The European Court ofJustice
and Direct Effect for the GATT:
A Question Worth Revisiting
Judson Osterhoudt Berkey*
Abstract
The European Court of Justice has long been criticized for consistently holding that the
General Agreement on Tariffs and Trade (GATT) does not have direct effect The end of the
GATT Uruguay Round prompted a renewed analysis of direct effect by Kees Jan Kuilwijk. In
his book The European Court ofJustice and the GATT Dilemma, Kuilwijk argues that the
continued denial ofdirect effect to the GATT 94 not only proves that the EC/ has protectionist
motives but also that it is unconcerned with individual rights. In addition to updating the
traditional critique of the Court's doctrine, Kuilwijk's book illustrates the tendency of that
critique tofail to acknowledge the full complexity of the direct effect question. Thus, a more
measured and thorough exploration of the legal political and economic Issues involved in
analysing the issue of direct effect may prove useful. Thispaper attempts such an analysis. Its
purpose is not to advocate a particular position but merely to illustrate the gaps in the
traditional critique.
1 Introduction
Whether the European Court ofJustice should permit enforcement ofprovisions of the
General Agreement on Tariffs and Trade within the European Community legal
system Is a difficult question to answer.1
In his book, The European Court ofJustice and
the Gatt Dilemma, Kees Jan Kuilwijk concludes that the Court should do so by means of
* Arent Fox, Klntner, Pbtkln & Kahn, Washington, D.C. I would like to thank two Individuals who
provided Invaluable assistance in the preparation of this article: first Professor Robert Z. Lawrence who
taught me concepts of international trade theory and provided useful comments on this paper during its
progress; second. Professor Joseph H. H. Weiler who not only helped me develop many of the ideas
contained In this paper but also taught me to think like a lawyer and a scholar.
1
With the entry into force of the Agreement Establishing the World Trade Organization, references to the
GATT are misleading and unldlomatic The term GATT now refers to the General Agreement on Tariffs
and Trade only and not the entire world trading structure as in the past Thus, the term 'GATT 47' in this
paper refers to the old GATT agreement and Its practice. The new GATT 1994 Treaty, which consists
European Journal ofInternational Law 9 (1998), 626-657
The European Court ofJustice and Direct Effect for the GATT 6 2 7
granting direct effect to the GATT 94.2
While Kuilwijk's book contains a sophisticated
legal, political and economic analysis of the direct effect question and is a worthwhile
addition to the prevailing academic view on this issue, its treatment of the relevant
concerns does not always fully elucidate their complexity.
This article thus attempts to illustrate the competing considerations facing the ECJ
in deciding whether to grant direct effect to the GATT 94. It will be shown that when
one thoroughly examines the legal, political and economic realities of the GATT 94 it
is no longer obvious, as Kuilwijk argues it is, that i) the ECTs jurisprudence requires
direct effect; ii) denying direct effect is detrimental to the Community's interests; and
ill) direct effect is necessary to protect individual rights. After a brief introduction to
the current ECJ doctrine and its critics, the paper will pursue three distinct lines of
analysis: doctrinal, motive and political economy. While the analysis is motivated by
Kuilwijk's book, the discussion will often explore issues beyond the bounds of his
critique.
A Current State of the ECJ Doctrine
Four principles govern the domestic legal effect of international agreements. These
principles are the following: direct application, direct effect, supremacy and interpretation.
The ECJ developed a common conception for each of these principles, with the
exception of direct effect.3
Although the Court granted direct effect to other
international agreements, it held in International Fruit Company, an Article 177b
preliminary reference decision, that individuals could not enforce GATT 47 provisions
because the agreement lacked direct effect ,
The Court reached this conclusion based on a consideration of the 'spirit, general
scheme and the terms of the General Agreement'.4
The Court held that, because the
largely of the GATT 47 as amended and several related understandings, will be referred to simply as the
GATT 94. Other agreements that were part of the Uruguay Round Agreements, such as the Agreement
on Safeguards, the Understanding on Rules and Procedures Governing the Settlement ofDisputes (DSU),
or the Agreement on Trade-related Aspects of InteJlectual Property Rights (TRIPs), will be referred to
Individually and separately.
