enforceability depends on whether the specific provisions of the WTO Agreement to which the Community’s bilateral agreements refer will be treated as forming part of the latter agreements. If this is the case, they should be analysed in the light of the nature and purpose of the Community’s agreements and not the WTO Agreement. Were this to be accepted and should they contain a clear, precise, and unconditional obligation, they will be enforced in national courts in accordance with the established case law.233 In the alternative, it can be argued that in such cases both the Community and the associated states intended to transpose specific WTO rules in their legal orders. 228 O.J. L118, 21/4/1998, p. 1. 229 Article 14(1) of the Agreement. 230 M. J. Trebilcock and R. Howse, The Regulation of International Trade, 3rd edition (Routledge, 2005), at p. 230. 231 Inter alia, Article 34 of the Partnership Agreement between the members of the African, Caribbean and Pacific Group of States of the one part, and the European Community and its Member States, of the other part, signed in Cotonou on 23 June 2000, O.J. L317, 15/12/2000, p. 3. 232 Case C-265/03 Simutenkov [2005] ECR I-2579 extended this principle to Partnership and Cooperation Agreements. See also, Case C-438/00 Deutscher Handballbund [2003] ECR I-4135; Case C-18/90 Kziber [1991] ECR I-199; Case C-262/96 Su¨ru¨l [1999] ECR I-2685; Case C-162/00 Pokrzeptowicz-Meyer [2002] ECR I-1049. 233 Ibid.

The European Union and WTO law 81

Extending the Nakajima doctrine to cover these cases too should not be ruled out.234 With reference to the effect of Panel and Appellate Body rulings, Article 85 of the Partnership and Cooperation Agreement (PCA) with Azerbaijan may be relevant in that it states : When examining any issue arising within the framework of the Agreement in relation to a provision referring to an article of the GATT/WTO, the Cooperation Council shall take into account to the greatest extent possible the interpretation that is generally given to the article of the GATT/WTO in question by the Members of the WTO.235

This provision affords the interpretation that institutional bodies set up by international agreements concluded by the Community must apply the interpretations rendered by the Panel and Appellate Body adopted by the DSB. This goes further than the nature of DSB recommendations under the WTO Agreement itself and possibly further than the Enabling Regulation, analysed above. In addition, should the conditions identified in the previous paragraph apply, and WTO rules become directly effective by virtue of reference to them in the Community’s international agreements, then Panel and Appellate Body interpretations of these provisions must be taken into account. If the opposite is the case, it will be difficult to argue in favour of direct effect of Panel and Appellate Body rulings for the reasons identified by the Court in the analysis of the reactive approach, in particular after the Court’s judgment in Van Parys. By contrast, it could be argued that the proviso in the PCA with Azerbaijan may simply extend the general interpretative duty and practice of the Community institutions to employ the interpretations rendered by the WTO bodies236 to organs established under international agreements concluded by the Community. The above analysis is illustrative of the emergence of WTO law as a standard at the carrying out of the Community’s internal and external policies. The importance of WTO law as the standard for the conduct of international trade regardless, it is difficult to surmise the purpose behind the institutional WTOphilia, particularly evident within the Commission. Clearly, it is the conflict-avoidance – and consequently, WTO dispute avoidance – strategy of the Commission than the suitability of the WTO norms. Observing the Community’s international commitments is important; however, they can only serve as a framework wherein the 234 As explained by the CFI in Chiquita at para. 124 ‘ The applicant rightly argues that application of the Nakajima case law is not, a priori, limited to the area of anti-dumping. It is capable of being applied in other areas governed by the provisions of the WTO Agreements where those agreements and the Community provisions whose legality is in question are comparable in nature and content to those just referred to above concerning the Anti-Dumping Codes of the GATT and the anti-dumping basic regulations which transpose them into Community law.’ 235 Partnership and Cooperation Agreement between the European Communities and their Member States, of the one part, and the Republic of Azerbaijan, of the other part, O.J. L246, 17/9/1999, p. 3. 236 Rosas, supra note 100.



Community formulates its policies in order to achieve its own constitutional objectives. These are, in some instances, significantly different from those of the WTO. Unless an express reference to the need for compliance with WTO is inserted in the preamble to the EC Treaty, the Community should primarily concern itself with the formulation of policies in the interests of the Community and its Member States which may not necessarily coincide with, and are normally more complex than, the rudimentary framework established by the WTO. While the proactive approach is, in general, welcome, limits to the proactive practice of the Community’s political institutions must be set.

5. Synthesis and critique The starting point for the assessment of the application of WTO law in the Community legal order should be the balancing act the Community institutions, including the Court, need to perform between two competing considerations : the supremacy of international law and the supremacy of Community law.237 While the supremacy of international law has always been appealing to commentators regarding the Community as a sui generis legal order founded by international law,238 the Court has been vigilant at the support of the supremacy of Community law.239 In fact, the Court, in Opinion 1/91, was quick to acknowledge the supremacy of the Community Treaties over provisions of the proposed EEA Agreement. It held that the jurisdiction of the proposed EEA Court affected the allocation of responsibilities as defined by the EC Treaty and therefore, undermined the autonomy of the Community legal order.240 Inevitably, the recognition of direct effect of WTO law would deprive the Court of Justice from the authority to uphold the supremacy of the Community. The scales in the balancing act could be represented by the concepts of monism and dualism ; monism being inherently prone to accord supremacy to international law and dualism to domestic/Community law. The catalysts for the balancing act in this sense are ‘compatibility ’ and ‘ direct effect ’. The starting point for the analysis of the relationship between international law and Community law should be Article 300(5) which provides that when an agreement ‘ calls for ’241 amendments to the Treaty, those must be adopted first before the agreement is concluded. Further, Article 300(6), which concerns the advisory jurisdiction of the Court, provides that where the Court of Justice finds that an envisaged international agreement is incompatible with the Treaty, the agreement may enter into force only if the Treaty is amended. These paragraphs represent a definitive statement 237 Klabbers, supra note 4 at p. 271. 238 Pescatore, supra note 31. 239 Christian Timmermans, ‘ The EU and Public International Law’ (1999) 4 European Foreign Affairs Review 181. 240 Opinion 1/91 [1991] ECR I-6079. 241 Emphasis added.

The European Union and WTO law 83

that the Treaties, as the ‘Constitution ’ of the Community, are supreme and cannot be subordinated to provisions of international agreements unless the authors of the Treaty so decide.242 Article 300(7) states ‘ Agreements concluded under the conditions set out in this Article – namely, the conditions enshrined in the previous six paragraphs of Article 300, including paragraphs 5 and 6 – shall be binding on the institutions of the Community and on the Member States’ ; does this presume that those agreements shall be binding insofar as they are compatible with the Treaty ? This constitutes a reasonable assumption. The obligation of compatibility of international agreements has recently been extended also to the internal rules. Indeed, the Council and the Commission are now charged with the task to ensure this at the negotiation of international agreements.243 While, from a WTO legal perspective,244 this amendment is not a novelty as it integrates a pre-existing obligation into the Community Treaties, the Council and the Commission will be faced with a considerable task at the conclusion of the Doha Round of multilateral trade negotiations if they are properly to discharge this responsibility. Nonetheless, while the ‘ compatibility ’ catalyst is relatively unexplored, its emergence heralds a notable shift of the Treaty authors to a dualist approach towards WTO law. This will be seen in the future, account being taken of the proliferation of studies comparing the substantive law of the WTO and the EU245 and the building of a body of jurisprudence by the Panels and Appellate Body of the WTO. Moving on to the ‘direct effect ’ catalyst,246 the Court has been unequivocal. Its outright denial of direct effect of WTO law points towards a dualist understanding and the supremacy of Community legal order over the international one. The notable exceptions analysed in the context of the coactive approach prove the rule. Beyond the reasons analysed previously, some further considerations and conditions for change of course will follow. The WTO – unlike the Community, which created a new legal order whose subjects are not only the Member States but also their nationals247 – does not enjoy such high aspirations.248 On the contrary, it was designed in the traditional public international law sense in which states and 242 Bourgeois, supra note 153 at p. 97. 243 Article 133(3) EC Treaty as amended by the Treaty of Nice. 244 Article XIV :4 WTO Agreement. 245 Marise Cremona, ‘Neutrality or Discrimination ? The WTO, the EU and External Trade’ in G. de Bu´rca and J. Scott (eds.), The EU and the WTO: Legal and Constitutional Issues (Hart Publishing, 2003) p. 151; F. Ortino, Basic Legal Instruments for the Liberalisation of Trade: A Comparative Analysis of EC and WTO Law (Hart Publishing, 2004); Joanne Scott, ‘International Trade and Environmental Governance : Relating Rules (and Standards) in the EU and the WTO’ (2004) 15 European Journal of International Law 307; M. Slotboom, A Comparison of WTO and EC Law, Do Different Objects and Purposes Matter for Treaty Interpretation? (Cameron May, 2006). 246 Klabbers, supra note 4 at pp. 292–298. 247 Case 26/62 Van Gend en Loos [1963] ECR 1 ; Opinion 1/91 [1991] ECR I-6079 at para. 21. 248 ‘ Following this approach, the GATT/WTO did not create a new legal order the subjects of which comprise both contracting parties or Members and their nationals.’ Panel Report in WT/DS152/R United States – Sections 301–310 of the Trade Act of 1974 at para. 7.72.



international organizations are the main subjects. Despite its broad scope, it is clear that its Members never wished it to become a new world legal order, comprising also their citizens. Its structure and mechanisms are such so as to exclude the citizens from being part of the system.249 Further, the WTO is characterized by the absence of a norm-generating mechanism.250 The rounds of multilateral trade negotiations strike a delicate balance of rights and concessions, which, owing to the intergovernmental nature of the agreement, confer public law rights belonging to the WTO Members rather than the individual traders. The jewel of this international law structure is the rigorous dispute settlement system established by the DSU. The clear and unambiguous nature of its provisions can be advocated as a reason for opening up the Community to the direct effect of WTO law. In this respect, one wonders what purpose such a rigorous dispute settlement serves if the WTO Agreement intended to confer rights on individuals enforceable in WTO Members ’ courts.251 It has been argued that since the old GATT flexibility is gone, there is no need to insist upon the lack of direct effect, as now there is a binding international adjudication. This argument is not convincing at all. On the contrary, the bindingness and rigour of the DSU militate against direct effect. Private parties have national avenues to invite their governments to pursue their interests in Geneva and the guarantee that once there, a report creating a binding obligation in international law will, sooner or later, serve their interests. This is further strengthened by the fact that, at least in the context of the TBR, the Court of Justice has granted private parties a significant avenue of control of the institutions at the exercise of their functions, thus offering a counterbalance for the denial of direct effect.252 Were the opposite the case, the choice of market access barriers to be challenged would be determined by the interests of private parties and not the WTO Members. This transfer of control clearly falls beyond what the parties have agreed when concluding the Uruguay Round Agreements and establishing the multilateral trading system under the administration of the WTO. From the point of view of encouraging compliance with the WTO, it should be pointed out that enforcement in national and Community courts can only have the effect of repealing the WTOinconsistent legislation. This hardly resembles the compliance envisaged within the DSU itself. From a policy point of view, it is in principle undesirable to have the WTO bodies and Community Courts handling the same cases. They are destined

249 With the exception of NGOs in an attempt to embrace civil society (Article V:2 WTO Agreement) and certain procedural rights. See, Steve Charnovitz, ‘ The WTO and the Rights of the Individual’ (2001) 36 Intereconomics : Review of European Economic Policy 98. 250 de Bu´rca and Scott, supra note 1 at pp. 2–7. 251 See also AG Lenz’s Opinion in Case C-469/93 Amministrazione delle Finanze dello Stato v. Chiquita [1995] ECR I-4533 at para. 21. 252 Case 70/87 Fediol v. Commission [1989] ECR 1825; Case T-317/02 Fe´de´ration des industries condimentaires de France (FICF) and Others v. Commission of the European Communities [2004] ECR II-4325.

The European Union and WTO law 85

to create conflicts, something that hardly facilitates the objective of the DSU to provide ‘security and predictability to the multilateral trading system ’.253 In the absence of a preliminary reference system, mirroring the one established under the Community constitutional order,254 direct effect of WTO law can hardly serve the multilateral trading system.255 A realistic argument however needs to be made. It is unfair for individual traders to pay the bill either by being subject to market access barriers or suffering the consequences from the suspension of concessions affecting their trade. Does this in any way mitigate the stance against direct effect of the WTO and, in particular, Community liability for breach of WTO law ? The answer should be in the negative. What this realization does is to expose the shortcomings of the WTO system and encourage proposals for its reform instead of challenging the proper understanding of this system by the Court of Justice. Rosas has suggested that the system of suspension of concessions should be abolished in favour of a system of compensation, whereby the Arbitrators under Article 22.6 DSU will calculate the amount of nullification or impairment of benefits suffered by a WTO Member and determine the sum of compensation due.256 This is a proposal which should muster support in the reform of the DSU process. Regarding the proactive approach, the WTOization of the Community legislative engine represents a controversial realization and raises issues of both substance and process. Regarding substance, opening up to a legal system with lower standards of environmental protection, public health and labour laws, and a comprehensive global membership represents a cause for concern.257 Moreover, it has been argued that the normative subordination of the Community to the WTO is destined to put the European social model at risk.258 The concerns raised are exacerbated by the dramatic change in the European Union’s economic and political architecture resulting from its enlargement.259 More importantly, regarding process, are we convinced that the WTO is better suited for the European Union than its own law-making processes, a result of no little effort, compromise and continuous debate ?260 Issues of democracy, legitimacy, and accountability are raised in this context too.261 The European Union despite its more developed characteristics in this respect, still has a long way to

253 Article 3.2 DSU. 254 Rosas, supra note 1. 255 See also Eeckhout, supra note 57 at p. 99. 256 Rosas, supra note 72 at p. 144. 257 But see, Marco M. Slotbloom, ‘ Do Public Health Measures Receive Similar Treatment in European Community and World Trade Organization Law? ’ (2003) 37 Journal of World Trade 553. 258 Antoniadis, supra note 143 at p. 343. 259 D. C. Vaughan-Whitehead, EU Enlargement versus Social Europe? The Uncertain Future of the European Social Model (Edward Elgar, Cheltenham, 2003). 260 Eeckhout, supra note 57 at p. 100. 261 Ibid.



tread.262 However, citizens affected in their everyday lives still find it difficult to accept the normative dominance of Community law encapsulated in the principle of direct effect and its concomitant, the principle of supremacy.263 This is more so within the WTO context. What is increasingly worrying is the impression given by the Commission that it treats WTO law as the ‘supreme law ’. The substantive falseness of this approach regardless, the absence of any public debate at the formulation of WTO rules – instead, the single undertaking procedure is followed in the multilateral rounds of trade negotiations – , the limited role reserved to the European Parliament at the negotiation and conclusion of the agreement on behalf of the European Union, and the fact that dispute settlement takes place behind closed doors,264 indicate that there is less likelihood for WTO law to be received with enthusiasm. The position of the Commission is inherently paradoxical, as the Commission itself has identified the need for increase in legitimacy and accountability within the European Union context.265

Conclusions As Trachtman put it ‘the question of direct effect is a political decision ’.266 The Community Treaties, along with the Council Decision concluding the WTO Agreement, represent the authentic political statement by the Member States on the issue of WTO law. The Court has appropriately responded and dismissed the calls from commentators to undervalue the normative merit of the relevant clause in the Council Decision concluding the WTO Agreement.267 If the combined interpretation of the case law, the legislative activity and the institutional practice means that the Community legal order is a dualist one for the purposes of the application of WTO law, then so be it.268 Following from this, unless the Community transforms WTO law into the Community legal system by means of transposition into its own legislative instruments, WTO law cannot have direct effect. Hermeneutically, this means that the Community chose WTO law as a second best set of rules. In its internal policy-making, it uses WTO law as a benchmark and accepts its primacy in its commercial policy instruments. Ordinarily, it tries to interpret legislation consistently with the WTO Agreements. 262 A. Arnull and D. Wincott (eds.), Accountability and Legitimacy in the European Union (Oxford University Press, 2003). 263 The disaffection towards the European Union looming in many Member States and made manifest in the recent French and Dutch referenda is illustrative of this proposition. 264 Which is not reversed by the groundbreaking decision to hold the proceedings in public in WT/ DS320 US – Continued suspension of obligations in the EC – Hormones dispute; WT/DS321 Canada – Continued suspension of obligations in the EC – Hormones dispute. 265 White Paper on European Governance, COM(2001)428, 25 July 2001. 266 Joel Trachtman, ‘ Bananas, direct effect and compliance ’ (1999) 10 European Journal of International Law 655 at p. 664. 267 Consequently, it is not the Court playing a political role as Klabbers suggested, supra note 4 at p. 298. 268 Eeckhout, supra note 27 at pp. 24–29.

The European Union and WTO law 87

Apart from those exceptions, which do not, in essence, involve the application of WTO law but its communitarized version, WTO law may not be invoked in the Community and national courts, predominantly so if it is to challenge Community legislation. In the external relations of the Community, WTO law may serve as a particularly useful benchmark. The use of WTO norms in the material provisions of the Community’s international agreements will facilitate trade liberalization and will have a positive impact on the establishment of a level-playing international trading field. In sum, the analysis of the case law and institutional practice leads to the conclusion that the Community possesses a finely tuned system for the application of WTO law, which is the result of the interaction between the Court and the political institutions. It is not as receptive to WTO law as some commentators would prefer, but neither as inconsistent as often accused. Until significant changes within the WTO and its law-making mechanism materialize to enable WTO law to play a more important role, this elaborate nexus of approaches should be considered as thoroughly satisfactory.

World Trade Review (2007), 6 : 1, 89–122 Printed in the United Kingdom

Dispute settlement corner WTO Case Law : The American Law Institute Reporter’s Studies This section publishes one case in each issue from an American Law Institute (ALI) annual volume published by Cambridge University Press. The ALI series is edited by Henrik Horn and Petros C. Mavroidis and commenced with The WTO Case Law of 2001 published in 2003. These new annual volumes are part of an ALI project on World Trade Organization Law. The project will undertake yearly analyses of the case law from adjudication bodies of the WTO. Each of the cases is jointly evaluated by an economist and a lawyer, both well known experts in the fields of trade law and international economics. The Reporters critically review the jurisprudence of the WTO adjudicating bodies and evaluate whether the ruling in each case ‘ makes sense’ from an economic as well as a legal point of view, and, if not, whether the problem lies in the interpretation of the law or the law itself. The Studies do not always cover all issues discussed in a case, but they seek to discuss both the procedural and the substantive issues that form the ‘ core ’ of the dispute.

f Gene M. Grossman and Alan O. Sykes doi:10.1017/S1474745606003107

United States – Definitive Safeguard Measures on Imports of Certain Steel Products G E N E M. G R O S S M A N Princeton University

A L A N O. S Y K E S Stanford University

Abstract : The Appellate Body decision in the steel dispute is the latest in a line of unsatisfactory decisions in the safeguards area. The problem stems from the fact that the treaty text regarding the preconditions for the use of safeguard measures is badly deficient. The Appellate Body with its usual emphasis on textualism has done little to resolve the puzzles that the text creates. WTO members are left with little guidance about the proper use of safeguards beyond some confusing and sometimes incoherent standards. We review these issues as they have arisen in prior decisions, and then discuss the details of the steel dispute. Our emphasis there is on the panel decision more than the Appellate Body report, as the latter breaks little new ground.



G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

1. Introduction Since the inception of the WTO, safeguard measures have regularly been the subject of dispute settlement proceedings. The latest in this chain of disputes concerns the definitive safeguard measures imposed by the United States on a wide range of steel products in 2002. The safeguards investigation of steel imports was initiated under US law by the US International Trade Commission (ITC) at the request of the United States Trade Representative (USTR) in June 2001. The request covered four broad categories of steel products, which were divided into 33 categories by the ITC for purposes of data collection. Ultimately, the ITC defined 27 separate ‘industries ’ producing steel products within the scope of the investigation. For each of these industries, the ITC proceeded to determine whether imports had increased, and, if so, whether increased imports were a substantial cause of serious injury or threat of injury. This analysis resulted in negative determinations for 15 industries, affirmative determinations for eight industries, and ‘ divided’ determinations (a 3–3 vote) for four industries.1 Under US law, a negative determination by the ITC precludes any action by the President to impose a safeguard measure. Affirmative determinations and divided determinations are forwarded to the President for consideration of possible relief, along with remedial recommendations that the President is not bound to follow. As to some products, the ITC recommended that imports from nations with which the United States has preferential trading arrangements – including Canada, Mexico, Israel, and Jordan – be exempted from any safeguard measures. After the conclusion of an inter-agency review process, the President instituted ten distinct safeguard ‘ measures’ covering various steel products, which generally excluded imports from preferential trading partners. Eight WTO members (including the European Community) challenged the measures pursuant to the WTO dispute resolution process, and the proceedings were consolidated before a single panel. The panel ruled against the United States on all of the challenged measures, on multiple grounds. The Appellate Body affirmed the ruling in considerable part, leading to the eventual dismantling of the measures in December 2003. The issues before the panel and the Appellate Body were by and large familiar ones, and their opinions broke little new ground. The panel decision is by far the more interesting of the two, but, in the end, we view both opinions as simply the latest in a long line of unsatisfactory WTO decisions in the safeguards area. Our critique in this regard draws heavily on our prior work, along with the work of other ALI reporters. See Grossman and Mavroidis (2004), Horn and Mavroidis (2003), Sykes (2003), Sykes (2004).

1 USITC, Certain Steel Products, Inv. No. TA-201-73, (USITC Pub. No. 3479, December, 2001), Vol. I (hereafter USITC Report).

Definitive Safeguard Measures on Imports of Certain Steel Products 91

Section 2 lays out some of the key legal issues in safeguards jurisprudence, and describes the state of the law prior to the steel dispute. Section 3 then considers the issues in the steel case, notes some interesting features of the panel report, and shows how the Appellate Body decision adds little to what had come before it. Section 4 offers some concluding thoughts about the future of safeguard measures in the WTO system and the need for coherent reforms.

2. Legal, historical, and economic background The legal basis for safeguards in the GATT system was found in Article XIX, which became outmoded with the passage of time in a way that created serious confusion regarding the legal prerequisites for the use of safeguards. With the advent of the WTO, a new Safeguards Agreement elaborated additional legal constraints on the use of safeguards, but failed to resolve many of the central puzzles that had arisen under Article XIX. Appellate Body decisions since the inception of the WTO have only made matters worse, to the point that the legal requirements for the use of safeguards are largely incoherent, and no nation can employ them without the near certainty of defeat in the dispute resolution process, should they be challenged. This section briefly reviews Article XIX, the evolution of national practice under GATT, and the legal developments since the adoption of the Safeguards Agreement and prior to the steel decision.

2.1 Article XIX and its interpretive puzzles GATT Article XIX provides in paragraph (1) : If, as a result of unforeseen developments and of the effect of the obligations incurred by a contracting party under this Agreement, including tariff concessions, any product is being imported into the territory of that contracting party in such increased quantities and under such conditions as to cause or threaten serious injury to domestic producers in that territory of like or directly competitive products, the contracting party shall be free, in respect of such product, and to the extent and for such time as may be necessary to prevent or remedy such injury, to suspend the obligation in whole or in part or to withdraw or modify the concession.

Unforeseen developments The first clause of Article XIX provides that safeguard measures are permissible only following ‘ unforeseen developments ’ associated with ‘ the obligations incurred by a contracting party’. What constitutes an ‘ unforeseen development ’? Unforeseen by whom, at what point in time ? How does one determine the ‘ effect of the obligations incurred ’ ? These questions had a natural answer at the outset of the GATT system. The original GATT negotiations concluded in 1947, with the expectation that GATT would be supplanted within a few years by a new institution to be called


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

the International Trade Organization (which, of course, never came into being because of a shift in the political winds). The first clause of Article XIX has a natural interpretation in the context of a trade agreement that was expected to be short-lived. The negotiators had made a number of trade concessions to each other in 1947, and Article XIX provided for their suspension in the event that those concessions had an unforeseen, adverse impact on import-competing industries due to a surge in import competition. To the questions posed above, therefore, one might answer that an ‘unforeseen development ’ was some development that caused the increase in imports following a trade concession under the original GATT to be greater than reasonably expected. It had to be unforeseen by the GATT negotiators, at the time of the 1947 negotiations. And the import surge had to result from one of the original GATT trade concessions, in the sense that it would not have happened but for some such concession. But how does one interpret the requirements of Article XIX (1), first clause, in an agreement that remains in force after many years ? Consider an import surge 30 or 40 years after the agreement was drafted. What would it mean to say that such a surge resulted from the ‘ obligations incurred ’, particularly if those obligations were incurred decades earlier ? Could any such surge have been ‘foreseen ’ given the passage of so much time ? By whom and when ? And how are the answers affected by the fact that GATT negotiations are ongoing, with new negotiating ‘rounds ’ every decade or so ? The requirements of the first clause no longer have a straightforward interpretation in an agreement that lasts for decades rather than a few years, and that is characterized by an ever-changing set of commitments. For these reasons, GATT practice evolved over time toward ignoring the requirements of the first clause in Article XIX(1).2 National laws to authorize safeguard measures soon made no mention of them. Section 201 of the US Trade Act of 1974, for example, simply requires the International Trade Commission (ITC) to determine ‘ whether an article is being imported into the United States in such increased quantities as to be a substantial cause3 of serious injury, or the threat thereof’.4 There is no requirement that developments be ‘ unforeseen ’ or that they result from earlier trade concessions. At this writing, this statute remains the basis for safeguard measures under US law. Such a development is an understandable consequence of the difficulties in giving content to the first clause of Article XIX in a long-lived agreement. But as shall become immediately apparent, the absence of this ‘ anchor ’ for the remainder of Article XIX(1) creates other problems.

2 See McGovern (1986: 291). 3 A ‘substantial cause’ is a ‘ cause that it is important and not less than any other cause’. 19 USC. ·2252(b)(1)(B). 4 19 USC. ·2252(b)(1)(A).

Definitive Safeguard Measures on Imports of Certain Steel Products 93

Serious injury and causation Under Article XIX, it is not enough that an unforeseen import surge results from a trade concession. The import surge must go on to cause or threaten ‘ serious injury ’. This phrasing raises other obvious interpretive issues – what is ‘ serious injury ? ’ How does one determine whether the ‘cause’ of such injury (or threat thereof) is ‘ increased quantities ’ of imports ? On the first question, the text appears deliberately vague. The drafters might have made reference to specifics in this regard – lost profits, unemployment, bankruptcies, and the like – but chose to leave the term undefined. Perhaps the best inference is that they did not want to constrain the concept unduly by attempting a definition, and that they would allow a variety of factors into the analysis. On the question of causation, the logic was nevertheless fairly clear. The unforeseen import surge, resulting from the trade concession, had to be responsible for serious injury. Put differently, the serious injury had to be ‘ caused ’ by the trade concession, via its effect on the level of import competition, in the usual ‘ but for ’ sense of the term ‘cause ’. Within this framework, the ‘ exogenous ’ variable is the trade concession, and the ‘ increased quantities ’ of imports were those resulting from that concession. Likewise, the level of imports in the absence of the trade concession serves as the baseline against which to measure the ‘ increase ’. But now imagine reading out the first clause of paragraph one, as GATT members began to do many years ago. Then, one must simply have ‘ such increased quantities [of imports] _ as to cause or threaten serious injury ’, as, for example, is required by the US law The baseline for the ‘ increase ’, import levels prior to a recent trade concession, is no longer available. Further, the only apparent candidate for an exogenous variable is the ‘increased quantities ’ of imports, as there is no longer any background event from which these ‘ increased quantities ’ result. Two considerable problems arise as a result. First, how does one now determine whether there are ‘ increased quantities ’ of imports at all – against what baseline is the increase to be measured? Second, and more fundamental, how can one treat increased quantities of imports as an exogenous or ‘causal ’ variable ? Elementary economics suggests that the forces of supply and demand will determine the quantity of imports, just as they do prices and domestic production. If imports and domestic products are perfect substitutes, for example, then the quantity of imports will equal the difference between domestic demand and domestic supply at the equilibrium price. That price, in turn, will be determined by the intersection between the domestic demand curve and the total supply curve, comprised of the sum of domestic supply and import supply. The exogenous factors in the economic framework are the determinants of domestic supply, domestic demand, and import supply. Domestic demand is affected by such things as domestic consumer tastes and incomes, domestic supply by the domestic costs of inputs into production


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

and the state of available production technology, and import supply by the parallel factors that affect supply and demand in other countries. The quantity of imports is then a result of the interaction of these forces ; it is not a causal variable at all. Likewise, changes in the quantity of imports will be the result of changes in the determinants of domestic supply, demand, and the world price. Increased quantities of imports may result, for example, from a fall in the world price due to falling input costs abroad, to improved production technology abroad, or to weakening demand abroad. Increased quantities of imports can also result from an increase in domestic demand attributable, for example, to rising consumer incomes. Finally, increased quantities of imports can result from increasing costs of domestic production reflected in a leftward shift of the domestic supply schedule. Against this backdrop, the question ‘ did increased quantities of imports cause serious injury to a domestic industry ? ’ is simply incoherent. Suppose, as an illustration, that the domestic industry suffers a decline due to rising costs. As domestic production falls at the world price, imports will increase to fill the rising gap between domestic demand and supply. Are ‘increased quantities ’ of imports the ‘ cause ’ of this ‘ injury ’? Certainly not in the usual sense of the term ‘ cause ’. By hypothesis, what have changed are the costs of domestic firms, and these changes result in reduced domestic production and increased imports. Hence, once the first clause of Article XIX (1) becomes a nullity, it is by no means clear how nations should operationalize their reliance on Article XIX. There is no longer any natural baseline against which to measure ‘ increased quantities ’, and there is no longer any intelligible exogenous variable to assess as potential ‘ cause ’ of serious injury.

2.2 The Safeguards Agreement For reasons that would take us afield here, relating primarily to the proliferation of ‘ gray-area ’ measures such as voluntary export restraints during the later years of GATT, the Uruguay Round negotiators concluded a new Agreement on Safeguards. Unfortunately, the Agreement did little to resolve the puzzles raised by Article XIX(1). On the basic preconditions for reliance on Article XIX, the Agreement largely parrots US law in stating that a ‘ Member may apply a safeguard measure to a product only if that Member has determined _ that such product is being imported into its territory in such increased quantities, absolute or relative to domestic production, and under such conditions as to cause or threaten serious injury to the domestic industry that produces like or directly competitive products ’.5 Like US law, it omits any reference to ‘unforeseen developments’ or the ‘effect of the obligations incurred ’. 5 Ibid. Art. 2(1).

Definitive Safeguard Measures on Imports of Certain Steel Products 95

The only guidance as to the meaning of ‘ serious injury ’ and to the analysis of causation is provided by Article 4 of the Agreement, which states in pertinent part : 1. For purposes of this Agreement : (a) ‘ serious injury’ shall be understood to mean significant overall impairment in the position of a domestic industry ; (b) ‘ threat of serious injury’ shall be understood to mean serious injury that is clearly imminent _ 2. (a) In the investigation to determine whether increased imports have caused or are threatening to cause serious injury to a domestic industry _ the competent authorities shall evaluate all relevant factors of an objective and quantifiable nature having a bearing on the situation of that industry, in particular, the rate and amount of the increase in imports of the product concerned in absolute and relative terms, the share of the domestic market taken by increased imports, changes in the level of sales, production, productivity, capacity utilization, profits and losses, and employment. (b) The determination referred to in subparagraph (a) shall not be made unless this investigation demonstrates, on the basis of objective evidence, the existence of the causal link between increased imports of the product concerned and serious injury or threat thereof. When factors other than increased imports are causing injury to the domestic industry at the same time, such injury shall not be attributed to increased imports.

Plainly, this provision does not seriously address, let alone resolve, the conundrums presented by modern GATT practice under Article XIX. It does not provide any guidance, for example, on what it means to say that increased imports are a causal variable, or any guidance on what is meant by ‘ factors other than increased imports _ causing injury to the domestic industry at the same time ’. The concept of ‘serious injury ’ is left quite vague, and members must simply ‘ evaluate ’ relevant factors. Further, although the Agreement nowhere refers to ‘ unforeseen developments ’ and the ‘ effect of the obligations incurred ’ as a predicate to safeguard measures, it does not specifically provide that Article XIX(1), first clause, may henceforth be ignored. Hence, fundamental questions regarding the legal prerequisites for safeguard measures remain unanswered by the WTO Agreement on Safeguards. These unresolved issues have found their way into WTO disputes.

2.3 Safeguards in Appellate Body Jurisprudence prior to the Steel Case The resurrection of unforeseen developments As indicated, GATT practice evolved toward ignoring the unforeseen developments requirement of Article XIX, and the Safeguards Agreement says nothing about that requirement. But in its first important ruling in a safeguards


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

dispute – Korea–Dairy6 – the Appellate Body overruled the findings of the dispute panel in the case to the effect that formal compliance with Article XIX(1), first clause, is no longer required. The Appellate Body instead held that a treaty interpreter ‘ must give meaning and effect to all the terms of the treaty. An interpreter is not free to adopt a reading that would result in reducing whole clauses or paragraphs of a treaty to redundancy or inutility ’.7 Article XIX and the Safeguards Agreement are to be read cumulatively, it says, and the first clause of Article remains a binding obligation. As for the proper interpretation of the obligation imposed by Article XIX(1), first clause, the Appellate Body opined : [I]it seems to us that the ordinary meaning of the phrase ‘as a result of unforeseen developments ’ requires that the developments which led to a product being imported in such increased quantities and under such conditions as to cause or threaten to cause serious injury to domestic producers must have been ‘unexpected ’. With respect to the phrase ‘of the effect of the obligations incurred by a Member under this Agreement, including tariff concessions’, we believe that this phrase simply means that it must be demonstrated, as a matter of fact, that the importing Member has incurred obligations under the GATT 1994, including tariff concessions.8

The Appellate Body went on to endorse the reasoning of the working party report in the old GATT Hatter’s Fur case,9 which stated : ‘‘‘ unforeseen developments’’ should be interpreted to mean developments occurring after the negotiation of the relevant tariff concession which it would not be reasonable to expect that the negotiators of the country making the concession could and should have foreseen at the time when the concession was negotiated ’.10 This line of reasoning was repeated by the Appellate Body in Argentina–Footwear,11 which also overruled the dispute panel in the case. Thus, the Appellate Body has fully revived the first clause of Article XIX, and has held in these and subsequent decisions that national authorities have failed to demonstrate their compliance with it. United States–Lamb,12 in particular, holds that WTO members must demonstrate their compliance with the Article XIX(1), first clause, prior to the time that a safeguards measure is undertaken. The US ITC’s failure to consider the matter in its lamb investigation was ‘not surprising ’,

6 Korea – Definitive Safeguard Measure on Imports of Certain Dairy Products, WT/DS98/AB/R (1999). 7 Ibid. "80. 8 Ibid. "84. 9 Sales No. GATT/1951-3 (Nov. 1951). 10 Ibid. "89. 11 Argentina – Safeguard Measures on Imports of Footwear, WT/DS121/AB/R (1999). 12 United States – Safeguard Measures on Imports of Fresh, Chilled or Frozen Lamb Meat from New Zealand and Australia, WT/DS178/AB/R (2001).

Definitive Safeguard Measures on Imports of Certain Steel Products 97

given the absence of any reference to it in the governing US statute ; but that was no defense for the United States under WTO law.13 One can certainly quarrel with the legal soundness of these decisions. Given the uniform practice of ignoring Article XIX(1), first clause, during the latter years of GATT, and its omission from the Safeguards Agreement, it is questionable whether the drafters of the Uruguay Round Agreements had any intention of reviving the obligation – had they wished to alter established GATT practice in this respect, one might argue, they would have so indicated with clarity. The difficult interpretive issues that the clause raises in a long-lived agreement, which led to its irrelevance in GATT practice, might also have been noted as a basis for letting it remain dormant. Having embraced the opposite view, the Appellate Body might have undertaken to explain coherently what Article XIX(1), first clause, now requires. At what point in time must the events in question have been unforeseen – the time of the last tariff concession ? What if the last concession on the product in question was decades ago – could anything today have been foreseen ? What if the product has been the subject of numerous tariff concessions over time – are expectations associated with the last concession the only relevant ones ? Why or why not ? How does one establish the expectations of trade negotiators as an evidentiary matter? What if there are many negotiators and their accounts of their expectations are incongruent ? What if most of them are dead ? This list of questions is assuredly incomplete. With regard to the ‘ effect of the obligations incurred’, by contrast, the Appellate Body apparently offers a construction that enables this requirement to be trivially satisfied in every case – a member simply needs to show that it has incurred some obligations with respect to the product in question. It is hard to imagine how a dispute could arise without such an obligation, since a member with an unbound tariff could always raise it unilaterally without any need to rely on a safeguard measure. The Appellate Body evidently does not require members to demonstrate that ‘ increased quantities ’ of imports are attributable directly to any recent trade concession. It suffices for them to argue that, in the absence of a tariff binding, they would be able to raise tariffs to eliminate the import surge. Increased quantities As noted, Article XIX originally contemplated that ‘ increased quantities ’ of imports would be measured against baseline levels prior to 1947 GATT concessions. Having revived Article XIX(1), first clause, therefore, one might perhaps have expected the Appellate Body to require a similar approach to establishing the baseline against which the existence of ‘ increased quantities ’ is assessed, perhaps

13 Ibid. "73.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

by looking to import levels prior to the most recent concession on the product in question. But it has not taken that approach. In Argentina–Footwear,14 the Appellate Body considered a case in which Argentina had adopted the approach embraced some years earlier by the US ITC – a five-year ‘rule of thumb ’ for establishing the import baseline. The dispute panel in the case concluded that it is ‘ reasonable to examine the trend in imports over a five-year historical period ’.15 But the Appellate Body focused on language from the second clause of Article XIX(1) and its counterpart in Article 2.1 of the Safeguards Agreement : ‘any product is being imported into the territory of that contracting party in such increased quantities and under such conditions as to cause or threaten to cause serious injury ’. The phrase ‘ is being imported’, according to the Appellate Body, ‘ indicates that it is necessary for the competent authorities to examine recent imports, and not simply trends in imports during the past five years – or, for that matter, during any other period of several years ’.16 In our view, the determination of whether the requirement of imports ‘ in such increased quantities ’ is met is not a merely mathematical or technical determination. In other words, it is not enough for an investigation to show simply that imports of the product this year were more than last year – or five years ago. Again, and it bears repeating, not just any increased quantities of imports will suffice. There must be ‘ such increased quantities ’ as to cause or threaten to cause serious injury to the domestic industry in order to fulfill this requirement for applying a safeguard measure. And this language in both Article 2.1 of the Agreement on Safeguards and Article XIX :1(a) of the GATT 1994, we believe, requires that the increase in imports must have been recent enough, sudden enough, sharp enough, and significant enough, both quantitatively and qualitatively, to cause or threaten to cause ‘ serious injury ’.17 Thus, the Appellate Body insists that imports must have increased ‘ recently’. But how recently, and in what amount ? The phrase ‘ recent enough, sudden enough, sharp enough, and significant enough, both quantitatively and qualitatively, to cause or threaten to cause ‘‘serious injury ’’ ’ hardly provides useful guidance. The insistence on ‘ not just any increase ’ but ‘ such increased quantities ’ as to cause injury is equally unhelpful. And one must again confront the fundamental issue that all of this verbiage avoids – what does it mean to say that increased quantities of imports ‘ cause ’ injury when they are, as an economic matter, a result of a variety of possible developments ? Far from lending badly needed clarification, the Appellate Body’s treatment of the ‘ increased quantities ’ requirement only adds to the confusion.

14 Argentina – Safeguard Measures on Imports of Footwear, WT/DS121/AB/R (1999). 15 Ibid. "130. 16 Ibid. 17 Ibid. "131.

Definitive Safeguard Measures on Imports of Certain Steel Products 99

Serious injury Like Article XIX and the Safeguards Agreement, the Appellate Body has not attempted to define ‘serious injury’ with any precision. Its focus has been primarily on the text of Article 4.2, which simply provides : the competent authorities shall evaluate all relevant factors of an objective and quantifiable nature having a bearing on the situation of that industry, in particular, the rate and amount of the increase in imports of the product concerned in absolute and relative terms, the share of the domestic market taken by increased imports, changes in the level of sales, production, productivity, capacity utilization, profits and losses, and employment.

According to the Appellate Body, the text requires that all of the listed factors be ‘ evaluated ’ in every case, and it has found safeguard measures wanting under WTO law whenever a member has failed to discuss one or more of these factors in its official report on safeguard action.18 The Appellate Body has further held that the obligation to evaluate ‘all relevant factors’ may extend to factors not raised by any of the parties to the safeguards investigation.19 Otherwise, the Appellate Body has simply insisted that serious injury represents ‘ significant overall impairment ’ as stated in Article 4.1 of the Safeguards Agreement.20 It characterizes this standard as ‘high ’ and ‘ exacting’.21 It is not necessary that every ‘relevant factor ’ reflect industrial decline, however, for serious injury to be present – ‘ a certain factor may not be declining, but the overall picture may nevertheless demonstrate ‘‘significant overall impairment ’’. ’22 On the whole, therefore, the Appellate Body has provided relatively little guidance on the meaning of ‘serious injury ’, a situation that is perhaps understandable given the vagueness of the pertinent textual obligations. Beyond a requirement that all factors listed in the Safeguards Agreement be ‘ evaluated ’ in each case, it remains unclear what conditions will support a finding of serious injury or threat, and what degree of deference on the matter will be afforded to national authorities.