K. J. Kullwl)k, The European Court of Justice and the GATT Dilemma (1996).
For example, the ECJ adopted a monist conception ofdirect application so that International agreements
and Community lawformpart of a single legal system. See Case 181/73, R.&V. Haegeman v. Belgian State,
[1974] ECR 449, and Joined Cases 21-24/72, International Fruit Company NV and Others v. Produktschap
voor Groenten en Fruit, [1972] ECR 1219. The ECJ also held that International agreements are supreme
over all secondary Community law. See Case 104/81, Hauptzollamt Mainz v. CA. Kupferberg& Cie. Kga.A.,
[1982] ECR 3641, and International Fruit Company. Finally, the ECJ held that similarly worded provisions
contained In both the EC Treaties (the European Community (EC) Treaty and the Treaty on European
Union (Maastricht Treaty) will hereinafter be referred to collectively as the Treaty') and Internationa]
agreements do not have to be Interpreted In the same manner. See Case 270/80. Polydor Umttedand RSO
Records Inc v. harlequin Record Shops Limited and Simons Records limited, [1982] ECR 329. and Case
70/87, EEC Seed Crushers' and Oil Processors' Federation (FFD101 m) v. EC Commission, [1989] ECR
1781.
International Fruit Company, supra note 3. at 1227.
628 E/ZL 9 (1998), 626-657
GAIT 47 'is based on principles ofnegotiations undertaken on the basis of "reciprocal
and mutually advantageous arrangements," [and] is characterized by the great
flexibility of its provisions, in particular those conferring the possibility of derogation,
the measures to be taken when confronted with exceptional difficulties, and the
settlement of conflicts between the contracting parties', it did not provide individuals
with rights which could be invoked in national courts.5
The ECJ also held that Member States could not enforce the GATT 47 provisions in
Article 173(1) actions before the Court.6
The ECJ used International Fruit Company as a
precedent in this decision to conclude that 'those features of 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'.7
Thus, the Court consistently held that the nature
and structure of the GATT 47 precluded any ofits provisions from having direct effect.
Although the ECJ denied direct effect to the GATT 47, It did allow some of the
agreement's provisions to have legal significance within the Community. The Court
held that GATT 47 provisions could be used to interpret the meaning of Community
legislation which expressly referred to those provisions. For example, in FEDIOL m the
ECJ held that the GATT 47 Article 3 prohibition against discriminatory taxes could be
used to interpret the meaning of 'illicit commercial practices' under the Community's
New Commercial Policy Instrument Regulation because this regulation required the
Community to comply with its international obligations. The Court distinguished its
previous direct effect holdings by stating that 'the GATT provisions have an
independent meaning which, for the purposes of their application in specific cases, is
to be determined by way of interpretation'.8
In the end, however, the Court found that
the contested measures did not constitute an illicit commercial practice.
The Court also held that GATT 47 provisions can be used to interpret Community
legislation when that legislation implemented a specific GATT 47 provision.9
For
example, in Nakajima, the Court held that the GATT 47 Anti-dumping Code could be
used as grounds for reviewing the legality of an anti-dumping margin determined
under the Community's Basic Anti-dumping Regulation. In a manner similar to its
decision in FEDIOL IH, the Court held that this was possible because the regulation
'was adopted in accordance with existing international obligations, in particular
those arising from Article VI of the General Agreement and from the Anti-Dumping
Code'.10
Again, however, the ECJ found that the substantive provisions of the
Anti-dumping Code had not been violated.
Because the ECJ has not addressed the direct effect question since the signing of the
Uruguay Round Agreements, the above decisions constitute the Court's doctrine
Case C-280/93, Federal Republic of Germany v. Council of the European Union, [1994] I ECR 4973.
Ibid, at 5073.
FEDIOL m, supra note 3, at 1831.
Case 69/89. Nakajima All Precision Co 1U v. Council of the European Communities, [1991] I ECR 2069.
ft/d. at 2178.
The European Court of Justice and Direct Effect for the GATT 6 2 9
towards the GATT 94. Therefore, the current ECJ doctrine continues to hold that the
GATT 94 lacks direct effect and only has legal significance when Community
legislation expressly refers to, or is intended to implement, specific GATT 94
provisions.
B Current Criticisms of the ECJ Doctrine
The ECJ's GATT direct effect doctrine has been criticized on many grounds. In fact,
Kuilwijk claims that it is 'one of its [the ECJ's] most seriously criticized doctrines
ever'.11
Kuilwijk's book provides one of the most sophisticated critiques to date. In the
mode of the other critics of the Court, Kuilwijk first argues that the GATT 94 is not
significantly different in nature and structure from the other international agreements
to which the ECJ has granted direct effect Thus, he concludes that it is
inconsistent to deny direct effect to the GATT 94. Second, he argues that granting
direct effect to the GATT 94 would force the Community to adopt a rule-based liberal
economic foreign trade policy which would maximize the economic welfare of the
Community.12
He therefore concludes that the Court harms the Community by
continuing to deny direct effect
Kuilwijk's arguments, however valid and informative, are often one-dimensional
11
See, e.g., the many writings of Emst-UWch Petersmann. In The EEC as a GATT Member — Legal
Conflicts between GATT Law and European Community Law', In M. Hllf, F. G. Jacobs and E.-U.
Petersmann (eds), The European Community and GATT (1986), Petersmann provides a broad critique of
the ECJ's policy that stresses the failure of the Court to apply general principles of Community law to the
GATT 47. 'Strengthening the GATT Dispute Settlement System: On the Use ofArbitration In GATT, In M.
Hllf and E.-U. Petersmann (eds). The New GATT Round ofMultilateral Trade Negotiations (1991), offers a
balanced treatment of the political economy function of International trade rules and the need to allow
Individual enforcement Finally, In 'National Constitutions and International Economic Law', In M. Hllf
and E.-U. Petenmann (eds), National Constitutions and International Economic law (1993), Petersmann
argues that the GATT 47 protects Individual economic rights and should be constltutlonallzed by means
of direct effect
For other critical perspectives, see Tumllr. 'GATT Rules and Community Law — A Comparison of
Economic and Legal Functions'. In Hllf. Jacobs and Petersmann, supra this note, who offers an analysis of
the economic effects of trade protectionism, and Bourgeois, Trade Pollcy-Making Institutions and
Procedures In the European Community'. In Hllf and Petersmann (1993), supra this note, who provides a
thorough examination ofthe application ofgeneral principles ofCommunity law to foreign trade. See also
Hllf, The Application of GATT within the Member States of the European Community, with Special
Reference to the Federal Republic of Germany', In Hllf. Jacobs and Petersmann. supra this note, for a
succinct list of the arguments for and against direct effect
With regard to the Uruguay Round Agreements, see Komuro, The WTO Dispute Settlement
Mechanism: Coverage and Procedures of the WTO Understanding'. 29 Journal of World Trade (1995) 5.
for an analysis of the new Dispute Settlement Understanding. See also Kul)per. The New WTO Dispute
Settlement System—The Impact on the European Community', 29 Journal ofWorldTrade (1995) 49, and
Lee and Kennedy, The Potential Direct Effect of GATT 1994 In European Community Law', 30 Journal of
World Trade (1996) 67, for analyses which caution against granting direct effect to the GATT 94. In
general. Kuilwt)k's analysis of the GATT 94 follows Petersmann's critiques.
12
A third line of Kuilwijk's argument Involves theories of political economy and legal rights. This
argument however, seems to collapse Into the other two as it Is used to provide philosophical support for
the notion that the lack ofenforcement of the GATT 94 obligations allows the Community institutions to
adopt protectionist trade policies which result in welfare losses for the Community as a whole and
unjustly harm Individuals.
630 SJIL 9 (1998), 626-657
and coloured by a conspiratorial tone that ultimately weakens his critique. For
example, he states that the ECJ denied direct effect to the GATT 47 in order 'to support
a "secret bond" between the Commission and the Council, on the one hand, and
import-competing Community producers, on the other'.13
He argues that the Court's
concern over reciprocity in the GATT 47 reflected a 'persistent adherence to the
traditionalist mercantilist, negative and defensive view of foreign trade policy'.14
Thus, he concludes that the Court should grant direct effect to the GATT 94 in order to
provide the 'divine guidance ... needed to attain economic paradise'.15
Given
Kuilwijk's tendency to slip into such one-sided advocacy, a more balanced treatment
of the direct effect question may be useful.
2 Doctrinal Analysis
One of the main criticisms of the ECJ's GATT 47 direct effect doctrine was that it was
inconsistent with the Court's doctrine regarding other international agreements. An
analysis of the cases in which the ECJ has dealt with other international agreements
reveals that the Court has examined four principle elements when deciding whether
such agreements are capable of having direct effect16
These elements are the
following: i) reciprocity in the initial balance of the obligations established by the
agreement; il) reciprocity in the ability to enforce the obligations established by the
agreement; ill) the possibility of derogating from the obligations established by the
agreement; and iv) the method of dispute settlement established by the agreement
International Fruit Company and the confirming line of cases17
show that the ECJ
denied direct effect to the GATT 47 based upon a consideration of these same four
elements. Critics of the Court's doctrine have argued that because the nature and
structure of the GATT 47 did not differ appreciably from the other international
agreements with respect to these four elements there was no justification for using
them as a reason to deny direct effect to the GATT 47.18
Kuilwijk attempts to
" Kuilwijk, supra note 2, at 247.
" Ibid, at 340.
15
Ibid, at 28.