18 See Argentina – Safeguard Measures on Imports of Footwear, WT/DS121/AB/R (1999), "121. 19 See United States – Definitive Safeguard Measures on Imports of Wheat Gluten from the European Communities, WT/DS166/AB/R (2001), "55. 20 To date, the Appellate Body has largely refrained from detailed commentary on the reasoning behind findings of ‘ serious injury’ by national authorities. The most notable exception is United States – Safeguard Measures on Imports of Fresh, Chilled or Frozen Lamb Meat from New Zealand and Australia, WT/DS178/AB/R (2001). The US ITC had found lamb prices in the United States to be ‘depressed’ even though they were generally higher than four or five years earlier. And it had found a threaten of serious injury even though prices had risen toward the end of its period of investigation. The Appellate Body held these findings to be insufficient to support the ITC determination. Ibid. ""157–159. 21 United States – Safeguard Measures on Imports of Fresh, Chilled or Frozen Lamb Meat from New Zealand and Australia, WT/DS178/AB/R (2001), "124. 22 Argentina – Safeguard Measures on Imports of Footwear, WT/DS121/AB/R (1999), "139.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

Causation The Appellate Body has addressed the causal relationship between increased quantities of imports and serious injury in several opinions prior to the steel case. None of them, however, provides clear answers to the conceptual difficulties identified in Section I. Argentina–Footwear briefly addresses the proper method for determining whether imports are the ‘cause’ of injury. The dispute panel in the case had indicated that ‘if causation is present, an increase in imports normally should coincide with a decline in the relevant injury factors ’.23 The Appellate Body agreed with the panel that in an analysis of causation, ‘ it is the relationship between the movements in imports (volume and market share) and the movements in injury factors that must be central to a causation analysis and determination’. Furthermore, with respect to a ‘coincidence ’ between an increase in imports and a decline in the relevant injury factors, we note that the Panel simply said that this should ‘normally’ occur if causation is present.24

Hence, in its first important statement on the subject, the Appellate tips its hat to the notion that correlation and causation are not the same, but implies that they ‘normally ’ go hand in hand. One has no sense that the Appellate Body is aware of (or at least troubled by) the profound conceptual difficulty in confounding the two in a setting where the ostensible ‘causal ’ variable is in fact endogenous. The other Appellate Body opinions on causal analysis focus principally on the so-called ‘ non-attribution requirement ’ of Article 4.2 of the Safeguards Agreement. It provides that safeguard measures may not be employed unless the ‘investigation demonstrates, on the basis of objective evidence, the existence of the causal link between increased imports of the product concerned and serious injury or threat thereof. When factors other than increased imports are causing injury to the domestic industry at the same time, such injury shall not be attributed to increased imports. ’ One question raised by this language during the course of various disputes has been whether the harm ‘caused ’ by increased imports (again suspending the issue of what it means to treat increased imports as causal) must by itself suffice to cause serious injury, or must simply contribute to serious injury, perhaps along with other factors. To this ill-posed question, the Appellate Body has responded that ‘ the Agreement on Safeguards does not require that increased imports be ‘‘ sufficient ’’ to cause, or threaten to cause, serious injury. Nor does that Agreement require that increased imports ‘ alone ’ be capable of causing, or threatening to cause, serious injury ’.25 23 Argentina – Safeguard Measures on Imports of Footwear, WT/DS121/AB/R (1999), "141. 24 Ibid. "144. 25 United States – Safeguard Measures on Imports of Fresh, Chilled or Frozen Lamb Meat from New Zealand and Australia, WT/DS178/AB/R (2001), "170. See also United States – Definitive Safeguard Measures on Imports of Wheat Gluten from the European Communities, WT/DS166/AB/R (2001), "70.

Definitive Safeguard Measures on Imports of Certain Steel Products 101

Although increased imports need not account for all of the serious injury, the Appellate Body nevertheless underscores the importance of ensuring that injury caused by ‘ factors other than increased imports ’ ‘not be attributed to increased imports ’. To make sense of these dual principles, one can only assume that the Appellate Body is concerned about situations in which increased imports have not made any causal contribution to serious injury, and where serious injury is nevertheless wrongly ‘ attributed’ to imports. It has found fault with members’ ‘ non-attribution analysis ’ on multiple occasions. In United States–Wheat Gluten,26 the volume of imports had risen 38 % during the five-year period of investigation employed by the ITC. Over the same period, US productive capacity had grown 68 %. Capacity utilization at US firms had fallen considerably along with profits, however, and the US ITC had linked the decline in profitability to declining capacity utilization rates.27 One issue before the Appellate Body was whether the US ITC had incorrectly ‘ attributed’ injury caused by the expansion of US capacity to rising imports. On this question, the Appellate Body noted that had US capacity not risen over the period of investigation, its capacity utilization rate would have fallen only modestly, even with the increased volume of imports.28 Further, had imports maintained their market share over the period of investigation, capacity utilization rates still would have fallen significantly due to the increased capacity brought on line.29 In the face of this evidence, the Appellate Body concluded that the US ITC had not ‘ adequately evaluated the complexities ’ and had not ensured that injury attributable to other factors is not attributed to imports.30 The Wheat Gluten opinion is problematic in a number of respects. First, as with the other Appellate Body opinions on causation, it does nothing to help with the question of how to conceptualize imports as a causal variable. Second, taking seriously for a moment the notion that imports are ‘ causal ’, it was undisputed in the case that they had risen substantially during the period of investigation, while the profitability of domestic producers had fallen. Given the Appellate Body’s earlier pronouncements that a correlation of this sort is ‘ normally’ present when a causal connection exists, and that imports need not account for all serious injury, one wonders why this evidence was not enough. The logic of the Appellate Body opinion seems to suggest that the problems suffered by US producers were caused by two factors – rising imports and investment in new capacity that proved unnecessary. In the absence of either factor, US producers would have been considerably more profitable. Why, then, is it inappropriate to attribute at least part of the ‘ serious injury ’ to imports ? 26 United States – Definitive Safeguard Measures on Imports of Wheat Gluten from the European Communities, WT/DS166/AB/R (2001). 27 Ibid. ""81, 84. 28 Ibid. "85. 29 Ibid. "86. 30 Ibid. "91.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

Finally, and as the United States had argued, much of the increase in capacity was put in place before imports began to increase.31 The sequence of events thus suggested that US producers had invested in new capacity in anticipation of growth opportunities, but that imports had increased to capture those growth opportunities and render the new investment uneconomical. It could thus be argued that the unexpected surge in imports was the real ‘ problem ’, and that investments in capacity were not a conceptually distinct cause of injury but rather a background predicate for the injury caused by imports. To this line of argument, the Appellate Body Responded : [T]he relevance of an ‘other factor’, under Article 4.2(b), depends on whether that ‘ other factor’ was, or was not, ‘causing injury’ ‘at the same time ’ as increased imports. Therefore, the possible relevance of the increases in capacity added during the period of investigation does not depend on the moment in time when the increases in capacity occurred, but on when the effects of those increases are felt, and whether they are ‘causing injury’ ‘at the same time ’ as increased imports.32

This response simply begs the question as to why domestic investments in new capacity should be considered an ‘ other factor ’ distinct from imports as a cause of injury, if indeed the anticipated recoupment of those investments was frustrated by an unexpected import surge. The US ITC had treated capacity investments as an alternative cause of injury to be sure (and dismissed them as less important), and the Appellate Body simply seemed to accept it as an ‘ other factor ’ without reflection. The decision in United States–Lamb33 is similar in this last respect. The US ITC had considered six factors other than increased imports that might have contributed to serious injury : ‘ the cessation of subsidy payments under the National Wool Act of 1954 ; competition from other meat products, such as beef, pork and poultry ; increased input costs ; overfeeding of lambs ; concentration in the packing segment of the industry ; and a failure to develop and maintain an effective marketing program for lamb meat ’.34 The Appellate Body again seemed to accept these factors uncritically, and simply inquired whether the United States had done enough to ensure that injury caused by these factors was not ‘ attributed ’ to imports. Once again, it found the analysis of the ITC wanting, suggesting that it consisted of conclusory assertions without reasoned explanation.35 Along the way it added : ‘We emphasize that the method and approach WTO Members choose to carry out the process of separating the effects of increased imports and the effects

31 Ibid. "87. 32 Ibid. "88. 33 United States – Safeguard Measures on Imports of Fresh, Chilled or Frozen Lamb Meat from New Zealand and Australia, WT/DS178/AB/R (2001). 34 Ibid. at "182, n. 57. 35 Ibid. ""185–186.

Definitive Safeguard Measures on Imports of Certain Steel Products 103

of the other causal factors is not specified by the Agreement on Safeguards. What the Agreement requires is simply that the obligations in Article 4.2 must be respected when a safeguard measure is applied’.36 Evidently, members can use any analytical method they wish that complies with Article 4.2, yet the Appellate Body offers no counsel as to what the set of permissible methods might include. Finally, in United States–Line Pipe,37 the US ITC had considered the possibility that decreased oil and gas drilling was a more important cause of injury than increased imports, and had concluded to the contrary. Once again, however, its analysis was deemed insufficient – the ‘ cited parts of the USITC Report do not establish explicitly, with a reasoned and adequate explanation, that injury caused by factors other than the increased imports was not attributed to increased imports. The passage on page I-30 of the USITC Report highlighted by the United States is but a mere assertion that injury caused by other factors is not attributed to increased imports ’.38 In sum, the Appellate Body decisions prior to the steel case regarding the causal analysis required by the Safeguards Agreement suggest the following principles : (a) correlation is typically the best evidence of causation ; (b) the ‘ other factors ’ considered by national authorities during the course of their investigations will be accepted uncritically without any reflection as to their logical relevance; and (c) the Appellate Body will not tell nations how to conduct their ‘ non-attribution analysis ’, but will insist that it contain ‘ reasoned and adequate explanation ’, which has so far been lacking in every case. In these decisions, the Appellate Body offers no theory as to how imports are to be viewed as causal, or as to how members should determine what constitutes a potential ‘ other factor ’. It faults the lack of ‘ adequate explanation ’ in the decisions of national authorities, yet its own explanation of the permissible role for safeguard measures could hardly be less instructive.

3. The steel dispute The recent steel dispute raised all of the issues noted above, and others. The panel decision hints at a partial resolution of the ‘unforeseen developments ’ puzzle, but makes little progress on other fronts. The Appellate Body decision breaks no new ground at all, and holds the steel safeguards imposed by the United States to be illegal for predictable reasons in light of the prior cases.

3.1 The panel decision The challenges to the US steel safeguard measures collectively attacked every aspect of their legal basis. The panel exercised judicial economy to avoid

36 Ibid. "181. 37 United States – Definitive Safeguard Measures on Imports of Circular Welded Carbon Quality Line Pipe from Korea, WT/DS202/AB/R (2001). 38 Ibid. "220.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

addressing arguments regarding the definition of ‘ industry’ in the US investigation and the existence of serious injury. Its findings instead focused on four areas : the existence of ‘ unforeseen developments’ as a predicate for safeguards; the question whether steel imports had increased in ‘such increased quantities ’ as to permit safeguards; the causal link between increased imports and injury; and the lack of ‘parallelism ’ between the injury analysis and the remedial measures.39 Unforeseen developments US law has yet to be amended to require that increased imports result from unforeseen developments, and the initial ITC decision in the steel case predictably paid little heed to the issue. But prior to the imposition of the steel safeguard measures in March 2002, the USTR requested additional information from the ITC regarding unforeseen developments, and received a supplemental report on that issue in February. Based in large part on this supplemental report, the United States argued before the panel that four unforeseen developments had contributed to the influx of imports that had injured the US steel industry : the Asian financial crisis, the drop in demand for steel due to the dissolution of the Soviet Union, the unexpected strength of US demand for steel, and the persistent appreciation of the US dollar.40 Among other arguments, the complainants urged that none of these events was ‘ unforeseen ’, and questioned whether they had resulted in increased imports sufficient to justify safeguards. In assessing these issues, the panel began by noting that the parties agreed that ‘the point in time at which developments should have been unforeseen is that of the completion of the Uruguay Round’.41 The panel allowed that the Asian financial crisis could constitute an unforeseen development ‘since it took place after the United States last negotiated its tariff concessions on the steel products covered by the investigation ’.42 The same was true of the consequences for the steel market of the dissolution of the Soviet Union, even though that process had begun prior to the end of the Uruguay Round. The ongoing strength of the US economy and the US dollar were harder to regard as unforeseeable, but the panel concluded that these factors were not viewed by the ITC as unforeseen developments in themselves, but simply circumstances that contributed to the increase in imports that resulted from the developments in Asia and the Soviet Union. Thus, the United States prevailed on the proposition that unforeseen developments had affected the steel market to some degree. According to the panel’s interpretation of the text of GATT Article XIX, however, the unforeseen developments must produce increased imports that cause serious injury or threat. It was on this issue that the United States failed to persuade 39 United States – Definitive Safeguard Measures on Imports of Certain Steel Products, WT/DS24849, 251-54, 258-59/R (July, 2003) (hereafter Panel Rep.). 40 Ibid. "10.40. 41 Ibid. "10.74. 42 Ibid. "10.80.

Definitive Safeguard Measures on Imports of Certain Steel Products 105

the panel, in part because the ITC findings of injury were all contained in the original ITC report rather than its supplemental report – at no time prior to its affirmative injury findings did the ITC identify the increased imports that had resulted from the unforeseen developments and analyze their impact on the domestic industry. Such analysis was required, according to the panel, for every line of steel products (every ‘ industry ’) in which a safeguard measure was taken.43 Instead, the ITC had simply asserted in its supplemental report, after the original injury findings had been made, that unforeseen developments had affected the steel market in a general way – as the panel stated, ‘ in light of the complexity of the matter, a more sophisticated and detailed economic analysis was called for ’.44 The panel’s treatment of the unforeseen developments issue might prove helpful in suggesting how to operationalize this requirement in practice, but at the same time underscores that it is a potentially severe hurdle for any nation that seeks to employ safeguard measures. As to the questions posed earlier in this essay (unforeseen by whom ? ; at what point in time?), the panel suggests that the relevant actors are trade negotiators, and that the relevant time is the point when the member seeking to use safeguards ‘ last negotiated its tariff concessions on the _ products covered by the investigation ’. The end of the Uruguay Round may be taken to have ‘ reset the clock ’ on this latter issue, as GATT members formally withdrew from their old GATT obligations at the end of the Round and entered a new (WTO) treaty, even if the tariff bindings on many products did not change. In deciding whether events were ‘unforeseen ’ by the negotiators at the relevant time, the focus in the first instance will be on whether the events in question took place before or after the conclusion of the negotiations. One can thus imagine the rules here evolving in a manner that is roughly consistent with the rules in nonviolation cases – a presumption might arise that negotiators foresee the results of events that take place before the conclusion of negotiations, and do not foresee the results of events that take place later.45 As tariffs decline to minimal levels and more and more time passes since the last concession, the set of ‘unforeseen ’ events will presumably expand and unforeseen developments should become easier to identify. The panel’s demand for linkage between unforeseen developments and particular import increases, and the further requirement that these imports be linked to serious injury or threat, is a substantially new interpretation of the obligations imposed by GATT Article XIX(1). The great virtue of requiring such a linkage between unforeseen developments and injury, via the effect of unforeseen developments on import quantities, is that a coherent exogenous variable thereby

43 Ibid. "10.128. 44 Ibid. "0.125. 45 See Japan – Measures Affecting Consumer Photographic Film and Paper, WT/DS44/R (1998), "10.79.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

resurfaces in the analysis – one asks not whether ‘increased imports ’ have caused injury in the abstract, but whether particular unforeseen developments have caused injury, via an effect on the relative competitive position of imported and domestic goods. The question that national authorities are asked to answer is once again economically intelligible, as it was at the outset of GATT. But the new obligations inherent in the panel’s interpretation do pose some substantial analytical challenges. It is hardly clear what sort of ‘ more sophisticated and detailed economic analysis ’ will suffice. Consider the steel case itself : By the panel’s reasoning, the United states should have ascertained precisely by how much US imports had increased, in each of 27 steel ‘ industries ’, as a result of the Asian financial crisis and the drop in demand for steel inside the old Soviet Union. It should then have analyzed whether this increase caused or threatened to cause serious injury to the relevant industry (as well as provide a convincing ‘ nonattribution ’ analysis, discussed further below). On the surface, such an analysis seems to call for a global general equilibrium model of each segment of the steel market, so that the effect of events in particular overseas markets such as Asia and the former Soviet Union can be simulated with precision. If that sort of analysis is indeed required, the time and expense involved could be enormous. The accuracy of such exercises is also subject to considerable doubt because the results often turn on controversial assumptions. And, for many industries, the data necessary to estimate the parameters for such models will be lacking, and modelers would have little choice but to fall back on simulations that rest on seat of the pants guesses about relevant supply and demand elasticities, cross-elasticities of demand, and the like. The potential boon for consulting economists is readily apparent, but one must harbor no illusions that the task of undertaking such analysis is straightforward – if done properly, it is expensive, time-consuming, and inevitably fraught with the potential for serious error. Perhaps a future reviewing panel would be satisfied with something less daunting (and thus less rigorous and even more error prone), but the question of what will suffice as a ‘ reasoned and adequate explanation ’ of the linkage between unforeseen developments and injury remains unclear at best. The problem of ‘ non-attribution ’ might also appear to become more tractable under the panel’s interpretation of ‘ unforeseen developments’ – one might say that as long as the unforeseen developments cause increased imports, in turn resulting in injury or threat, then by definition injury has not been ‘attributed ’ to any factor other than imports. The task of identifying the other factors to which importrelated injury must not be ‘attributed ’, and of assessing their impact, could arguably be put to the side. Conceptual issues still remain, however, as to what is permitted to ‘ count ’ as an unforeseen development. Imagine, for example, an ‘unforeseen ’ shock in the domestic market for inputs into steelmaking that raises the cost of steel production in the United States. US steelmakers raise their prices to cover costs, and imports flood into the US market to undercut the price increases and market shares of domestic firms. Such ‘ increased imports ’ may surely be said to cause injury (relative to a counterfactual world in which the imports are not

Definitive Safeguard Measures on Imports of Certain Steel Products 107

permitted to increase), and to result from the unforeseen developments in the domestic input market. But are safeguard measures appropriate when the root cause of injury is a shock in the domestic economy ? Or would safeguards to remedy such injury be impermissible because the injury caused by the domestic shock is wrongly ‘ attributed ’ to imports ? Nothing in the WTO decisions thus far affords much help with such matters, an issue about which we will say more below in connection with the ‘non-attribution’ problem. Increased imports and the baseline question Taking its cue from earlier Appellate Body decisions, the panel held ‘ that the use of the present tense in the verb phrase ‘‘is being imported’’ in both Article 2.1 of the Agreement of Safeguards and Article XIX :(1)(A) of the GATT 1994 indicates that it is necessary for the competent authorities to examine recent imports and that the increase in imports was ‘‘recent’’’.46 Further, ‘ the emergency nature of safeguard measures calls for an assessment of whether imports increased suddenly so that the situation became one of emergency’.47 And finally, ‘ [i]n light of the Panel’s above conclusion that the competent authority must have determined that imports increased suddenly and recently, the Panel will generally focus its analysis on the situation of imports in the more recent period that preceded the end of the period of investigation ’.48 The panel then proceeded to examine the data on import trends for each of the ten product categories covered by the challenged measures. In most instances, the panel generated a graph representing the import volume and market share data over the five-year period of investigation employed by the ITC. Because the panel’s focus was on the ‘more recent period that preceded the end of the period of investigation ’, the graph for each product category says much about how the panel came out in each case. We reproduce four of the graphs below for purposes of illustration (see Figure 1). For both hot-rolled bar and stainless steel rod, the panel found that the ITC report did not contain a ‘ reasoned and adequate ’ explanation of why imports had increased. Regarding hot-rolled bar, the panel focused on the ITC’s failure to account for the most recent data from interim 2001 _ The decrease from interim 2000 (1.34 million tons) to interim 2001 (952,392 tons) represented a decrease by 28.9 %, whereas the increase in the year-to-year period before (1999 to 2000) that was characterized as ‘ rapid and dramatic’ was merely 11.9 %. In light of this decrease in the most recent period, the Panel does not believe that the trend of imports from 1996 to 2000 (an increase by 52.5 %) is sufficient to provide a basis for a finding that, at the moment of the determination, hot-rolled bar ‘is being imported in such increased quantities’.49 46 Panel Rep. "10.159. 47 Ibid. "10.166. 48 Ibid. "10.175. 49 Ibid. "10.205.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

Figure 1 HOT-ROLLED BAR Imports (Tons) 3,000,000




0 1996




1999 2000 2001 Annual data Semi-annual (Jan-Jun) data



Imports (Tons) 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0 1996




1999 2000 2001 Annual data Semi-annual (Jan-Jun) data

REBAR Imports (Tons) 2,000,000


1,500,000 1,000,000 500,000 0 1996




1999 2000 2001 Annual data Semi-annual (Jan-Jun) data



160,000 140,000 120,000 100,000 80,000 60,000 40,000 20,000 0 1996




1999 2000 2001 Annual data Semi-annual (Jan-Jun) data

Definitive Safeguard Measures on Imports of Certain Steel Products 109

The analysis was similar in many respects for stainless steel rod. The USITC relied on the increase occurring between 1996 and 2000, with the largest increase from 1999 to 2000 (25%). The decline between interim 2000 and interim 2001 was acknowledged, but the USITC did not give an explanation why it nevertheless found that there was an increase of imports in absolute numbers. This failure is particularly serious since this decrease (by 31.3 %) was sharper than the preceding increase, and, as a matter of proportion, offset the increase of the two preceding years.50

For rebar, by contrast, the panel accepted the ITC finding of increased imports : In light of the tripling of imports, the decrease over the last 18 months is not significant enough in order to stand in the way of a conclusion that rebar ‘is being imported in such increased quantities’.51 Likewise, as to stainless steel bar, the panel found that ‘ in the light of the significant increase from 1999 to 2000 (19.3 percentage points), the decline by 3.3 percentage points from interim 2000 to interim 2001 is, contrary to what the European Communities has stated, insignificant ’. It simultaneously does not detract from a finding that imports, relative to domestic production, remain at high levels so that stainless steel bar ‘is being imported in (such) increased quantities’. 52

For two product categories, tin mill products and stainless steel wire, the panel ruled that the United States failed to demonstrate the presence of increased imports because it had relied on the separate opinions of ITC Commissioners who had defined the relevant ‘industries ’ in different ways. For example, some treated ‘ tin mill products ’ as a separate industry, while another included tin mill products in a broader industry. The panel was of the view that the separate findings reached in such fashion could not collectively constitute a ‘ reasoned and adequate ’ explanation for the finding of increased imports. The panel’s reasoning focused on two issues in each instance : whether imports had risen substantially over the entire period of investigation (the five-year period ordinarily used by the ITC as its baseline), and whether any recent downtrend in imports had undercut the finding of an overall increase. Implicitly, recent trends carry more weight than the five-year trend, but a modest recent decline in imports would not prevent a finding of increased imports if the five-year trend was more dramatically upward. The panel’s approach is at least somewhat puzzling, for a couple of reasons. First, having held earlier that the United States should have linked import increases to the Asian financial crisis and the dissolution of the Soviet Union, one might have expected the panel to suggest that the timing of those events defines the baseline for measuring the increase in imports. Instead, as under longstanding US practice, the panel (and the complainants) seems to accept that the arbitrary five-year baseline is 50 Ibid. "10.267. 51 Ibid. "10.225. 52 Ibid. "10.254.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

permissible in principle: ‘The complainants do not challenge the choice of a fiveyear period of investigation per se. Complainants rather disagree with the fact that, generally, the USITC did not focus sufficiently on the situation of imports in the latest part of the period of investigation. ’53 Second, had the United States in fact linked an increase in imports to unforeseen developments such as the Asian financial crisis and the dissolution of the Soviet Union, and had it shown that the extra imports resulting from these developments were sufficient to cause serious injury, would that not have been enough to establish a basis for safeguard measures? Why must the United States additionally show a sudden, sharp, and significant increase in total imports ? Should the right to safeguard measures be eliminated, for example, if the effects of the unforeseen developments, working through increased imports, had been reinforced by a slowdown in US demand, which might have caused additional injury but served to temper the growth in import volume ? The great emphasis on the most recent year or months of data is peculiar in another respect. The time series for imports of any good may exhibit significant volatility for a variety of reasons, and the notion that WTO members would wish to condition the right to use safeguards heavily on the most recent import fluctuations, which may be quite unpredictable when an investigation is initiated, seems odd even if it has some arguable textual basis. The preamble to the Agreement on Safeguards emphasizes the importance of ‘ structural adjustment’, much as US law has long set forth the alternative goals of promoting industrial competitiveness or facilitating an orderly industrial contraction.54 If these stated goals are to be taken seriously, they concern measures to address long-term structural trends. Likewise, Article 7 of the Safeguards Agreement provides that safeguard measures may be imposed for four years, with the possibility of an extension to eight years. The potential duration of the measures is also suggestive of the notion that they address long-term trends in industrial competitiveness. If this is right, why should the opportunity to utilize safeguard measures turn critically on recent import fluctuations rather than long-term import trends ? One wonders whether the panel here, and the Appellate Body generally, has turned the matter completely on its head. Causation and the ‘ non-attribution ’ problem The panel found fault with the ITC analysis of causation for nine of the ten ‘industries ’ covered by the US safeguard measures. In each instance, it held that the ITC failed to demonstrate a causal link between increased imports and injury, that it failed to ensure that injury caused by other factors was not ‘attributed ’ to imports, or both. Demonstrating the causal link to imports. Again taking its cue from prior decisions, the panel suggested that a causal linkage between increased imports 53 Ibid. "10.160. 54 See 19 USC. ·2251(b).

Definitive Safeguard Measures on Imports of Certain Steel Products 111

and injury might be established in one of two ways: through a ‘coincidence ’ analysis, or through an analysis of the conditions of competition. A coincidence analysis examines the ‘temporal relationship between the movements in imports and the movements in injury factors’.55 Such coincidence is ‘ normally’ evident ‘ if causation is present ’,56 although the suggestion that a temporal lag may exist between import increases and injury ‘ may have merit in certain cases ’.57 Where a clear coincidence exists, ‘ no further analysis is required of the competent authority ’, save for a careful non-attribution analysis.58 Where coincidence is lacking or an analysis of coincidence has not been undertaken, the competent authority must explain its absence and must show causation convincingly through other means. According to the panel, an analysis of the conditions of competition requires the competent authority to consider the factors enumerated in Article 4.2(a) of the Safeguards Agreement : changes in import volume, import market share, domestic sales, production, productivity, capacity utilization, profits and losses, and employment. Other unenumerated factors may also be relevant.59 Further, ‘ price _ in the Panel’s view, is an important, if not the most important, factor in analysing the conditions of competition in a particular market _ we consider that relative price trends as between imports and domestic products will often be a good indicator of whether injury is being transmitted to the domestic industry _ given that price changes have an immediate effect on profitability, all other things being equal ’.60 Against this backdrop, the panel proceeded to consider the analysis of the ITC as to each of the challenged measures. In the important category of certain carbon flat-rolled steel (CCFRS), for example, it found : that there was no coincidence between, on the one hand, import trends and the situation of the domestic industry of CCFRS, as reflected in data for production, net commercial sales, productivity and capacity utilization of the domestic CCFRS. We have also found that there was a lack of coincidence between import trends and declines in domestic operating margin _ We did discern coincidence, albeit lagged, between increased imports, on the one hand, and employment, on the other hand _ Having taken into consideration all of the foregoing, in the Panel’s view, overall, coincidence did not exist. ’61 ‘Given a lack of coincidence between import trends and the injury factors, it was for USITC to provide a compelling explanation as to why a causal link was considered, nevertheless, to exist.62 55 Panel Rep. "10.299. 56 Ibid. ·10.300. 57 Ibid. "10.310. 58 Ibid. "10.307. 59 Ibid. ·10.318. 60 Ibid. "10.320. 61 Ibid. ""10.374–375. 62 Ibid. "10.376.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

The ITC’s analysis of the conditions of competition for CCFRS was then found deficient as well. The panel first suggested that the product category might be too broad for such an analysis to be undertaken at all in convincing fashion. Further, the ITC apparently relied heavily on evidence of import underselling and downward price trends for both imports and domestic products for two sub-products in the CCFRS category, without explaining ‘ why pricing data for the other three items that constituted CCFRS were not specifically considered ’. And, ‘ while some of the domestically produced constituent items were undersold by the import counterparts at particular points during the period of investigation, this was not necessarily the case for the entire period of investigation ’.63 Thus, the conditions of competition analysis failed to support the existence of a causal link between increased imports and injury. As to other product lines, the panel found the ITC’s analysis more convincing. For fittings, flanges, and tool joints (FFTJ), the panel examined the relation between imports and indicators of injury, and noted that ‘ clear coincidence exists between the upward trend in imports and the downward trend in the injury factors, except for productivity ’. Because the ITC report had not analyzed this coincidence in detail, however, the panel also found that a condition of competition analysis was required to support the ITC’s finding of a causal link.64 In that regard, the panel reviewed data assembled by the ITC showing that imported products significantly undersold domestic products during the period of investigation, and on that basis concluded that the conditions of competition analysis supported the existence of a causal link. In the case of hot-rolled bar, the panel noted that the ITC had not undertaken a coincidence analysis. But as part of its analysis of the conditions of competition, the ITC assembled data on market penetration by imports along with import and domestic prices. The data were presented in Figure 2. Based on these data, the panel concurred that a causal link was present – ‘ [t]he USITC explained that domestic prices declined in an effort to mitigate the erosion of market share _ On the basis of the foregoing, overall, we find that the USITC’s conditions of competition analysis was compelling. ’65 It would be unfair to fault the panel for following the analytic lead of the Appellate Body and the ITC, but, in doing so, it followed them to the land of economic gibberish. The panel insists that the linchpin in the search for causation is a search for coincidence. The irony of that phrasing is glaring – the Random House English dictionary defines ‘coincidence ’ as ‘ a striking occurrence of two or more events at one time apparently by mere chance ’. The most elementary statistics class teaches that correlation is not causation, and the problem is not ameliorated by relabeling correlation as coincidence. 63 Ibid. "10.379. 64 Ibid. "10.516. 65 Ibid. ""10.429–430.

Definitive Safeguard Measures on Imports of Certain Steel Products 113

Figure 2













0 1996







Imports vs imports unit values and domestic unit values (Tons and $/ton)

0 2001

Years Imports Domestic unit values Domestic unit values (semi-annual)

Imports (semi-annual) Imports unit values Imports unit values (semi-annual)

Further, as explained earlier, the problem is actually much more fundamental. It is not disreputable to examine correlation as an aid to an exploration of causation with appropriate caveats. But one can only do so when one variable is a logical candidate for the cause of the other. Import quantities do not cause anything – they are simultaneously determined along with prices, domestic output, domestic employment, and so on. It makes no more sense to say that increased imports caused a decline in domestic production, for example, than to say the exact opposite. At this point, the panel’s analysis would seem to involve internal inconsistency. On the one hand, it would require the complainant to identify ‘ unforeseen developments ’ that are linked to increased imports which in turn have caused serious injury or threat. But, on the other hand, it would prescribe a separate causation analysis in which the unforeseen developments play no role. If the panel’s requirements for establishing the existence of unforeseen developments have been met – namely, that the domestic industry experienced exogenous shocks that caused import volume to rise and conditions to deteriorate – must not the requirements for causality have been met as well ? What additional information can the examination of coincidence provide ? The ‘ conditions of competition ’ analysis that serves as an alternative to coincidence analysis is no more comforting. To the extent that the panel, like the ITC, finds evidence of import ‘underselling ’ to be persuasive evidence of causation, an economist would respond that persistent underselling by imports is simply evidence that they are of lower perceived quality for some reason. It says nothing


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

about a causal link between anything and anything else. Likewise, to the degree that a high degree of correlation exists between the price series for imported and domestic goods, that fact is some evidence that the goods are reasonably close substitutes in consumption. Again, no inference of ‘ causation’ is supported, as indeed there is no intelligible causal variable under examination. To be sure, cases will arise in which ‘ coincidence ’ is relatively stronger or weaker. Cases will arise in which imported goods undersell domestic goods, and many cases will arise in which import and domestic price trends are highly correlated. It will thus be possible for importing nations to demonstrate ‘ causation’ with some regularity using the tests that the panel applies. But if any relationship exists between that set of cases, and the set of cases in which safeguard measures are appropriate on some principled basis, it will arise only by ‘coincidence ’. The non-attribution problem. Following the lead of the Appellate Body, the panel makes clear that the presence of a ‘causal link ’ between imports and injury, established as above, is not enough to satisfy the requirements of the Safeguards Agreement. If other factors have also contributed to injury, the ‘ competent authorities must separate and distinguish the injurious effect of the increased imports from the injurious effects of the other factors ’.66 This exercise is required even though imports need not be solely responsible for injury, but must merely have contributed to it. A proper non-attribution analysis also determines the permissible scope of the safeguard remedy – the Appellate Body had ruled in United States–Line Pipe that safeguard measures may only remedy the injury attributable to increased imports, not that attributable to other factors.67 For a number of product lines, the panel found the ITC’s non-attribution analysis to be lacking. The panel’s approach, as in previous cases, was simply to accept at face value the ‘ other factors ’ put forth by the respondents at the ITC, with no discussion as to how or why they are appropriate or inappropriate. The panel would then check to see whether the ITC had confidently distinguished the injury attributable to the factor in question, and ensured that such injury was not attributed to imports. We offer one illustrative example of the analysis : In the case of hot-rolled bar, as noted above, the panel accepted the ITC’s analysis of the conditions of competition as a basis for finding a causal link between increased imports and injury. But the respondents argued before the ITC that injury was caused, inter alia, by increased input costs for domestic producers, what the panel termed increases in the costs of goods sold (COGS). The ITC acknowledged that COGS had risen during part of the period of investigation, but argued that import competition had suppressed prices and prevented domestic firms from recouping their higher costs. On that basis, the ITC concluded that imports were the more important cause of injury.

66 Ibid. "10.329. 67 Ibid. "10.338.

Definitive Safeguard Measures on Imports of Certain Steel Products 115

The panel evidently considered this analysis too cursory, thus falling short of a ‘ reasoned and adequate ’ explanation. The panel hinted, however, that if the ITC had gone further in its analysis, it might have been able to defend its conclusion. In particular, the panel noted that there was a general lack of ‘coincidence ’ between changes in COGS and operating margins for domestic producers. Had changes in COGS ‘ played a significant role in the situation of the domestic industry, one would have expected operating margins to increase while COGS was decreasing ’.68 The panel’s discussion of the non-attribution requirement suffers from the same logical flaws as the treatment of the issue in prior cases. Its analysis of the COGS factor for hot-rolled bar illustrates the fundamental problem. An increase in input costs for domestic firms will lead them to institute price increases if they can. But, as may have been the case in the hot-rolled bar market, import competition may prevent such price increases. The price increases that are attempted by domestic firms, caused by rising input prices, may thus be the cause of greater import volumes, which restrain price increases and leave domestic firms in a weakened financial situation. The panel insists, however, that the injury attributed to rising COGS must be distinguished from the injury attributed to imports. As the above discussion makes clear, this task is logically impossible. The imports themselves result from increases in COGS, and so how can the effects of the two possibly be distinguished ? Putting it differently, the causal variable in this scenario is an increase in domestic input costs. The result is both an increase in imports and a weakened financial situation for domestic firms. For the same reason, the proposition that a permissible safeguard measure can address the injury caused by increased imports, but not the injury caused by rising COGS, is also fundamentally incoherent. The panel’s suggestion that one can analyze the importance of COGS by looking at the coincidence between changes in COGS and operating margins is silly. Other things being equal, increases in COGS will surely tend to lower operating margins, but many other factors in the market are variable over time, and the absence of a ‘ coincidence ’ between changes in COGS and operating margins simply indicates that other things are happening simultaneously – a clear demonstration of why ‘ coincidence ’ and causation are two different things. Obviously, the presence of other factors varying simultaneously cannot negate the fact that increases in input costs, other things being equal, are disadvantageous for domestic producers. The ultimate issue here is a simple one – should safeguard measures be permitted when the cause of injury to a domestic industry, and the cause of rising imports, is a shock to the cost structure of domestic firms? There may be good reasons to answer this question yes or no, but the analysis of the panel merely masks and confuses it.

68 Ibid. "10.440.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

Parallelism The ‘ parallelism ’ requirement stems from Argentina–Footwear. The Appellate Body there held that a correspondence must exist between the imports included in the analysis that led to the injury determination, and the imports covered by the safeguard measure.69 Thus, for example, if imports from Canada were part of the data on which the injury finding rests, imports from Canada cannot be exempted from any subsequent safeguards remedy. The legal basis for the parallelism requirement is shaky. Pauwelyn (2004) criticizes the requirement and argues forcefully that the real issue is whether Article XXIV of GATT, pertaining to the formation of customs unions and free trade areas, either requires or permits members of such entities to exempt imports from other members from safeguard measures. This question thus far has no clear answer. Whatever its merits, the parallelism requirement is established in the cases, and the United States undertook to argue that it had complied. The original ITC determination had aggregated imports from all sources, however, while the eventual safeguard measures had largely excluded imports from NAFTA countries, Israel and Jordan. The ITC was asked to revise its analysis to exclude these imports in its supplemental report to USTR. It did so, and reached the same conclusions for each industry. The panel took issue with the analysis for several reasons. In several instances, the ITC had not made clear that it had properly excluded imports from Israel and Jordan in its revised analysis. The ITC also failed to explain to the panel’s satisfaction why its findings remained the same despite the fact that a smaller quantity of imports was involved after the parallelism adjustments. In addition, the panel held that the ITC was obliged to repeat its non-attribution analysis based on the revised import totals, and that it had failed to do so in the supplemental report. In the interest of parallelism, therefore, the panel would have the ITC revise its import data, and thereafter undertake the same conceptually flawed analyses of coincidence, conditions of competition, and non-attribution. The excluded imports would become an ‘ other factor ’ to which injury from the included imports could not be attributed. Such exercises are no more valuable with the revised data than with the original. Until a logically sound approach to the question of causation emerges, the requirement of parallelism is just a sideshow. It is also somewhat peculiar that neither the panel nor the prior pertinent decisions draw any connection between the parallelism issue and the unforeseen developments issue. If the United States is obliged by Article XIX to draw a connection between injury and developments in Asia and the former Soviet Union, is it not possible that such injury was transmitted through an effect on imports from particular sources, rather than an effect on all imports or on world prices? Does the answer to that question have any implications for the permissible scope 69 Ibid. ""10.590–591.

Definitive Safeguard Measures on Imports of Certain Steel Products 117

of the safeguard remedy, or for the possible exclusion of imports from particular sources ?

3.2 The Appellate Body decision The Appellate Body affirmed the panel in most pertinent respects, or exercised judicial economy to avoid reaching the issues raised.70 We can thus address its decision with considerable brevity. Unforeseen developments Much of the US appeal on this issue consisted of a challenge to the ‘standard of review ’ employed by the panel – a requirement that the US competent authorities provide a ‘ reasoned and adequate ’ explanation for their findings. The Appellate Body essentially affirmed the standard employed by the panel, and we have no quarrel with the general principle that national authorities should set forth findings with enough clarity and logical reasoning for reviewing panels to be able to assess them. The Appellate Body also concurred with the panel that a member proposing to invoke safeguards must demonstrate that unforeseen developments have led to increased imports for each ‘ industry’ covered by a safeguard measure.71 Here too it affirmed the finding that the ITC had failed to make such a showing with a ‘ reasoned and adequate’ explanation.72 We have no quarrel with this principle either as a general matter. The treaty text requires the preconditions for safeguards to be met for any ‘ product’ covered by a safeguard measure. When, as in the steel case, national authorities determine that a number of different ‘ products’ are involved and proceed to define a number of separate ‘industries ’ to investigate, it is appropriate to require that the preconditions for safeguard measures be met in each industry. But the Appellate Body opinion does not speak to the deeper issues raised by the unforeseen developments requirement. It expressly states that it offers no ruling on the question whether the developments identified by the United States – the Asian financial crisis, the dissolution of the Soviet Union, and so on – actually constitute ‘ unforeseen developments’ as a legal matter.73 It thus offers no guidance on how one determines what is ‘ unforeseen ’, or what is permitted to ‘ count ’ as an unforeseen development. Most importantly, it does not address the ultimate issue, raised implicitly by the panel’s analysis, as to whether national authorities must convincingly link all of the ‘serious injury ’ caused by ‘ increased imports ’ to the underlying ‘ unforeseen developments ’. If indeed they must, then the requisite analysis becomes somewhat better grounded in a coherent economic theory on the 70 United States – Definitive Safeguard Measures on Imports of Certain Steel Products, WT/DS24849, 251-54, 258-59/AB/R (November, 2003) (hereafter App. Body Rep.). 71 Ibid. "319. 72 Ibid. "326. 73 Ibid. "269.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

one hand, but the task of producing a ‘ reasoned and adequate ’ justification for a safeguard measure becomes all the more daunting on the other. Increased imports Here, the Appellate Body reaffirmed the principle set forth in Argentina–Footwear that a mere increase in imports is not enough to satisfy the requirement of ‘such increased quantities ’ as to cause injury. It quoted with approval its prior reference to a requirement that ‘the increase in imports must have been recent enough, sudden enough, sharp enough, and significant enough, both quantitatively and qualitatively, to cause or threaten to cause ‘‘serious injury’’’,74 and found that the panel had interpreted this standard correctly. In so doing, the Appellate Body did nothing in our view to clarify the practical application of the standard. Likewise, the decision does little to explain the logic behind the standard – why do the consequences of ‘ unforeseen developments ’ not unfold slowly at times, so that the resulting increase in imports is not particularly sudden or sharp ? And, at the risk of seeming repetitive, what sense does it make to fret about the time period over which increased imports may or may not have caused injury, when the law offers no coherent theory of how imports cause anything? On these fundamental conceptual issues relating to the increased imports requirement, the marginal contribution of the opinion is nil. Regarding the panel’s findings with respect to specific product categories, the Appellate Body for the most part affirmed the panel’s findings to the extent that they were appealed by the United States. We critiqued the panel’s analysis at some length above on these issues, and will not repeat the discussion here. The Appellate Body did reverse the panel’s findings with regard to tin mill products and stainless steel wire. Contrary to the panel’s conclusion, the Appellate Body held that the findings of different Commissioners, who had defined the industries in varying ways, could in principle suffice as a ‘ reasoned and adequate ’ explanation for a finding of increased imports.75 Nothing in the treaty text precludes an aggregation of judgments in this fashion, nor is it logically inconceivable that analyses based on different conceptions of the ‘ industry ’ might nevertheless justify a finding of increased imports. Having reversed the panel on this point, however, the Appellate Body found it unnecessary to complete the panel’s analysis because it held the safeguard measures for tin mill products and stainless steel wire to be illegal on other grounds.76 The reversal of the panel here may be of some modest comfort to the United States, with its history of ITC Commissioners who often arrive at different definitions of the relevant ‘ industry ’ in safeguards proceedings, but seems to be of modest conceptual importance. In sum, as with its treatment of the unforeseen developments requirement, the Appellate Body decision adds nothing of significance to its previous decisions 74 Ibid. ""345–346. 75 Ibid. ""416, 429. 76 Ibid. "431.

Definitive Safeguard Measures on Imports of Certain Steel Products 119

beyond tacit approval of the analysis conducted by the panel on several of the product categories. Our thoughts on that analysis are set forth above. Parallelism The Appellate Body agreed with the panel that the ITC had not shown that it had properly excluded imports from Canada, Mexico, Israel, and Jordan in reaching its findings. It emphasized especially that the ITC had not considered the excluded imports as an ‘ other factor ’ in a proper non-attribution analysis.77 We have no quarrel with the factual proposition that the ITC’s reasoning on this front was murky, although the task of fixing the problem remains confounded by the absence of any coherent way to identify the impact of the relevant set of ‘ increased imports ’, or to separate the injury caused thereby from injury caused by other factors. Until a conceptual framework for identifying the relevant exogenous variables and assessing their impact is articulated, it is simply not clear how to implement a requirement of parallelism. And absent an intelligible theory as to when safeguard measures are appropriate and when they are not, it is impossible even to say whether a ‘ parallelism ’ requirement makes sense. Causation Because its rulings on the unforeseen developments, increased imports and parallelism issues sufficed for finding that each of the ten challenged measures violated WTO law, the Appellate Body declined to consider the panel’s analysis of the causation issue. It simply referred the parties to its prior rulings for further ‘ guidance’, rulings with which we have already taken issue. In sum, unlike its prior rulings in the Safeguards area, the Appellate Body ruling in the steel dispute offers virtually nothing of conceptual importance. To the extent that any important new law was made, it consisted of affirming the reasoning of the panel on certain key points that have already been discussed extensively above. Nothing in the opinion resolves any of the conundrums raised by prior Appellate Body decisions.