" In his study, Kuilwijk examines the following five cases: Case 87/75. Concerto Danldc Bresdani v.
Ammlnistrazione ltallana delle Finame, [1976] ECR 129, Kupferberg, supra note 3, Case 17/81, Pabsl &
Rlchan KG v. HauptzoUamt Oldenburg, [1982] ECR 1331, Case C-192/89, SZSevlnce v. Staatssecretaris Van
Justltlt, [1990] I ECR 3461, and Case C-18/90, Office national de 1'emploi (Onem) v. Bahia Kziber, [1991] 1
ECR 199. His analysis relies most heavily on Kupferberg and Sevtnce.
17
Case 9/73, Corf SchluUr v. HaupUoUamt Lorrach, [1973] ECR 1135. Case 126/78, Douaneagenl der NV
NedeHandse Spoorwegen v. InspecteurderInvoerrechten en acdjmen, [1975] ECR 14 39, Case 266/81, Sodeta
ltallana per l'Otcodotto Transalpine (S1OT) v. Mlnistero delle Flnanze and others, [1983] ECR 731. Joined
Cases267-269/81, Ammlnistjvziom delle FinamedelloStaUiv.SoxieUiPetrouferaltaliane SpA (SP1) andSpA
Mlchelin ltallana (SAM1), [1983] ECR 801. Joined Cases 290-291/81. Compagnia Singer SpA and Gdgy
SpA v. Amminlstrazione delle Finame dello Stale, [1984] ECR 847. FED10L m, supra note 3. Nakajlma, supra
note 9, NMB (Deutschland) GmbH and others v. EC Commission and another, [1992] I ECR 1689, and
Germany v. Council supra note 6.
" See, e.g.. Maresceau, The GATT in the Case-Law of the European Court of Justice', In Hilf, Jacobs and
Petersmann, supra note 11, and Pescatore, The GATT Dispute Settlement Mechanism; Its Present
Situation and Its Prospects'. 27 Journal of World Trade (1993) 5.
The European Court of Justice and Direct Effect for theGATT 631
strengthen this criticism by arguing that the existence of the new Agreement on
Safeguards, which limits the ability of the contracting parties to derogate from their
GATT 94 obligations, and the increasingly judicial nature of the Understanding on
Rules and Procedures Governing the Settlement of Disputes (DSU) result in there
being even less reason for the Court to deny direct effect to the GATT 94. Indeed, he
claims that a continued denial of direct effect shows that 'the provisions of GATT are
instead treated as the "children ofa lesser God", not worthy of the Court's attention'.19
A GATT Reciprocity May Differfrom That in the Other International
Agreements
Although it is certainly true that the GATT 47 and the other international agreements
were similar in many important ways, it may not be entirely correct to state that there
were no reciprocity differences between them which justified different direct effect
doctrines. While the other international agreements were designed to implement
various stages of economic integration between the Community and its treaty
partners, the GATT 47 was structured to create a mechanism for negotiating
non-discriminatory and reciprocal tariff reductions among its contracting parties.
Thus, the nature of the other international agreements may have been more
conducive to direct effect than the GATT 47.
In Kupferberg, for example, the ECJ considered the Free Trade Area Agreement
between the Community and Portugal which eliminated all tariffs and other barriers
to trade between the two parties. In Pabst & Richarz and Sevinct, the Court considered
the Greek and Turkish Association Agreements, respectively, which created customs
unions, a higher form of integration than a free trade area, between the parties.
Finally, the Yaounde Convention, considered by the ECJ in Brcsciani, provided for
non-reciprocal duty-free entry into the Community of goods from the ACP countries.
These agreements, which established rights and obligations similar to those existing
between the Member States ofthe Community, all created higher forms of integration
than the GATT 47.
The Community and its treaty partners in the other international agreements also
had historical ties which did not exist between the contracting parties ofthe GATT 47.
For example, at the time that the ECJ was considering the free trade area agreement in
Kupferberg, Portugal had already applied for membership in the Community. In Pabst
& Richarz, the ECJ made explicit references to the fact that the association agreement
was intended to prepare Greece for entry into the Community. The ACP countries in
the Yaounde Convention were by definition those which had historical ties to
Community Member States. Even Kuilwijk admits that these historical and potential
future relationships between the Community and its treaty partners may have
" KuUwiJk. supra note 2. at 343.