4. Concluding commentary Our review of the steel dispute suggests how difficult it will be for WTO members to use safeguards going forward without a prospect of near certain defeat when a complaint is brought against them. Members must demonstrate the existence of unanticipated developments, persuade that they were ‘ unforeseen ’, convincingly trace their impact on increased imports, demonstrate that much of the import surge is sufficiently ‘ recent ’, convincingly show the relation between the imports

77 Ibid. "456.


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

and serious injury to an appropriately defined ‘ industry ’, and convincingly show that ‘other factors ’ have not caused the injury attributed to increased imports. And they must accomplish these things in a theoretical vacuum, where neither the treaty text nor the Appellate Body decisions to date offer any clear explanation of what it even means to say that ‘ increased imports ’ have ‘ caused ’ or threatened to cause serious injury, how one identifies the ‘other factors ’ that might be responsible for injury, and how one attributes injury among its various causes. Because it is unclear what the law requires as a conceptual matter, it is exceedingly difficult to comply with it, and one can hardly fault national authorities for their inability to offer a ‘ reasoned and adequate’ demonstration of their fidelity to the law. To be sure, some observers may welcome these developments. For those who believe that safeguard measures are nothing more than wasteful protectionism, insurmountable hurdles to their use will have appeal. But the literature is in fact rather agnostic on the use of safeguards. Bagwell and Staiger (1990, 2002) suggest that safeguard measures may be understood as a device for reducing the pressure on nations to cheat on trade agreements, and thus can reduce the danger that the agreements may unravel. Sykes (1991) argues that the opportunity to employ safeguard measures ex post may facilitate more trade concessions ex ante in an environment where negotiators face political uncertainty about the consequences of their trade concessions. And the experience of the GATT suggests that if safeguard measures become unavailable, nations may return to arrangements such as voluntary export restraints that are even worse from the ex post standpoint because no standard restricts their use or limits their duration – the political pressures to protect troubled industries through trade policy will remain regardless of the state of the law, as will their potential to cause mischief. It is thus plausible that safeguard measures have a constructive role to play in the trading system. If they are to play any role at all in going forward, however, the law must evolve in a way that makes clear to WTO members what circumstances are appropriate for safeguards, and how to go about demonstrating the existence of those circumstances. For the reasons given above, the treaty text is woefully deficient in this regard, and the decisions of the Appellate Body have only compounded the problem. One way or another, what is need is a fresh start. We can imagine it coming in two ways. First, the Appellate Body might change course dramatically, and initiate a ‘ common law ’ evolution toward coherent standards for the use of safeguards, much as US courts have done in fleshing out the vague and imprecise standards of the US antitrust laws. To do so, however, the Appellate Body would have to abandon its insistence on grounding every principle in treaty text, in favor of ascertaining the ‘ object and purpose ’ of the Safeguards Agreement from other sources, perhaps even by some direct appeal to economic theory. Such a process would represent a major departure from the Appellate Body’s usual approach to cases, however, and many might question the legitimacy of such a departure. Alternatively, the WTO membership might simply renegotiate the Agreement on

Definitive Safeguard Measures on Imports of Certain Steel Products 121

Safeguards, with an eye toward resolving the fundamental issues that we have identified (and no doubt some that we have not). At this writing, neither possibility appears terribly likely in the foreseeable future. Whatever the near term likelihood of legal reform, the economic literature does offer some ideas on possible directions for change. A number of economic scholars advocate an interpretation of the legal prerequisites for safeguards that melds easily with basic price theory, an approach that actually surfaced briefly in the reasoning of a few ITC Commissioners.78 Grossman (1986), Kelly (1988), and Irwin (2003) would divide the potential causes of injury into three groups: forces that cause shifts in the domestic supply schedule, forces that cause shifts in the domestic demand schedule, and forces that cause shifts in the import supply schedule. Any harm to the domestic industry that can be attributed to shifts in the import supply curve will be deemed to result from ‘increased imports ’ ; any harm attributable to rising domestic costs that shift the domestic supply schedule will be deemed to result from causes other than increased imports. Likewise, harm due to shifts in domestic demand will be attributed to causes other than imports, unless the shift in demand is due to a price reduction on imperfectly substitutable imports. This approach also has the great virtue of economic coherence, shifting the inquiry to ask whether changing conditions of import supply, rather than increased quantities of imports, are causally responsible for injury. It does not, however, address the problem of what counts as an ‘ unforeseen development ’, or whether such a requirement is desirable and appropriate. And we do not pretend that it would always be easy to implement as a practical matter. The data requirements for confident estimation of the pertinent supply and demand relationships may often be lacking, and the task of identifying and specifying those relationships can be controversial. Critics may also argue that the import supply approach is not the only coherent way to implement a safeguards system, and may preclude safeguard measures in cases where WTO members might agree they ought to be permissible. The old GATT Hatter’s Fur case noted above is instructive in this regard. The US position in that case was that a decline in domestic demand for the types of hats produced by domestic firms was the ‘ unforeseen development ’ that resulted in an import surge and that justified safeguards action. The working party appeared to accept this theory, at least in principle, while quibbling as to whether the facts supported it. Sykes (2003) suggests a slightly different set of principles. Noting that the original structure of Article XIX was aimed at protecting against the ‘unforeseen ’ consequences of trade concessions that might produce import surges, he suggests that safeguards may be appropriate when unanticipated shocks lead exporters to enjoy unanticipated prosperity, while import-competing firms simultaneously 78 See US International Trade Commission, Wood Shakes and Shingles, Inv. No. TA-201-56, Pub. No. 1826 (1986)(views of Commissioners Liebeler and Brunsdale).


G E N E M. G R O S S M A N A N D A L A N O. S Y K E S

suffer severe and unexpected declines. Such an approach could permit a safeguard measure on the facts of the Hatter’s Fur case, for example. This approach too leaves many details to be worked out, most especially the issue of how to determine whether shocks are unanticipated, and is no more than an initial step toward articulating the role of safeguards in the system. Our goal here is not to opine definitively on the proper logic of safeguards. That task, of course, is a matter for the WTO membership. We simply suggest that economically oriented scholars may offer some useful ideas for the reform of safeguards law, and that they can at least help to suggest what sort of framework will stand the test of logical coherence. Current WTO jurisprudence flunks that test miserably, and without reform will continue to present insuperable obstacles to the use of safeguard measures. We are unsure about the systemic consequences of this state of affairs, but are by no means confident that the consequences are benign.

References Bagwell, Kyle and Robert W. Staiger (1990), ‘ A Theory of Managed Trade’, American Economic Review, 80: 779–795. Bagwell, Kyle and Robert W. Staiger (2002), The Economics of the World Trading System, Cambridge, MA: The MIT Press. Grossman, Gene M. (1986), ‘Imports as a Cause of Injury: The Case of the US Steel Industry’, Journal of International Economics, 20: 201–222. Grossman, Gene M. and Petros C. Mavroidis (2004), ‘United States – Definitive Safeguard Measures on Imports of Circular Welded Carbon Quality Line Pipe from Korea’, in H. Horn and P. C. Mavroidis (eds.), The WTO Case Law of 2001, American Law Institute Reporters Series, Cambridge: Cambridge University Press. Horn, Henrik and Petros C. Mavroidis (2003), ‘ United States – Safeguard Measures on Imports of Fresh, Chilled or Frozen Lamb Meat from New Zealand and Australia: What Should be Required of a Safeguard Investigation? ’, in H. Horn and P. C. Mavroidis (eds.), The WTO Case Law of 2001, American Law Institute Reporters Series, Cambridge: Cambridge University Press. Irwin, Douglas (2003), ‘Causing Problems? The WTO Review of Causation and Injury Attribution in US Section 201 Cases’, World Trade Review, 2: 297–325. Kelly, Kenneth (1988), ‘ The Analysis of Causality in Escape Clause Cases’, Journal of Industrial Economics, 37: 187–207. McGovern, Edmond (1986), International Trade Regulation, 2nd edition, London: Gobefield. Pauwelyn, Joost (2004), ‘ The Puzzle of WTO Safeguards and Regional Trade Agreements’, Journal of International Economic Law, 7 : 109–142. Sykes, Alan O. (1991), ‘Protectionism as a ‘Safeguard’ : A Positive Analysis of the GATT ‘Escape Clause’ with Normative Speculations’, University of Chicago Law Review, 58: 255–305. Sykes, Alan O. (2003), ‘ The Safeguards Mess: A Critique of Appellate Body Jurisprudence ’, World Trade Review, 2 : 261–295. Sykes, Alan O. (2004), ‘The Persistent Puzzles of Safeguards : Lessons from the Steel Dispute’, Journal of International Economic Law, 7 : 523–564.

World Trade Review (2007), 6 : 1, 123–133 Printed in the United Kingdom f Mathias Risse doi :10.1017/S1474745606003119

Review essay

Economic justice in an unfair world : towards a level playing field By Ethan B. Kapstein Princeton: Princeton University Press, 2006

Kapstein begins by stating that ‘If any norm characterizes international economic relations, it is probably that they should be carried out on a ‘‘ level playing field’’’ (p. xi). He then sets out to offer his vision of what such leveling would amount to. The book offers what he calls a ‘liberal internationalist’ perspective on global economic justice, a view defined in opposition to two other views introduced and explored throughout: communitarianism, on the one hand, which according to him takes states too seriously by denying, beyond a certain minimal level, obligations to those who do not belong to them; and cosmopolitanism, on the other hand, which he thinks does not take states seriously enough by failing to acknowledge, to a sufficient degree, the special moral importance of shared citizenship. By way of contrast, liberal internationalism takes states as given, but acknowledges that there are moral obligations tying together states into a society of states, a society that, according to Kapstein’s ideal, is supposed to be inclusive, participatory, and welfare-enhancing for all states. What makes such a society of states possible, according to Kapstein, is the core insight of international economics : the mutually beneficial nature of free trade. A society of states with the three features Kapstein champions can be created if the trading system is reformed accordingly. It is through the proper involvement of all countries (an involvement that takes seriously that all those states are sovereign and in that sense morally equal) that the trade system becomes participatory and inclusive. The moral equality of states is supposed to shape negotiations and agreements. And it is through trading itself, and through increasing trade liberalization, that each country will benefit and improve its overall welfare. Possibly, with some assistance, all countries can discover their comparative advantage, use it to integrate themselves into the society of states and from then on shape their economic destiny through policy choices. This society of states is committed to the goals of economic growth and convergence in the long run. Even powerful states are said to have an interest in maintaining such a system because of its robustness, presumably because it provides political stability and creates economic opportunities in ways in which possible alternatives (which would deny equal standing to less powerful states) could not. The vision of global economic justice Kapstein proposes happens to be in every state’s enlightened self-interest anyway. As far as the creation of this society of states is concerned, prudence and morality coincide. To wit: the main thrust of my criticism later on will be that this picture is overly simplistic. When it comes to global fairness, self-interest and morality do not always coincide, at any rate not in such a way that this could become the guiding idea of global economic justice. Kapstein does not just think of this entirely as a vision for the future ; he believes the more powerful states are already doing their part in creating such a system (p. 62), 123



and finds that ideas of fairness have ‘animated the agreements in ways that have tempered mercantilist dictates at least to some limited degree’ (p. 64). With reference to the special and differential treatment afforded to developing nations throughout the trade negotiations in recent decades, Kapstein thinks ‘it is notable that the trade regime has recognized something like a Rawlsian ‘‘difference principle among nations’’ ’ (p. 71). This principle, recall, requires (domestically) that social and economic inequalities, to the extent that they are even consistent with an equal measure of basic liberties for everybody and with an ideal of substantive equality of opportunity, must be regulated in such a way that they work to the benefit of the least advantaged in society. Rawls himself, to the chagrin of some of his defenders, rejected the extension of this principle to the international order. And to put it mildly, it goes against the mainstream of both the literature on globalization and the literature on global justice to claim that anything approximating the difference principle is in effect internationally. There is a vast difference between giving special and differential treatment to developing countries, and designing the global political and economic order in such a way that the position of the respectively least-advantaged members is as good as it could be. Be that as it may, trade liberalization, so Kapstein believes, can solve the world’s problems. He is aware that some people, and some countries as a whole, will lose out in the process (subsidies, after all, do benefit some), but gains from trade can be used to compensate them. The goal of global economic justice is to ‘provide the background conditions that make it possible for all states to exercise their comparative advantage and achieve sustained growth’ (p. 22). The WTO itself is supposed to be organized not around plain reciprocity, where each state concedes as much as it gains, but around so-called relaxed or diffuse reciprocity, where the smaller participants are expected to concede less than the bigger ones. Diffuse reciprocity is applied especially to AIDS : ‘The society of states should confront the AIDS problem and others like it on the basis of diffuse reciprocity, with each member contributing resources in order to help the poorest states meet the health challenges they face.’ Again we are told that ‘[s]uch a policy is both moral and self-interested’ (p. 78). According to Kapstein, it is through a ‘ combination of free trade and foreign aid’ (p. 42) that disadvantageous circumstances in which states may find themselves can be offset – and from then on it is up to domestic policies to determine their economic fate. Kapstein does indeed express strong confidence both in the transformative potential of free trade and in the effectiveness of foreign aid. Nevertheless, again, he is aware that there will be losers from trade. For them, there need to be safety nets, primarily domestically, but, if those fail, the international community would have to step in (p. 49). The goal of aid is to ‘ assist countries that are seeking to exercise their comparative advantage and enter the global trading system, by focusing on infrastructure support, capacity building, the knitting of social safety nets, and balanceof-payments support ’ (p. 112). Kapstein adds two chapters on migration, labor laws, and direct foreign investment. Throughout, the concern is to seek for arrangements that are both moral and prudent, and throughout, Kapstein continues to be directed by his confidence that such arrangements are possible. I think this approach seriously oversimplifies problems of global economic justice, and I will make this point for the case of trade only. Focusing on trade seems appropriate anyway, since the guiding



thought in Kapsteins’ book is that trade liberalization can be employed to everybody’s advantage.1 Making trade liberalization the cornerstone of an account of global fairness, as Kapstein does, encounters a number of serious problems, political, economic, and philosophical. It is the presence of these problems that stands in tension with Kapstein’s confidence that, at the level of global economic justice, prudence and morality coincide. In the remainder of this review, I will spell out these problems. Consider the political problems first. Kapstein’s approach is meant to be pragmatic. While he does not think that his proposals will be immediately acceptable to each and any government, he means for them to be acceptable without requiring changes in the basic political arrangements of the global order. This means in particular that the legitimacy of states per se is not questioned. And this, by itself, is fair enough : no single book on global justice can address all the questions that arise about it, one of which is, of course, the question of the basic legitimacy of states in light of the creasing global political and economic interconnectedness. However, since Kapstein’s approach is indeed driven by this sort of pragmatism, one must wonder why there is not more reflection on countries as political actors. Kapstein’s view is that trade liberalization creates gains, and that ongoing trade liberalization, in connection with appropriate aid, will eventually put every country in a position to benefit from trade. Yet regardless of whether countries as a whole benefit from increasing trade liberalization, at least in the short run there will be those whose relative economic positions benefit and those whose fortunes suffer. Kapstein, of course, knows this, and proposes that countries that win on balance are supposed to redistribute some of their gains to those who lose out in these countries themselves, but also to other countries that, on the whole, lose out in the process. To illustrate, suppose the US, EU, and Japan agree to eliminate agricultural subsidies. This would cause serious problems for agricultural producers in these countries (while benefiting consumers there), but also for consumers in net-food-importing developing countries. Subsidies for food-exporting producers are ipso facto subsidies for foodimporting consumers elsewhere. Most of the world’s least developed countries are net-food-importers. According to Kapstein’s proposal, the government of the US, the EU, and Japan would have to raise funds from their consumers (the main beneficiaries from the changes) to support the producers in ways that are non-trade-distorting, as well as to support consumers elsewhere to survive and perhaps also to find ways of turning the new parameters shaping the new parameters on the food market to their advantage. (Governments of net-food-exporting developing countries would presumably have to raise funds as well.) This is a splendid vision, but it glosses over a legion of political problems. First, the gains from trade liberalization do not arise centrally in the bank accounts of an international agency, but in a highly decentralized manner. They arise in the pockets of consumers in developed countries as well as in the accounts of exporters in developing

1 My own views on fairness in trade are developed in Risse (forthcoming), and throughout I will draw on the views developed there.



countries. Or they arise because now markets are organized around different sets of incentives, which motivates many producers, and potential producers, to reconsider what to produce, how much to produce, and where to sell it. Who does what will change in response to the new trade parameters, and it is hard to predict just what the shape of these changes will be (just as we did not know that the Multi-Fiber Agreement was effectively subsidizing certain countries against competition from China because China had never had the chance of exploring its potential on the textile market, prior to the abolition of that Agreement). That is, many of those who will benefit from the changes will do so in a rather indirect way, in a manner that will often not make it obvious to them that they are the beneficiaries of a specific set of measures from which a specific, straightforwardly identifiable group of other people suffer harm – people who at any rate would have no other claim to such redistribution than that they suffered harm from a governmental measure with which nothing may be considered wrong per se. The median voter might not easily be persuaded of the wisdom of such transfers. If, for instance, the German government were to come into the possession of additional funds because the EU budget would no longer have to cover agricultural subsidies, there would be political pressure to invest the money into the deeply troubled pension system. If the EU itself would be the financial beneficiary of such cuts, there would be political pressure to invest it into its own least-developed countries in Eastern Europe so as to ease tensions that threaten to undermine the sheer functionality of the EU. But this is not the biggest political problem that arises here. After all, in many developed countries, governments would at least be sufficiently accountable and effectively organized to see through such transfers, were they able to persuade their voters of the moral importance and prudence of such transfers. However, this cannot be taken for granted for many of the developing countries. One would hope, in particular, that African countries will be the beneficiaries of trade reforms, primarily in the sense that African countries would be given more incentives to participate in world markets. But many of those countries for which Kapstein’s proposal would have us pursue that goal do not have governments that are accountable to their populations, nor are they ostensibly committed to improving the general well-being of their populations. Instead, they are ruling elites who would pocket whatever immediate gains come their way. Moreover, economic development in the short run has a tendency to keep in power whoever happens to be in power at the time ; but in the long run it tends to promote democracy.2 While that may indeed be in the overall interest of any given country, it is not in the interest of any given ruling elite. Perhaps all these problems are solvable, but they are serious, and they make it difficult to think of trade liberalization as straightforwardly mutually beneficial – it is so, at best, if certain rather severe political problems can be solved. (There is, however, more to come by way of problems.) These political problems arise because states are not unitary actors, but entities that themselves are composed of groups with rather 2 Londregan and Poole (1996) suggests that, in the short run growth supports whoever is in power, whereas in the long run, increasing income has a small democratizing effect. Przeworski et al. (2000) are more reserved : they do not find that growing income makes democracies more likely to arise, but that democracies, regardless of how they have arisen, are more stable in wealthy countries once in existence (cf. chapter 2).



divergent interests. A philosophical approach to global justice concerned with the kind of world that we should aspire to obtain eventually, or with purely conceptual issues, might not have to engage much with these problems. (Again, one cannot solve all the world’s problems at once.) But more reflection on these issues would indeed have been desirable in an explicitly pragmatically minded account such as Kapstein’s. Let us continue with the economic problems. Trade and foreign aid, so Kapstein thinks, will offset possibly disadvantaged circumstances in which countries might find themselves. The vision here is that just about any country has something to contribute if it can only figure out what its comparative advantage is, and while aid may be needed to put it in a competitive situation to begin with, there will be a stage where it is because of trade openness that countries can determine their economic fate themselves through their policy choices. While hardly anybody would deny that trade is in some sense ‘important ’ for development, not everybody would agree that additional trade liberalization is as central to improving the situation of the poor as Kapstein thinks. In fact, this question features prominently in an ongoing debate about the sources of economic growth. One standpoint in that debate is the institutional stance – the view that what is central to economic prosperity is the quality of domestic institutions. Of course, there is no contradiction between saying that domestic institutions are key to economic prosperity and pointing out that trade is important for growth as well. However, the institutional stance does put pressure upon thinking of trade liberalization as the panacea that would finally bring about a fair global economic order, and would do so by way of offering a rather straightforward mechanism all states should adopt out of sheer self-interest. Birdsall et al. (2005), for instance, have argued that developing countries already encounter all the openness on the world market they could make use of. That is, those countries would not benefit from further trade liberalization, short of policy reforms. Kapstein could respond that, then, the aid portion of his proposal would have to become effective at this stage: that is, countries that currently cannot benefit from further trade liberalization would have to undergo policy reform, and they would have a claim to assistance with this process. Yet on this subject, Kapstein does not tell us much. What is needed here is specific investigations into just what needs to be done by way of policy reforms, and just how one could go about realizing them. In the background lurks the question of how much could be gained from trade liberalization. Kapstein does not explicitly discuss this matter, but for the purposes of our discussion it will be useful to look at it for a moment. Following Anderson (2004: 550), estimated gains from liberalization range from $254 billion annually ($108 billion for non-OECD countries, in 1995 dollars), to $832 billion ($539 billion for non-OECD countries, 1997 dollars), depending on how the estimates are made. (These are estimates of full trade liberalization, not merely by rich countries, which primarily benefits their consumers.) Anderson and Martin (2006b) say that ‘[f]reeing all merchandise trade and eliminating agricultural subsidies are estimated to boost global welfare by nearly $300 billion a year by 2015. Additional gains would come from whatever productivity effects that reform would generate ’ (p. 11). They add that 45% of the gains would go developing countries, which would be above their one-fifth share of global GDP.



However, such estimates vary widely with the predictive model used. Hertel and Keeney (2006) estimate that eliminating agricultural subsidies and liberalizing trade in goods and services would lead to gains of $151 billion, $34 billion for developing countries. But while that is still a sizable sum, it is by magnitudes smaller than the maximal estimates. These differences arise because the effects of trade liberalization are not currently observable. Estimates depend on assumptions about what would happen once subsidies are removed and a new constellation on the world market is being created. If we take this institutional stance seriously (as I think we must), we must be skeptical about an approach to global economic justice that thinks of trade liberalization as the royal road to a state of affairs characterized by more global economic justice. Some, who have pointed out that the importance of removing especially agricultural subsidies has been exaggerated (e.g., Birdsall et al. 2005), assert that other development programs would produce significantly higher gains (especially a temporary work permit program). Again, then, we are led to conclude that, in an approach that is so explicitly pragmatic about how to obtain more economic justice, such matters should have received more attention. At the very least, these issues make Kapstein’s questions a lot more complicated than they appear throughout his book. This now takes us to the philosophical problems. Recall the first sentence of Kapstein’s book: ‘If any norm characterizes international economic relations, it is probably that they should be carried out on a ‘‘ level playing field’’. ’ ‘Leveling the playing field’ is a metaphor taken from team sports: such leveling is in order to make sure that, if the goal is to place a ball somewhere, it is not the case that one team plays uphill and the other downhill. After all, the point is to put certain skills against each other, and therefore the outcome of the game should be an actual function of these skills. It is because of this obvious appeal in the domain of team sports that, once the metaphor is taken and applied to other domains, it commands instant consent: of course, whatever is relevantly like ‘leveling the playing field’ in sports ought to be done in any other domain as well. Yet the disagreement begins when we need to sort out just what that is supposed to amount to for a given domain. In particular, then, we must ask: just what are the features of the international global order that need to be ‘ leveled’ in the sense in which the playing field for a game of soccer ought to be leveled, and what are the reasons why we should find that as compelling when it comes to the interaction among states as we find the leveling of the playing field for team sports? We can attempt a first answer of these questions by relating this metaphor to the notion of equality of opportunity, more specifically, to one version of that notion. The ideal of equality of opportunity holds great attraction for citizens as well as for theorists across the political spectrum. It offers a vision of a society in which starting positions are regulated just enough for individuals with the relevant prerequisites to move ahead. Such surface harmony, however, covers up serious underlying disagreements. One version of equality of opportunity might merely demand that positions of power and influence be given to those with the relevant qualifications; but another version might demand that there be institutions in society that make sure everybody gets to develop their potential to acquire these qualifications. The former is sometimes called formal equality of opportunity, whereas the latter is called fair or substantive equality



of opportunity. The leveling metaphor can plausibly refer to this second notion of equality of opportunity; this, at any rate, is both intuitive and done by Roemer (1998), to whom Kapstein turns for inspiration. In this way, the metaphor of ‘leveling the playing field’ can be connected to a notion open to analytical treatment. Yet this sets us up for the next question. Just why should it be that, as a matter of justice, the society of states Kapstein envisages should be devised in accordance with such an ideal of equality of opportunity across states ? To see why this is far from obvious, note why domestically one can plausibly argue that substantive equality of opportunity is a demand of justice. The crucial point is that citizens in a state are jointly subject to a coercive structure that regulates access to positions of influence and power, and that regulates especially property holdings. Assuming that the specific shape of such an arrangement is a matter of convention, one can argue that substantive equality of opportunity is a condition on the moral acceptability of any such arrangement. What drives this argument, however, is that the state is a coercive system; it is in light of the kind of pervasive and ongoing interference (or threats thereof that would become effective if, among the myriad of possible actions one has at one disposal every day one happens to choose any actions other than those comparatively rather few that are in accordance with legal norms) with what individuals might do with their bodies and assets that the system must be arranged in certain ways. Regardless of what one thinks about the actual nature and normative relevance of international organizations, the world as such is not quite like that, if for no other reason then because most property arrangements continue to be made domestically (with some notable exceptions), and because equality of opportunity is a notion that is closely tied to such property arrangements, directly or indirectly. Thus on the face of it it is hard to see what the argument is for why the same notion of equality of opportunity should apply within states and between states. Kapstein never provides such an argument, he simply starts with the leveling metaphor and creates a sense that all we need to do is spell out just what this might mean. But we need to do much more : we need to know why we should even think like this about the global order. Let me make the point stronger. At this stage of our reflection on fairness at the global level (that is, after the subject matter has attracted a good deal of attention over a number of years), starting with the ‘ leveling’ metaphor, and thus starting with the assumption that a notion of substantive equality of opportunity applies at the global order is just no longer helpful. There are other views in the literature that find it obviously incorrect that the leveling metaphor applies internationally. Burtless et al. (1998), chapter 5, even claim that talk about fairness in trade is conceptually muddled. For, as they say, fairness itself is indeed tied to the image of ‘leveling the playing field;’ however, ‘leveling ’ suggests a requirement of the equalization of certain conditions, whereas trade thrives on differences. It is because A and B differ in their productive abilities that they trade. So how could ideas about ‘ leveling’ apply here ? On this view, it makes no sense at all to talk about fairness in trade. We cannot proceed beyond this clash of intuitions without further illumination. That Kapstein never provides an argument to back up his starting point is frustrating in particular since he writes that often moral philosophy tends to be ‘ preachy’ rather than ‘ analytical’ (p. xv), but then



himself does not offer the kind of analytical treatment of fairness needed for the questions at hand. The problem is indeed that fairness requires more theorizing than it gets when tied to the leveling metaphor, or even when tied to an analytically more accessible notion of equality of opportunity. Next, I offer an account of fairness I believe is much better suited for the kind of analytical challenges that talk about fairness in the global economic order raises, an account inspired by Broome (1999). Fairness is concerned to ensure people receive what they are owed. I refer to demands people have because they are owed something as stringent claims. Distributions of burdens or benefits are not fair (unfair) merely because they meet (violate) any criterion just mentioned. They are unfair only if they fail to deliver what people are owed. A philanthropist is not unfair if he gives more to one university although the other has bigger needs (neither having a stringent claim). Furthermore, fairness does not aim at satisfying stringent claims per se, but at their proportionate satisfaction. Suppose we are all owed a medication, and the more we take of it, the more we recover. No considerations other than medical need enter (disregard ownership, who is more deserving, etc.), and the needs are equal. Suppose there is not enough to restore everybody completely. Nobody can complain that her claim is not fully satisfied if all are satisfied equally. To develop these ideas, different questions must be answered. First, what are the bases for stringent claims? What are people owed? Second, if stringent claims arise on different bases, how should such bases be compared ? Third, what does it mean to satisfy claims made on different bases respectively ‘in proportion? ’ Much moral theory and conceptual work is required to answer these questions: that notion, speaking figuratively, comes conceptually rather late. To illustrate, consider two sketches of how to answer these questions. Utilitarians may only count well-being as basis for stringent claims ; would argue that what is owed to people is maximization of collective well-being; would hold that the only weighing involved is utilitarian aggregation; and would maintain that claims have been satisfied in proportion if each entered the aggregation in its strength. Fairness talk supervenes on talk of aggregate utility. Deontological views may isolate claims of some urgency as bases of stringent claims, isolating need, desert, and entitlement as such bases ; argue that what is owed to people is satisfaction of these claims ; would then explain how different sorts of urgent claims are compared (e.g., how needs- and entitlement-based claims are compared, possibly assigning some form of priority to one) ; and would offer some view of how claims on different bases can respectively be satisfied in proportion. Fairness now operates in more complex ways than on the utilitarian picture, and often opposes consequentialist considerations. A full-fledged theory involves two accounts of balancing claims: first, an account of how to compare need-, desert-, and entitlement-based claims, and, second, how to satisfy claims within each category proportionately. The latter exercise only arises at a rather practical level. Suppose individuals have entitlement-based claims to an indivisible object, and no other bases of claims matter. One way of thinking of proportionality is to hand the object to individuals for time periods proportionate to the strength of their claim. Such issues do not arise when we talk about fairness of the global economic order, however, because of the relatively abstract



nature of such matters; so I mention them here only for the sake of spelling out my account of fairness. With this account of fairness in place, let us briefly look back at the notion of substantive equality of opportunity. Sometimes it will be the case that what is owed to people is such equality of opportunity; for instance, above I sketched an argument to the effect that this is the case before the background of the kind of property regimes that are maintained by states as we know them. But fairness can still apply even when equality of opportunity does not. That is why Burtless et al. (1998) are wrong to claim that fairness in trade is conceptually muddled (a claim that depends on tying fairness to the leveling metaphor – a link that is plausible only if mediated through the notion of equality of opportunity). But, similarly, if one wants to argue that fairness requires equality of opportunity of sorts, one will need to supply an argument that the kind of balancing of stringent claims that becomes relevant in a certain context is indeed one that amounts to realizing equality of opportunity. What is lacking in Kapstein’s book is the kind of conceptual work that would allow him to do this, and thus to provide the kind of analytical work needed to overcome the preachy versions of moral thought. To see how this approach to fairness applies, consider agricultural subsidies, paid by the governments of the US, the EU, and Japan. Demands for the removal of these subsidies are common. What does fairness require here ? Note how subsidies operate. If governments pay producers for exporting products, this harms domestic consumers because it raises prices on the domestic market (because there are now fewer goods on it), and it benefits domestic producers. Internationally, it is the other way around, however : subsidies lower world market prices and thus benefit consumers elsewhere, while harming producers. As we saw above, then, agricultural subsidies to farmers in the EU, US, and Japan are ipso facto subsidies to consumers in net-food-important countries, which is what many of the least developed countries are. But subsidies do not just redistribute : they also cause deadweight losses, net-losses for the economy that arise because people produce more and consume less than they otherwise would. What the overall economy invests in subsidies is more than what the producers end up receiving. What is owed to whom here ? Net-food-exporting developing countries, or developing countries that, absent agricultural subsidies, would turn into net-food importers, could argue that the removal of subsidies is owed to them because, first, developed nations have a duty of assistance to them, and, second, trade liberalization is essential to development. We have seen above that this second claim is empirically controversial (contested, that is, by those who think that domestic policy reform is where the focus should be, rather than trade reform), but this sort of claim is often made, and is certainly within the realm of what is plausible. Net-food-importing developing countries, however, would be harmed, and thus would argue for the continuation of such subsidies also by reference to a duty of assistance. Now consider the farmers in developed countries. Do they have a claim in fairness to subsidies from their government? Under certain conditions they do indeed, in particular if they live in so-called coordinated market economies, such as Germany and Japan, where labor markets tend to be regulated in such a way that a high degree of specialization becomes the norm: workers invest heavily in such skills, and do so



because they are employed under conditions that render life-long employment in the same company rather likely, and that are indeed meant to foster such arrangements. Coordinated market economies, like other forms of capitalism, are characterized by institutional complementarities : labor markets work in certain ways, and so does, say, ownership of firms, all of this fits together into an efficient system. So we have three parties that could make a credible claim that something is owed to them. What is characteristic about discussions about fairness at the international level is that these claims would be made on entirely different bases. The consumers in netfood-importing developing countries and the producers in net-food-exporting developing countries can make their demands based on some views according to which development aid is owed to them (but one side would spell this out for and the other against agricultural subsidies). The producers in developed countries, at least to the extent that they are coordinated market economies, can argue in support of subsidies based on their status as citizens in a system that has shaped their lives in certain ways, namely in a way that requires a form of risk-taking that is acceptable to them, if indeed brought about by labor market regulation, only if the state is willing to support them if changes in trade policy, or other factors, make their specific life choices turn out to be unfortunate. (As opposed to this, farmers in liberal market economies, such as the US, may well not have this sort of claim. In such an economy, everybody is on the labor market as an entrepreneur. At any rate, labor legislation as such is not designed to encourage specialization, and thus at least on that basis participants in such economies cannot demand a safety net against failure on the market.) There is no politically feasible solution to this situation that would be beneficial to everybody. (One could, of course, think of paying some parties to this conflict a lot of money so that they change what they do to such an extent that they are no longer a party to this conflict, but in general this would not be a politically feasible solution. If one did have the resources to do something like this, it might well cease to be true that certain things are owed to some people, which would change how fairness applies to this situation.) What is required, from a standpoint of fairness, is an assessment of how to balance these different claims (to subsidies, or to the elimination of subsidies) proportionally. One proposal would be to require developed countries to stop forms of support for their citizens that are trade distorting; that is, they should pay them to do something else rather than give them export subsidies. For those affected by this in the developed world this would be a decidedly second-best scenario. Ideally, they would want to continue in their chosen line of work, given the strains of changing one’s economic niche involuntarily. But it may be what is required given the needs of developing countries. This solution would be good for the net-food-exporters in the developing world, but the net-food-importers would have to receive aid to help them out for a while. To some extent, this proposal would coincide with what Kapstein suggests; but the difference is that here fairness is explicitly understood as involving the balancing or weighing of stringent claims made on different bases, as a balancing of conflicting claims, which seems much closer to the moral reality of the situation than to think of it as a mutually beneficial arrangement of sorts. Fairness is concerned with balancing claims when they cannot all be satisfied entirely. Occasionally, no such balancing is required because an arrangement is possible that does benefit everybody. But trade



liberalization is not like that, even if, as trade theory has it, trade liberalization is in the long-term interest of all countries involved. If one merely focuses on that point of trade theory, one loses sight too much of certain matters of fairness. Recall at this stage also the political and economic reflections I have offered above. The proposal I just made, which is supposed to capture a balancing of claims rather than an immediately mutually improving arrangement, is the more plausible the more important trade is for development (which is a contested matter); and it involves the resolution of severe political problems. These problems are troublesome for the way of thinking about subsidies that I just proposed, but they are so much the more troublesome for Kapstein’s way of thinking about fairness that does not really get these difficulties into proper sight. Kapstein, again, overemphasizes the extent to which prudence and morality coincide for global economic justice, and ipso facto understates the extent to which global justice is an area in which competing claims conflict, as well as the extent to which an account of fairness should be concerned with making room for the task of balancing such conflicting claims. MATHIAS RISSE Kennedy School of Government, Harvard University

References Anderson, Kym (2004), ‘ Subsidies and Trade Barriers’, in Bjorn Lomborg (ed.), Global Crises, Global Solutions, Cambridge: Cambridge University Press. Anderson, Ky, and Will Martin (eds.) (2006a), Agricultural Trade Reform and the Doha Development Agenda, Palgrave Macmillan and the World Bank. Anderson, Kym and Will Martin (2006b), ‘Agriculture, Trade Reform, and the Doha Agenda’, in Anderson and Martin (2006a), pp. 3–37. Birdsall, Nancy, Dani Rodrik, and Arvind Subramanian (2005), ‘How to Help Poor Countries’, Foreign Affairs, July/August. Broome, John (1999), ‘ Fairness’, in John Broome (ed.), Ethics out of Economics, Cambridge: Cambridge University Press. Burtless, Gary, Robert Lawrence, Robert Litan, and Robert Shapiro (1998), Globaphobia. Confronting Fears about Open Trade, Washington, DC: Brookings Institution. Hertel, Thomas and Roman Keeney (2006), ‘What is at Stake: The Relative Importance of Import Barriers, Export Subsidies, and Domestic Support’, in Anderson and Martin (2006a), pp. 37–63. Londregan, John and Keith Poole (1996), ‘ Does High Income Promote Democracy? ’, World Politics, 49: 1–30. Przeworski, Adam, Michael Alvarez, Joses Cheibub, and Fernando Limongi (2000), Democracy and Development: Political Institutions and Well-Being in the World, 1950–1990, Cambridge: Cambridge University Press. Risse, Mathias (Forthcoming), ‘Fairness in Trade’, Kennedy School of Government Working Paper, No. RWP05-004. Roemer, John (1998), Equality of Opportunity, Cambridge, MA: Harvard University Press.

World Trade Review (2007), 6 : 1, 135–148 Printed in the United Kingdom f J. Michael Finger doi :10.1017/S1474745606003120

Review essay

Trade negotiations and developing countries Negotiating Trade : Developing Countries in the WTO and NAFTA. Cambridge, Cambridge Edited by John S. Odell Cambridge University Press, 2006, pp. xi, 298

Behind the Scenes at the WTO : The Real World of International Trade Negotiations By Fatoumata Jawara and Aileen Kwa London New York, Bangkok : Zed Books in association with Focus on the Global South, updated edition, 2004, pp. lxxxv, 329

The WTO, we hope, is an institution that mutes the importance of raw power – provides a system for working out problems among countries in which the interests of smaller countries are not always overwhelmed by those of larger. The two books reviewed both address this issue, but in different ways. The Odell volume (a collection of studies by different analysts) reviews a number of WTO events in which developed and developing country interests were at odds ; e.g., the ‘bananas dispute’ involving Ecuador, the US, and the European Communities. The studies in that volume document the skill of developing country negotiators to use the system to their advantage; they demonstrate that the WTO process often came to outcomes more favorable to smaller countries than a simple weighing of relative power would imply. Jawara and Kwa analyze only one event, the WTO agreement at Doha in 2001 to open a new round of negotiations. In contrast with the Odell studies that focus on strategies and approaches – what developing country delegates did, what worked, what did not – Jawara and Kwa focus on ‘the WTO,’ concluding that ‘[d]eveloped countries are benefiting from the WTO, as are a handful of (mostly upper-) middleincome countries. The rest, including the great majority of developing countries, are not’ (269). My major aim in this review is to retell in shortened form some of the stories the books present. I want to provide in a form convenient for introduction into conversation and debate examples – evidence – the reader can cite in favor of or against the usefulness of the WTO for smaller countries. (Where I have added information to what the authors provide, I provide references.) John Odell, in his introduction, attempts to identify common factors and strategies that contribute to outcomes favorable to weaker members. After summarizing the findings in his volume, I offer a short comment to the effect that the uncommon elements and the skills of the negotiators are more important. When I turn to the Jawara and Kwa volume, I take pains to tell their story in their own words. It is, to my reading, less a story about ‘the WTO’ than about developing 135



country delegations being systematically outfoxed by developed country delegations, particularly the US delegation. I argue however that these authors offer little evidence to support their allegation. They offer no evidence that developing countries considered the agreement in conflict with their interests, or that in it was in either political or economic substance. Given that they have not demonstrated that a ‘ crime’ was committed, their identification of the ‘guilty party’ – developed and developing country delegations or ‘the WTO’ – falls apart.

Highlights from the studies in the Odell volume

Bananas; Ecuador (Chapter 8: James McCall Smith) As part of its ‘ single market ’ reforms of the 1990s, the European Union established COMB, the Common Organization for the Marketing of Bananas. Its system of quotas favored ACP countries over Latin America and the Caribbean. Its allocation of import licenses was to distributor companies rather than to countries, and favored those who had ripening and storage facilities inside the EU. In 1993 Colombia, Costa Rica, Guatemala, Nicaragua, and Venezuela raised a GATT complaint against the COMB. The EU blocked establishment of a panel but did enter into negotiations with the complaining countries. The result in 1994 was a Banana Framework Agreement with Colombia, Costa Rica, Nicaragua, and Venezuela. The agreement would increase these countries’ quotas and allow them to allocate 70% of their quotas to distributors of their choice. Other Latin American producers, Guatemala, Ecuador, Honduras, Mexico, and Panama, would lose market share. Ecuador is sometimes called the Saudi Arabia of bananas: one-third of global banana exports, lowest cost producer, production mostly locally owned (not by multinationals as in Honduras and Nicaragua), and the home of significant trading companies. Not a GATT Contracting Party, Ecuador had not joined in the 1993 GATT complaints against the COMB. The multinational distributors Chiquita and Dole controlled at the time more than one-half of world banana trade. Dole, to accommodate COMB, invested in ACP production and in ripening facilities around Europe. Chiquita took no such steps and saw its share of the EU market fall from 30% in 1992 to 19% in 1995. Chiquita used Section 301 of US trade law to apply pressure to the US Trade Representative to take WTO action against the Framework Agreement. The US, not a producer of bananas, needed Ecuador to give credibility to its case, but Ecuador was not a member of the WTO. From this situation, Ecuador drew leverage in its WTO accession negotiations. Moreover, the Latin American countries party to the Framework Agreement came under US pressure to withdraw and to support the US–Ecuador complaint against the EU banana regime. In September 1997, the WTO dispute settlement process ruled against the EU banana regime and gave the EU 15 months to comply. Neither the US nor Ecuador considered the modifications the EU proposed to be sufficient for compliance. At this point ensued a heated dispute between the US and the EU over the next procedural step. The US, citing one article of the WTO dispute settlement understanding, asked for a panel to arbitrate the retaliatory action that would be



appropriate – in effect, give the US a choice between accepting what the EU proposed or an approved retaliatory response. The EU, citing another article, insisted that the next step would be to constitute a panel to evaluate the concurrence with WTO rules of the new EU regime. Ecuador found a third course, asked the original panel to consider the adequacy of the EU adjustments. The same persons ended up acting as three different panels, requested by the US, the EU, and Ecuador. They found a way to moderate differences by releasing simultaneously reports from each of the three panels, again finding the EU not in compliance. The US moved quickly toward imposition of sanctions but Ecuador again followed a separate course. Ecuador continued to negotiate with the EU, but to little result, and in time began steps for retaliatory sanctions. As Ecuador’s merchandise imports from the EU are small, she was able to convince the WTO panel that cross-retaliation – on intellectual property – should be authorized. In its list of items on which retaliatory sanctions might be imposed Ecuador listed not only merchandise but also music copyrights, geographical indicators, and industrial designs. Several developing countries openly complemented Ecuador for this innovative approach to strengthening the hand of smaller WTO Members. Another source of comfort for smaller countries was perhaps that Ecuador had cleverly tweaked the nose of the US. It was at US insistence that the dispute settlement understanding provides for cross-retaliation. The US saw its capacity to retaliate on merchandise exports of developing countries an important tool to defend US intellectual property interests. Negotiations over an outcome soon focused on an eventual tariffs-only regime, preceded by a transition period of quotas. Here the interests of the US and Ecuador differed. USTR’s constituents in the matter were distributors, with strong interest in rents from the transition system of quotas allocated to distributor companies. Ecuador, as a low cost producer, would gain from the elimination of quotas, but was also the home of several distributors who wanted larger shares of the transition quota allocations. Moreover, a tariffs-only regime would still leave Ecuador disadvantaged relative to ACP countries, to whom the EU was offering preferential arrangements. Ecuador rejected a proposal put together by the US and the EU and forced significant modifications that provided for Ecuador not only a larger share of EU imports but also a stronger position for its trading companies. Some observers suggest that Ecuador also received EU support at the IMF, at development banks, and with private creditors to help to allay a financial crisis. In the end, the EU recognized Ecuador’s principal supplier status in the process that will establish the EU’s tariff rates. The EU also accepted in principal that the new EU banana regime will not diminish the total market access of Latin American banana exporters. In these negotiations, Ecuador exhibited a mastery of WTO legal and procedural technicalities and an entrepreneurial capacity to create and to exploit the many dimensions of leverage that the situation allowed. Taking a position independent of the US was risky, but at the same time widely respected among other WTO members. Moreover, the EU Member states were divided on what the outcome should be; Sweden, Finland, Austria, and Germany publicly supported Ecuador when she objected to the US–EU plan. Ecuador also used to advantage the familiar complaint



that whether the WTO elephants dance or fight, the interests of the smaller Members are likely to be trampled.

Scallops, sardines and catfish ; Peru, Chile, Vietnam (Chapter 7: Christina L. Davis) In May 1995, Canada requested consultations with the EU over new French requirements for the labeling of scallops. The new French rules barred Canada from labeling its species of scallops as ‘noix de coquille Saint Jacques’, the familiar name for scallops in France. Peru and Chile, in similar situations vis-a`-vis their scallops exports, joined the request for consultations and the subsequent request for a WTO panel. The panel issued its interim ruling in March 1996. Such rulings are circulated only to parties involved in the dispute, but apparently found the French regulation in violation. While the EU Commission had been reluctant earlier to press the French to change their regulation, with this (apparent) legal leverage in hand the EU announced that the French regulation would be replaced by an EU regulation that allowed marketing scallops under the name ‘noix de coquille Saint Jacque’, so long as the label also provided the species and country of origin. Ms Davis reviews also a complaint Peru raised against EU regulations on the labeling of sardines. Peru had developed in Europe a market for ‘Pacific sardines’, but the relevant EU regulation (previously not enforced against Peru) barred marketing fish as ‘sardines ’ unless they were a species common to the Atlantic Ocean and the Mediterranean Sea. Otherwise, they must be labeled ‘pilchards’ or ‘sprats ’. A key element in the case was the relevant international code, the Codex Alimentarius. Codex lists the Peruvian species among the species it defines as sardines, but prescribes that only the European species can be labeled simply ‘ sardines’. Others should be labeled with the country, region, species, or common name prefixed. In this instance, the outcome was a new EU regulation that allowed the use of the name ‘ sardines’, but required that it be preceded by the scientific name of the species. The new EU sardines regulation meets the Codex standard and parallels the new regulation for scallops. It is not the outcome the Codex allowed that would have been most favorable to Peruvian interests, but it is better for these interests than the rule that it supersedes. As well as providing a standard against which the questioned regulation would be evaluated, the WTO system required that the EU address the complaint within the system and provided the exporting countries a venue for bringing together common interests. In the sardines case, Peru received legal assistance from the Advisory Center on WTO Law, an agency created to help developing countries in the WTO dispute settlement process. In both cases, there were strong European consumer and NGO interests in favor of a liberal regulation. The interests of the EU Commission in displacing the authority of national administrators on such matters was perhaps another factor. (Vinod Rege, 2002, has explained that such concern influenced the Commission’s shift of position on customs valuation during the Uruguay Round.) The legal leverage of the international rules helped to tip the balance toward European interests favoring a more liberal rule.



In a dispute over how Vietnamese-grown catfish could be labeled in the US market, the wording of the relevant rules was similar to the scallops case above. The relevant section of the US–Vietnam bilateral treaty provided the same wording as the article in the WTO TBT agreement relevant in scallops case. Vietnam, not a WTO Member, did not, however, have at its disposal the assets that membership provides. The result: the Vietnamese species of catfish cannot be labeled ‘ catfish’ in the US market, only the species common to the US South-East can be so labeled.

Intellectual property in the Uruguay Round and later (Chapter 2: J. P. Singh; Chapter 3: John S. Odell and Susan K. Sell) One of the major concessions accepted by developing countries in the Uruguay Round was the bringing of intellectual property under WTO regulation. Intellectual property came into the Uruguay Round agenda at the insistence of developed country industries, whose initial focus was to control counterfeiting of brand names and recorded entertainment. Agribusiness and software soon saw potential. Before the run-up to the Uruguay Round, intellectual property had been little discussed among developing country negotiators or analysts. One of the key elements in its inclusion in the Uruguay Round agenda was the threat of continuing unilateral US action under ‘301 ’. Moreover, North IP interests framed the TRIPS issue as piracy versus virtue rather than as economics. The advanced Asian developing countries saw themselves as near-term winners on IP, and were not active in opposition. And, as Sylvia Ostry has pointed out, at the time, few delegates or analysts (such as this reviewer) appreciated the size of the liability TRIPS would create for intellectual property users. This initial outcome is explained largely by relative power and preparation ; later events demonstrated how developing countries were able to use the system to win back some lost ground. To now, this recovery has focused on public health – HIV/ AIDS – and the relevant TRIPS provisions that provide for compulsory licensing. Many interests, vocal NGOs and the news media viewed TRIPS not as a matter of piracy versus honesty but as poor people’s health versus large companies’ profits. A number of international organizations, such as the World Health Organization, lent their support on the public health side of the issue. With the changed perspective and the marshalling of public health interests, intellectual property users have been able to expand the rights of WTO Members to use compulsory licensing – in the Declaration on the TRIPS Agreement and Public Health, adopted in November 2001 by the Fourth (Doha) Session of the WTO Ministerial Conference, the Decision of the General Council of August 2003, Implementation of paragraph 6 of the Doha Declaration on the TRIPS Agreement and public health and two other Council decisions.

The Mexican petroleum sector in NAFTA (Chapter 6: Antonio Ortiz Mena L. N.) Despite pressure from the US, Mexico in negotiating NAFTA successfully resisted any relaxation of her prohibitions against foreign investment in oil exploration,



extraction, refining, and retailing. The chapter author identifies several factors behind this success. . Mexican President Salinas, aware that the US Congress would raise strenuous objections, asked that migration be included in the negotiations. By focusing Mexico’s exclusion of the petroleum sector on similar exclusions by the US and Canada – the US excluded maritime services as well as migration ; Canada excluded cultural industries – Mexico minimized the need to ‘pay’ for this exclusion by giving up demands in other sectors. . Mexico in addition contracted with front-rank consulting firms in the United States to inform themselves – independently of the USTR – of industry, congressional, etc. positions. . The petroleum industry in the US had already adopted itself to Mexico’s policies, e.g., by trading with Mexican companies in petroleum products at the wholesale level, and hence had existing relationships that were valuable. They did not strenuously press the USTR to break the Mexican position.

WTO and the Singapore Issues (Chapter 4: Amrita Narlikar and John S. Odell) The Like Minded Group began as an eight-member coalition in the preparatory process leading up to the Singapore Ministerial Conference of 1996. Its initial agenda was to block inclusion of the eventually called Singapore issues – investment, trade facilitation, transparency in government procurement, and competition policy – in the next round of WTO negotiations. The group came in time to include 14 countries (all developing, one of whom described itself as an observer rather than a member) that together provided about 5% of world exports. It became a major sounding board for the idea that the Uruguay Round outcome had been unbalanced against developing countries, and for the proposition that before another round could be opened the developed countries must make up this imbalance through additional concessions. Their listing of what these concessions should be covered most of the existing WTO agreements and contained many demands in circulation for several decades under the label special and differential treatment – more rhetorical attempts to influence developed country behavior than workable proposals to regulate it (Finger and Wilson, 2006). After the Doha Ministerial Conference of 2001 launched a new round of negotiations, the group ceased to function. Of the Singapore issues, only trade facilitation, in a circumscribed way, has found its way into the Doha Round agenda. As to the unilateral concessions the group had specified the developed countries must make before a new round could be launched, they achieved little. The chapter does however report a number of other payoffs members of the group received to accept the launch of a new round. For example, Pakistan received a larger US aid package and larger EU import quotas for clothing. The African Group won the 2001 TRIPS–Public Health Declaration and a WTO waiver for EU–ACP preferences. Moreover, the group helped to bring attention to the need for ‘aid-for-trade’ to make trade an effective development tool. This awareness has led to significant increases in aid-for-trade budgets and to a WTO working group on aid-for-trade to prepare recommendations for Doha Round negotiators.



This reviewer’s conclusions from the Odell studies

1. Developing country negotiators did well In support of that conclusion I offer the summaries above, I also note that there is a minimum of second-guessing in the book. Furthermore, what second-guessing there is suggests that the negotiators did as well as could be done. For example, chapter 4 posits that the Like Minded Group might have been more successful had they asked for less on geographical indicators, for more on US acceptance of disciplines on antidumping.

2. Generalizations on strategy are risky Odell speculates that coalitions defined over narrower issue areas are likely more effective than those defined over more collective issue areas, but he adds in a footnote that one of the chapter authors argues to the contrary. He also speculates that developing countries that combine with a developed country will do better, but Chapter 7 documents that Peru, going it alone, got the same result on sardines as the coalition of Peru, Chile, and Canada obtained on scallops. Likewise, Chapter 8 highlights how Ecuador gained by separating itself from the US on important parts of the bananas issue.

3. The editor’s generalizations on strategy are too general to be helpful to practitioners An example of the lessons the introduction draws: ‘The larger the coalition the less it will lose and the more it will gain, provided that it manages the fragmentation problem’ (p.14). Chapter 2 reports that managers of a developing country compulsory licensing coalition kept the coalition together when the US devised a clever ploy to split it, eventually won the 2001 declaration on compulsory licensing. From a practitioners perspective, however, the valuable information is less that the managers held the coalition together than how. Absent such information, the generalization is left with a ‘bell the cat’ air about it. Describing events in an artificial language – e.g., offensive distributive negotiating strategy, purely integrative strategy – rather than clarifying what happened, this language makes the presentation cumbersome. By my instincts, formed from 30 years a bureaucrat and six years an academic, the stories are useful, but the introduction appeals to the status markers of an academic rather than of a practitioner. Success, I learn from these cases, lies in the capacity to apply imagination within the system, and that requires (1) entrepreneurship and (2) a mastery of the rules, processes, and traditions of the system. The relevant language is the language of the WTO rules. Seeking a ‘typology for uniformly classifying general courses of action available to negotiators, regardless of the issue’ (15) calls attention to the ordinary and overlooks the importance of the extraordinary – the skills, imagination, and daring of negotiators. A point that chapter 2 documents but does not state is that without Felipe Jaramillo and Enrique Iglesias the Punta del Este agreement to launch the Uruguay Round would not have been achieved. Likewise, without Felipe Jaramillo’s knowledge and the respect he had earned over the years in acquiring and applying it, the services



negotiations would not have ended successfully. For example, the compromise he found (Chapter 2 reports it) that services would be, technically speaking, a separate negotiation, but conducted by the same people as the other parts of the Uruguay Round, provided a way for the Indian and the US delegations to get around domestic constraints that would otherwise have prevented the launch of the Uruguay Round. The uncommon, not the common, made the difference. People matter; a lot. The assembly hall for the GATS Council should have a portrait of Felipe Jaramillo behind the dais.

Behind the Scenes at the WTO The strength of Jawara and Kwa’s book is the exciting story they tell. To ensure an accurate telling I will relate it in large part in their words. Everything below in italics is quoted from the book; numbers in parentheses are page references. The book recounts what happened before, during and after the Fourth WTO Ministerial Conference in Doha, Qatar, in November 2001. It tells how developing countries were bullied and coerced into acquiescing with an ‘ agreement’ with which most of them profoundly disagreed. (xv) _ The whole process leading up to Doha was carefully orchestrated by the USA, the EC and the [WTO] Secretariat, to ensure that there would be a new round of negotiations, including the ‘new issue’, irrespective of the views of the Southern majority of the organization’s membership. (78) The central actor is the United States, who, in steps, expand their circle of influence to (a) the EC, (b) the Quad (US, EC, Japan, Canada), (c) other developed countries, (d) middle-income countries, (e) other low-income countries, (f) finally, the least developed countries. As the circle expands, the countries taken in help to seduce or cajole into line those in the next larger circle. (The progression is diagrammed on 57.)

How did the United States do it at Doha? 1. They were led by Superman. Robert Zoellick is a highly cerebral workaholic. [He] came to Doha extremely well prepared, having dissected what went wrong in Seattle in minute detail and studied a RAND Corporation report on how negotiating styles varied across cultures. He had worked relentlessly to court ministers and presidents on different continents _ He impressed the Kenyan trade minister by speaking to him in his local language, Swahili _ and badgered Japan to play a more cooperative role (84) 2. They had Stuart Harbinson (Hong Kong ambassador to the WTO and chair of the WTO General Council) prepare a biased draft ministerial declaration and then employ a parliamentary trick to gain acceptance of it as the working document for the conference (89) – a document which broadly coincided with what the Quad – and particularly the USA and the EC – wanted, while largely ignoring the concerns of the developing countries. (118) 3. They manipulated proceedings. The Quad members were looking forward to ‘a controlled bedlam’ _ in which they could manipulate other participants to achieve their objectives; and the Doha draft gave them just what they needed to do so. (74)



4. They gave a bit on superficial procedural matters so as to make the conference chairman Yousef Hussain Kamal, Minister of Finance, Economy and Commerce of Qatar look less of a stooge for the powerful nations and the Secretariat. (89) 5. They hand-picked issues managers, ‘friends of the chair’ who, they were confident, would stay in line. (George Yeo, Singapore; Pierre Pettigrew, Canada; Pascal Cuochepin, Switzerland; Heraldo Munoz Valenzuela, Chile; Luis Ernesto Derbez Bautista, Me´xico; Alex Irwin, South Africa) so that [w]hile the facilitators played the key public role, behind the scenes it was the Secretariat staff headed by Andrew Stoler [of the United States], who were running the show, taking charge of the administration and the drafting process under the watchful eye of the Quad. (91) 6. US citizens dominated the WTO staff not by numbers but by application of super-powers similar to Zoellick’s. Jawara and Kwa report that, re the number of US citizens on the WTO staff, there are five times more French, three times more Brits, eight more Swiss, and two more Canadians. (201) 7. The affair was less a power play than a seduction. Such was Zoellick’s new-found friendship with African ministers that nearly all of them clamored to sing his praises after the deal in Doha was brokered. ‘Mr. Zoellick put this whole thing together,’ said Tanzania’s Trade Minister Iddi Mohamed Simba. Morocco’s ambassador at the time, Nacer Benjelloun-Touimi, added, ‘That is what leadership is about. He [Zoellick] has been brokering the deal between the Europeans and the African countries.’ (116)

The Machiavellian food chain The US, atop the food chain, enlists allies only so long as it can profit from them. When the Bretton Woods Agreement was put into effect, it produced dramatic changes in the world economic and financial order, which helped propel the USA into superpower status, while Britain [its ally in advancing the agreements] watched helplessly from the sidelines. (138) As had happened in 1944 at Bretton Woods, [h]aving largely secured the terms it wanted at Doha, through a show of unity with the other members of the Quad, the USA proceeded over the following months to ignore major elements of its obligations under the WTO agreements, creating a rift between itself and the EC Japan and Canada. (80) With the developed countries co-opted, the next stage was to win over the middle-income countries. [C]ertain middle-income countries that had been openly apprehensive about US and EC positions before Doha could be found secretly lobbying their less affluent counterparts to adopt these very positions at the ministerial itself _ Offers of bilateral trade and other agreements played a key role in such changes of heart. (149) The co-opted became aggressive allies. Some middle-income countries go beyond stabbing other developing countries in the back when necessary to get ahead of the game. (150) Jawara and Kwa name names.



Mexico _ was one of the middle-income countries that aligned themselves most closely with the positions of the powerful countries. (163) Chile is another favorite of the big countries and top Secretariat staff. (164) Costa Rica is sometimes viewed as a US ‘clone ’ within the WTO. (164) After Seattle, the EC, and the USA decided to delegate part of the responsibility for aligning African positions with theirs to South Africa. (168) _ [South African] trade officials started their campaign by ‘ persuading’ their less affluent neighbor, Lesotho, to support their position, which Lesotho duly did. (169) South Africa’s payoff included $923 million of exports to the USA duty-free under AGOA, $9 million to train customs officials (promised by Zoellick), $8.7 million in trade capacity building for SADC members, $500 million technical assistance for COMESA, $300 thousand to support COMESA’s court of justice (169). Moreover, half of the $12 billion in aid already earmarked for developing countries following the March 2002 Monterrey Summit on Financing for Development would go to Africa, plus an additional $1 billion for debt relief. (170) Just before the Doha meetings, Zoellick announced that the USA would expand Indonesia’s benefits under the Generalized System of Preferences. After Doha, Indonesia received supportive visits from Zoellick and US Secretary of State Colin Powell. (172) Five months after Doha _ the USTR announced that it had signed a Trade and Investment Framework Agreement with Senegal and seven other West African countries. (174) Kenyan Trade Minister Nicholas Biwott quickly established a rapport with Zoellick and was enthralled by the attention he received _ When the [ACP] waiver was granted, he saw this as a huge victory and consolidated his support for the powerful countries’ positions in other areas such as the ‘ new issues’. (176) A week after Doha the World Bank and the IMF suddenly announced that Tanzania would receive external debt relief of $3 billion _ in February 2002 Canada announced the cancellation of a further $80 million in bilateral debt for Tanzania. (180) To the reticent, the United States applied power ; disinformation campaigns to undercut developing country Geneva representatives with their governments at home. They got five Geneva representatives sacked, and placed at least two more on the US black list (150–151). For example, [t]he former ambassador [of Egypt] was quietly relieved of her duties at the request of the USA during the Doha Ministerial, and promoted when she returned to Egypt. (171) The former ambassador of Pakistan to Geneva, Munir Akram, was given two months notice to leave his post in February 2002, for a United Nations post in New York As a Southern delegate put it: ‘ If you become too outspoken you disappear, simple as that ’ (175–176). The tour of duty of Dr Alfredo Suescum, Geneva Ambassador of Panama, was not renewed when it ended in December 2001, a month after the Doha Conference. (167)

It was the United States versus a bunch of chumps Except for Robert Zoellick and Andrew Stoler, everyone mentioned in the book comes out a chump. For all the praise and support Kenyan Trade Minister Nicholas Biwott and other African delegations lavished on Robert Zoellick, Jawara and Kwa conclude



that they got, in return, only crumbs. They offered the crumbs of [the Declaration on the TRIPS Agreement and public health] and the ACP waiver to buy the support of developing countries, then simply wore down the opposition (112). The African delegates are portrayed as cuckolds; the Latin Americans as cornudos. Some of the Asians come out no better. Indonesia sold out for tariff preferences and diplomatic visits from Zoellick and Powell. (172) Minister Fafida Aziz of Malaysia was very much on the side of developed countries and even implored the Africans to go along with the deal (165). We learn that developing country delegates, particularly African delegates, are in it for the glory of diplomatic victories. Developed country delegates, particularly US delegates, are in it for their constituents’ money. When the United States goes around developing country Geneva delegations and deals with ministers and presidents in national capitals, it deals with representatives even further separated from their own peoples’ interests. Wow ! It takes courage for Jawara and Kwa to suggest such things. They offer however no proof.

The book’s weaknesses While the strength of the book is its exciting story, its weakness is the technical knowledge of the authors – of WTO rules and procedures and of research techniques – that bring forward doubts that the story is true.

No evidence that a crime was committed Developing country officials Jawara and Kwa assert, ‘ accepted an agreement with which most of them profoundly disagreed ’. If true, this is a serious indictment of the developing country officials and of the system that brought them to this acceptance. To convict someone of a crime the prosecutors must first establish that a crime was committed. This Jawara and Kwa do not do. With few exceptions, the developing county officials called as witnesses (i.e., quoted in the book) testify that the agreement was a good one. Jawara and Kwa’s own witnesses refute their key premise. This moves the issue to ‘ should have disagreed ’, but leaves the conflict between the agreement and developing country interests solely a matter of Jawara and Kwa’s insistence – not developing country governments’ insistence. In support, they offer little except to assert that any belief that developing countries would benefit is based on ‘a combination of ideology, paternalism, and missionary zeal. The true believers in globalization and liberalization feel sure that they know best _ ’ (269). Jawara and Kwa’s rampant second-guessing of developing country governments suggests that the true unbelievers likewise feel sure that they know best.

Victories unrecognized Another analytical flaw is that the authors do not recognize victory when developing countries achieve it. I provide two examples. The first relates to Jawara and Kwa’s principal marker for developing country success at Doha : whether or not the agreement provided for negotiations on the ‘ new issues ’ of investment, competition



policy, government procurement, and trade facilitation. They report the text of the agreement and Chairman Kemal’s explanation. This explanation stated that ‘ in my view this would give each member the right to take a position on modalities that would prevent negotiations from proceeding after the Fifth Session of the Ministerial Conference until that member is prepared to join in an explicit consensus ’. That sounds like ‘No. ’ Moreover, we know now that because no such consensus was achieved in later Ministerials (except, in Hong Kong, for limited negotiations on trade facilitation), there have been no such negotiations. Yet Jawara and Kwa report, ‘failure to stop the new round, for example, by blocking negotiations on the ‘‘new issues ’’’ (306) On Kemal’s explanation, they conclude: ‘ this ‘‘ clarification ’’ has proved almost worthless’ (111). A second example, when developing countries asked that ‘ credit’ be given in future GATS negotiations for previous unilateral liberalization, [t]he major powers, clearly unwilling to move on the matter, threw several spanners in the works, asking for the developing countries to ‘bind’ their liberalization under GATS. _ It was finally agreed that binding would not be required, but that it would be taken into account in determining the extent of a country’s credit’ (35). The authors’ conclusion: ‘ This largely defeated the purpose of the exercise’ (35). Would victory be the outcome only if binding could not be taken into account ?

Evidence of arm-twisting As to pressure on developing country negotiators, few would insist that it was not applied, but Jawara and Kwa hardly provide ‘smoking gun ’ evidence that the pressure applied was ‘ undue’ or ‘unusual’. For example, they report that ‘ On 25 October [2002] assistant USTR Rosa Whitaker sent a letter to sub-Saharan African trade ministers (Box 9.1) attempting to pressure them into accepting the US position linking the issue [TRIPS – public health] with market access in textiles under the African Growth and Opportunity Act (AGOA) ’ (248). Ms Whitaker’s letter, reproduced in the book, states as follows: Sadly, while HIV/AIDS has taken its greatest toll in sub-Saharan Africa, most of the region’s representatives to Geneva are not attending meetings related to this issue or engaging in the debate. The attached paper summarizes the main elements of this issue and the U.S. proposal. I urge you to instruct your officials in Geneva to work with the U.S. and other African countries to ensure that the solution the TRIPs Council develops benefits African countries and responds to the region’s needs. (250–251)

On market access in textiles and clothing (a separate section of the letter), Ms. Whitaker notes that Contrary ‘to positions conveyed by the African Union and many African Trade Ministers, the African Group has submitted a proposal supporting the early elimination of quotas [before the January 2005 elimination mandated by the Uruguay Round Agreement] ’ (251). She points out that African countries already have quota-free access to the US and other markets under AGOA and other preference schemes. African exports benefit from the quotas still in place on the exports of China and other powerful exporters. Ms. Whitaker concludes, ‘I hope that you will immediately instruct your representatives in Geneva to act to protect the region



from these simultaneous assaults on Africa’s interests _ I am confident that you too will recognize the convergence of interests between the US and sub-Saharan Africa’s interests on issues related to TRIPS and textile-apparel quotas ’ (251). Otto von Bismarck, is reported as once saying, ‘The less people know about how sausages and laws are made, the better they’ll sleep at night’ – I doubt this was the bloodiest step in making the Doha sausage. However, other than statements from unidentified persons that pressure was applied – conclusions of the interviewees, not evidence on which the ‘jury ’ (reader) might base her own conclusions – it is the only evidence Jawara and Kwa present. Likewise, the story that Jawara and Kwa provide about the Supachai–Moore deadlock over who would be appointed Director General is to me weaker evidence of the undemocratic nature of WTO proceedings than Jawara and Kwa want it to be. As they report, in a telephone conversation, US Secretary of State, Madeleine Albright, suggested to her Thai counterpart, Surin Pitsuwan, that both Supachai and Moore should withdraw and allow a search for a third candidate. In reply, Surin suggested successive three-year terms for Moore and Supachai. ‘ ‘‘That sounds interesting! ’’, she [Albright] said ’ (189). Some Geneva conversations later, Australia proposed the successive Moore– Supachai terms ; the WTO General Council enacted them. From this, Jawara and Kwa conclude : ‘That was how Dr Supachai’s position was secured – at the highest diplomatic lever between Bangkok and Washington. Not at the WTO General Council in Geneva, as many had come to believe ’ (189). A sausage this cleanly made would be welcome on my table.

WTO technicalities I will offer three examples. First, tariff preferences for developing countries require special legal arrangement not because WTO rules would otherwise require reciprocity (as Jawara and Kwa state, p. 152), but because GATT Article I requires MFN treatment. Second, an EC proposal on trade facilitation that members should ‘take administrative actions, decisions or rulings affecting importers or exporters only where a legal basis to do so is established ’ (243) does not mean that ‘members’ ability to take action at the border should be limited to what is allowed by a WTO trade facilitation agreement’ (243). The limit on the discretion of national administrators would be national regulation. Limits on national legislation would be a separate matter. Among the recommendations, Jawara and Kwa offer are the following : . Mini-ministerial meetings should not take place. (280) . No green room meetings should be held. (281) . There should be transcripts of the negotiations. (281) One technical problem with these is that defining a mini-ministerial would be difficult, e.g. was the Albright–Surin telephone conversation a mini-ministerial ? In addition to the two mini-ministerials at which, Jawara and Kwa allege, the United States marked the Doha Ministerial deck, they list six preparatory meetings among developing countries. Would they likewise be illegal ?



‘Green room meetings ’ would be likewise difficult to define. Does the ban apply to all meetings of less than full WTO membership ? What about meetings that take place outside the WTO building, such as those of the ‘Group de la Paix’, at which a draft for the Punta del Este Ministerial Declaration was worked out? Conversations over coffee, or telephone conversations among delegates ? Which of these are ‘ negotiations’, for which transcripts would be mandated?

WTO agreements should be evaluated This reviewer agrees, of course that each proposal and each actual WTO agreement should be viewed with skepticism – before, during and after negotiations (and can claim some experience doing so). Will the agreement advance each country’s national economic interest? Sylvia Ostry’s observation that many of the Uruguay Round agreements were accepted before their effects were understood merits more attention than it has received. The view should however admit the possibility that the answer is ‘Yes ’. Had Jawara and Kwa been willing to admit that possibility, they would have written a different book. Note: The reviewer would like to thank Mauricio Reina of Fedesarrollo, Bogota´, for useful comments. J. MICHAEL FINGER Washington, DC

References Finger, J. Michael and John, S. Wilson (2006), ‘ Trade Facilitation, Implementation, the Doha Development Agenda’, World Bank Staff Working Papers, forthcoming. Ostry, Sylvia (2002), ‘ The Uruguay Round North–South Grand Bargain: Implications for Future Negotiations’, in Daniel M Kennedy and James D. Southwick (eds.), The Political Economy of International Trade Law: Essays in Honor of Robert E. Hudec, Cambridge and New York, Cambridge University Press, Ch. 10, pp. 285–300. Rege, Vinod (2002), ‘ Customs Valuation and Customs Reform’, in Bernard Hoekman, Aaditya Mattoo, and Philip English (eds.), Development, Trade and the WTO: A Handbook, World Bank.

World Trade Review (2007), 6 : 1, 149–154 Printed in the United Kingdom f Cambridge University Press

Book reviews

The WTO and Sustainable Development By Gary P. Sampson . . . David Robertson Economic Development and Multilateral Trade Cooperation By Evenett, Simon J. and Bernard M. Hoekman . . . . . . . . . . . . . . . . . . . Kevin P. Gallagher


The WTO and Sustainable Development By Gary P. Sampson United Nations University Press, Tokyo, 2005, 315 pages

In the decade since the WTO began to operate, serious questions have arisen about its prospects. In these circumstances, it is disappointing to find an experienced commentator, like Gary Sampson, prepared to compromise that system. He supports extending WTO responsibilities to accommodate the demands from UN agencies and advocacy groups (NGOs), which seek to exploit the leverage of trade negotiations to pursue environmental, development, and social goals. The failure of other international agencies to deal with these problems indicates the global trading system would be jeopardized. My difficulties with this book begin with its title. ‘Sustainable development ’ is a convenient political catch phrase for journalists, publicity-seeking politicians, and NGOs, but in serious debate it must be defined. Sampson avoids this by resorting to ‘a convention’. He adopts the rather tired and controversial definition adopted in the Brundtland Report (1987); namely, that present actions should not compromise the opportunities of future generations. Since technical and innovative advances vastly increase future potential and facilitate substitutability, the prospects of future generations seem assured. So what is the WTO required to do ? Sampson sifts his 20 years’ experience with the GATT/WTO in search of an enhanced role for ‘sustainable development ’. By importing details of two specialist WTO committees – the Committee on Trade and Development (CTD) and the Committee on Trade and Environment (CTE) – he argues that ‘sustainable development’ already influences the work programme. This is drawing a long bow, because at different times North–South divisions have prevented constructive negotiations in both these committees. Moreover, these committees are not central to WTO activities. Since the conclusion of the Uruguay Round, the WTO has had an overloaded agenda. Outstanding issues include quarantine and other technical standards, biotechnology (including genetically modified organisms covered in the Cartagena protocol), intellectual property rights, trade in services, and trade and development issues as set out in the Doha Round agenda. This book has chapters on each of these subjects. Other ‘ trade linked ’ issues, placed on the WTO work programme since the Uruguay Round was concluded, focus on liberalizing ‘beyond the border ’: investment rules, 149



competition regulations, labour market controls, and government procurement policies. Discussion of these issues was blocked by developing countries at the Singapore ministerial meeting in 1996, and they remain wary that any agreement might reduce their policy freedom. At Cancun in 2003, the EU and Japan continued to press demands for negotiations on these ‘ link’ issues, to the point where the chairman terminated that meeting early to prevent a total breakdown. At several points, Sampson offers his view that these issues should not be included in WTO negotiations. Yet elsewhere, he argues that the environment and development fall within the WTO’s ambit, based on his experience with the CTD and CTE. For many years, environmental groups have argued that GATT/WTO dispute procedures should be used to provide environmental protection. In the 1990s, environmentalists and other NGOs launched campaigns to protect dolphins and turtles from dangerous fishing methods used to catch tuna and shrimp, respectively. WTO dispute panels on tuna–dolphin and shrimp–turtle heard the complaints, and decided trade restrictions were inappropriate instruments to protect endangered species. After NGOs’ lobbying and criticisms of the WTO, these decisions were reversed on appeal. After a thorough review (chapter 4), Sampson expresses doubts about these appeal decisions. The difference between the panel decisions and the judgments on appeal was that the former were decided in accordance with GATT/ WTO trade rules, whereas the appeal lawyers chose to take a broader view to accommodate environmental interests, based on the Preamble to the Uruguay Round Final Act and their interpretation of GATT article XX covering general exceptions. These ambiguities remain open for future appeals. Trade plays a major role in economic development, but lately the rationale of development theory has become lost in political aspirations about income redistributions, social justice, and charity in this age of popular compassion. Providing trade preferences to promote poor countries’ exports may promote development, but domestic trade liberalization offers wider opportunities for specialization and productivity gains across the whole economy, not just the export sector. This is seldom acknowledged in the current development debate. Sampson proposes that more cooperation between the WTO and UN agencies could achieve ‘sustainable development’. However, cooperation within the United Nations’ family of agencies, with NGOs as members of advisory commissions, is a tough proposition that would weaken further the consensus-based, inter-governmental WTO. The WTO’s decision processes are based on national sovereignty and policy autonomy to pursue economic and social goals. By admitting this in the introduction, the author announces his concerns about the UN proposals for global governance and ‘participatory democracy’, which have become a mantra in annual reports by UN agencies and in World Commissions ’ reports. In the introduction he claims, ‘I do not advocate a greater role for sustainable development in the WTO. ’ However, when he reviews outstanding WTO problems in detail, he waivers. (For a review of global governance and ‘participatory democracy’, see Robertson, 2006). While demonstrating detailed and comprehensive understanding of WTO committees and UN agencies dealing with the environment and development matters, this book is repetitive and becomes tedious in places. The conclusions in the final chapter are weak and especially confusing. The main proposal is that the WTO and trade



policies can contribute to environmental regulation without loss of economic efficiency. There is no attempt at cost–benefit analysis to justify this assertion. While acknowledging that the WTO is legally binding, which UN agencies are not, Sampson concludes, ‘ there is potential for mutually supportive and consistent policies between the WTO and specialized agencies of the United Nations ’. His support for this cooperation is founded on the Uruguay Round declaration on ‘Greater coherence in global economic policymaking’, which has brought cooperation between WTO, IMF, and World Bank officials. Consequently, his principal proposal is to establish another committee of UN agencies to ‘explore their cooperation’! The other three proposals in the final chapter are equally insubstantial: . another scholarly review by independent experts ; . combining CTE and CTD to study ‘sustainable development ’; . exploring new mechanisms to clarify WTO agreements, short of binding declarations. The striking feature of these proposals is their similarities with conclusions of various UN studies and World Commissions, where officials seek to by-pass governments. This would eviscerate the WTO and drag it down to UN levels. Perhaps such sacrifices are necessary to get a book published by the United Nations University ! Past GATT/ WTO successes have attracted interest from UN agencies and lobby groups that have been less successful in achieving their goals. Little thought is given to the reasons why one international agency has been effective where others have failed. One of the major reasons why the GATT/WTO has been effective is that its aims were limited ; to liberalize international trade by negotiating rules agreed by consensus among national governments. This depends on member governments’ willingness to negotiate as determined by their national economic policies. On the other hand, UN agencies vote on draft propositions, or seek reconciled texts, both of which tend to exacerbate differences or lead to ‘empty ’ agreements. Demands that the WTO’s work programme should be expanded, or that its management should be ‘democratized ’ by allowing NGO participation or some kind of representative (peoples’) council to advise the WTO Council, could only undermine the organization and its effectiveness. The unanswered question is ; would sacrificing the WTO enhance ‘sustainable development ’ ? DAVID ROBERTSON

Reference Robertson, D. (2006), International Economics and Conflicting Politics, Edward Elgar.


Economic Development and Multilateral Trade Cooperation By Evenett, Simon J. and Bernard M. Hoekman New York and Washington: Palgrave Macmillan and World Bank, 2006

Formally designating the mission of the current round of global trade negotiations as a ‘development round’ has proved to be a complex task for the more than 50-year



old global trade regime, traditionally focused solely on market access. Indeed, a development litmus test and an unprecedented number of developing countries taking part in the negotiations may prove to be the reason why the Doha round will become the longest round that ever was, or wasn’t. Simon Evenett and Bernard Hoekman have produced the best volume in the World Bank’s Trade and Development series so far. Unlike some of the past volumes, the editors include a wide range of authors and approaches. The volume is dominated by academics and negotiators from outside the Bank. What’s more, the volume addresses political economy and policy concerns in addition to the economic aspects of trade policy. The volume still falls short of being an all-in-one handbook for policy makers and scholars hoping to understand the politics and economics of trade and development. However, it goes a long way toward that end and should be required reading for anyone seriously engaged in current trade policy. In the editor’s introduction, they acknowledge that the developmental impacts of the Uruguay Round were less than optimal and ‘ highly skewed toward benefiting rich countries’. They single out the TRIPs agreement, which other World Bank studies show have cost the developing world $41.2 billion annually in patent rents, as particularly skewed in this manner (World Bank, 2002). They recognize these asymmetries as the impetus for a development round and argue that new, development friendly trade policies should have at least one of the following three components : . market access in developed countries for poorer nations, especially those that benefit the terms of trade of poor nations ; . lowering of domestic barriers to trade in the developing world; and . the establishment of complementary domestic policies and institutions in the developing world to support development. The editors also acknowledge that this is a difficult task from a political economy perspective because poorer nations have little to offer in terms of market access and concessions themselves in exchange for the dismantling of the remaining mercantilist policies of the developed world. Particular attention should be paid to the chapters on agriculture, given that this issue has formed the heart of controversy in the current round. Patrick Messerlin provides a good overview of the limited impact of OECD agricultural reforms under the Uruguay Round and some of the reasons for continued stalemate in agricultural negotiations. The former analysis is particularly useful in light of the ongoing tensions in the Doha Round, because it highlights the empirical evidence of the limited and shrinking benefits of farm policy reform in rich countries. Overall, the chapter still tends to take further liberalization in agriculture as a worthy end in itself. The empirical evidence does not necessarily support this view, suggesting that the main beneficiaries of US and EU agricultural policy reform would be a small number of wealthy or middle-income countries, while the majority of WTO members would gain little or would lose from the agreement. This is the one serious flaw in the chapter’s framework. Missing from the analysis are policy alternatives other than liberalization. For example, to the extent production-limiting policies can reduce overproduction and raise agricultural prices, there could well be benefits to a greater number of developing country farmers than would benefit from



liberalization. Most evidence suggests that liberalization would result in very modest decreases in Northern production and very small increases in prices (see Bouet, 2006). The author’s analysis of the political economy of protection provides a clear and illuminating look at the immense protections and lobbying efforts toward trade policy by the agriculture sector. Messerlin argues for rallying agribusiness to the cause of liberalization. This has limited potential, at least in the US, for one reason: Agribusiness is the largest beneficiary of current policies. Industrial consumers of commodities get unlimited supplies below production cost. Input suppliers get money pumped into the system to keep buying farm equipment, seeds, petrochemicals, etc. Finally, the author proposes a two-track farm policy based on large and small farmers, which is interesting, but WTO negotiations may well be better served by focusing on traded versus non-traded goods. Allowing nations full policy space for non-traded agricultural products, while severely limiting the subsidization of traded commodities, would be far more useful than focusing on small versus large farms. Kishore Gawande’s chapter on the political economy of agriculture contains a good review of the literature on lobbying activity, and its regression analyses confirm previous findings on the influence of agricultural lobbying, as well as its basis in quid pro quo lobbying rather than informational lobbying. The one finding reported as surprising by the author is interesting: that there is little association between protection and measures of trade. This suggests the important role of pressure over economic factors. Unfortunately, the author does little beyond demonstrating the obvious. Among the unasked questions in the chapter is how agribusiness firms, versus farmers themselves, behaved over time. One of the most commonly observed changes in advocacy within US farm policy is the rise of agribusiness to prominence over farmer groups. This change is widely seen as behind the major farm policy reforms from 1985 to 1996. Chapters by Joseph Francois, Will Martin, and Vlad Manole and by Alexander Kerk and Patrick Low address the development impacts of manufacturing tariff liberalization and special and differentiated treatment. They correctly show that political economy concerns have led to developed country resistance to granting market access to developing countries, especially in low skilled manufacturing sectors. They powerfully show how non-linear tariff reductions can help smooth the peaks in these protectionist holdouts in the developed world. However, the authors of both chapters also argue that developing countries should drastically reduce their domestic tariffs as well. Such advice does not clearly flow from the peer reviewed literature on the subject. There is considerable debate regarding the extent to which tariff liberalization is positively correlated with economic growth in developing countries. What is clear is that nations tend to liberalize trade as they get richer. This was clearly the case in the United States and most of Europe, in many East Asian economies from the 1970s to about 1990, and in China and Vietnam today (Rodrik, 2002). To the author’s credit however, they do not argue for throwing out differentiated treatment for developing countries, but rather argue that nations should have special treatment for specific policy measures rather than indexed wholesale to their level of development. Nevertheless, the majority of studies show that some countries will receive positive net benefits from liberalization. This leaves a ‘pareto possibility’ where the trade



‘winners ’ could compensate the ‘losers’. However, domestic politics often makes such compensation impossible. In addition, there are winners and losers across countries as well as within them – and no mechanism for cross-border compensations. The adjustment costs of liberalization can be substantial. UNCTAD estimates that the tariff revenue losses of a Doha deal for manufacturing alone would be approximately $63.4 billion for developing countries. In a recent article, Jagdish Bagwati has said ‘if poor countries that are dependent on tariff revenues for social spending risk losing those revenunes by cutting tariffs, international agencies such as the World Bank should stand ready to make up the difference until their tax systems can be fixed to raise revenue in other, more appropriate, ways’ (Bhagwati, 2005). While falling short of arguing for such a sweeping level of compensation, a rather sharp chapter by Susan Prowse recognizes the need for addressing the adjustment issue and makes the case that ‘aid for trade ’ can help nations make the necessary adjustments to make trade benefit their nations and offers a useful conceptual model for understanding the logic of aid for trade. Perhaps the most innovative chapter in the volume is by Jonathan Barton and Keith Maskus. These authors show how developing countries are increasingly losing access to knowledge because information traditionally held in the public domain is now restricted by intellectual property rules. These authors propose a new agreement for the WTO that would allow open access to basic science and technology. They remind us that knowledge is a public good and that by restricting access to such goods through tight intellectual property rules the trading system is denying developing countries the tools to innovate and spur innovation that could lead to more competition and higher levels of welfare in the future for developing and developed countries alike. The volume also includes must read chapters on investment incentives in developed countries, the merits of the single undertaking, and other issues. Even with its limitations, this book is essential. The volume provides policy-makers and trade scholars with accessible yet in-depth analyses of the relationship between global trade agreements and development. The volume is particularly strong in presenting the arguments that favor trade liberalization as a route to economic growth and development. Those who are interested in an alternative view should look elsewhere (see perhaps RIS, 2006). KEVIN P. GALLAGHER, Boston University

References Bouet, Antoine (2006), How Much Will Trade Liberalization Help the Poor? Comparing Global Trade Models, Washington, DC: International Food Policy Research Institute. Research and Information Service for Developing Countries (2006), World Trade and Development Report, 2006, New Dehli: RIS. Rodrik, Dani (2002), Global Governance as if Development Really Mattered, New York: United Nations Development Program. World Bank (2002), Global Economic Prospects, Washington, DC: World Bank.