632 E]H 9 (1998), 626-657
resulted In an 'extra-territorial solidarity' or an 'anticipated solidarity' which did not
exist between the GATT 47 contracting parties.20
When dealing with the GATT 47, in contrast, the Court always described the
agreement's purpose by referring to its preamble which stated that it was an
agreement for 'entering into reciprocal and mutually advantageous arrangements
directed to the substantial reduction of tariffs and other barriers to trade and to the
elimination ofdiscriminatory treatment in international commerce'.21
Thus, the Court
noted that the GATT 47 contracting parties only agreed to reduce tariffs and to
eliminate discrimination between them and not to pursue economic integration. The
GATT 47 did not anticipate the creation of a single market like that ofthe Community,
but rather allowed the contracting parties to preserve a diversity of market structures
as long as they did not unfairly differentiate between contracting parties. Because the
GATT 47 did not attempt to integrate disparate legal and economic systems like the
other international agreements did, It may not have required the enhanced
enforcement and uniform application of its provisions that would have resulted from
granting direct effect
Finally, one must recognize that the practical position of the Community in the
GATT 47 was different than its position in the other international agreements. The
other international agreements Involved countries which were in a much weaker
bargaining position than the Community. Because these agreements allowed the
parties to withdraw from the agreement if they so desired, the EC could use the threat
of doing so, and the resulting loss of treaty benefits for the other party, as a means of
forcing compliance with the agreement's obligations. Thus, the Community had a real
ability to influence the policies of the other parties to these agreements even when
these parties did not grant direct effect to the agreement
In the GATT 47, however, the Community was not in as strong a position to
influence the policies of the other contracting parties. The United States and Japan, the
two other large contracting parties, had an equal amount of bargaining power as the
Community. The Community could not threaten to withdraw from the system if it felt
its rights were not sufficiently respected because such a threat lacked credibility. Thus,
the Community was limited to working within the GATT 47 system in a way that it
was not in the other international agreements. In this case, it may have been
appropriate for the Court to insist upon explicit reciprocity from the other contracting
parties before granting direct effect to the GATT 47.22
The reasons for being cautious about drawing comparisons between the direct
effect granted to the other international agreements and direct effect for the GATT 47
Although Kullwijk does recognhe that these relationships may distinguish the reciprocity In the other
International agreements from that found In the GATT 47. he contends that Kupferberg 'seems to Indicate
that the Court does not attach too much Importance to the different "reciprocity" aspects of the legal
nature of an International agreement when considering its capacity to produce direct effect'. Kullwijk.
supra note 2, at 121.
Emphasis added. See. e.g.. International Fruit Company, supra note 3. at 1227, Schhiter, supra note 17. at
1157, and Germany v. Council, supra note 6. at 5072.
Lee and Kennedy, supra note 11, at 83.
The European Court of Justice and Direct Effect for the GATT 633
also apply to the GATT 94. The GATT 47 preamble, which remains the preamble to
the GATT 94, conveys the message that the world trading system, as structured by the
Uruguay Round Agreements, is intended to function as a forum for negotiating
multilateral negotiations directed at trade liberalization. And while it is true that the
Uruguay Round did produce some agreements requiring harmonization, most
noticeably the Antidumping Code, the Agreement on Subsidies and Countervailing
Measures, and the Agreement on Trade-Related Aspects of Intellectual Property
Rights (TRIPs), it is also true that those agreements contain provisions which seem to
require national laws that provide for individual rights.25
Hence, one must be careful
to maintain a distinction between those Uruguay Round Agreements which appear to
be Intended to have something akin to direct effect and the GATT 94 itself.
Ofcourse, the existence ofthese other Uruguay Round Agreements could be cited as
support for the argument that the ECJ should grant a form of conditional direct effect
to the GATT 94 whereby direct effect in any specific case would depend on whether
the other countries involved had also agreed to grant direct effect. However, as will be
argued below, granting even conditional direct effect might deprive the Community of
rights which it possesses under the GATT 94. Thus, there are still substantial
differences between reciprocity in the GATT 94 and the other Internationa]
agreements, and Indeed even the other Uruguay Round Agreements, which show
that the Court may not be deliberately treating the GATT 94 as the child ofa lesser God
but merely as a different species of agreement.24
B The Structure of the GATT Safeguards May Not Contemplate Direct
Effect
The ECJ always noted the flexibility of the GATT 47 Article 19 safeguard clauses,
which allowed contracting parties to suspend GATT 47 concessions in response to
increased imports that 'cause or threaten serious injury to domestic producers',
among its reasons for denying direct effect25
Critics have claimed that the Court acted
inconsistently in its analysis of the GATT 47 because it was not bothered by similar
safeguard clauses when granting direct effect to other international agreements.26
Kuilwijk goes further than the Court's other critics in arguing that the ECfs
treatment ofthe GATT 47 safeguard clauses 'mistakes GATT law for GATT practice'.27
He contends that the safeguard clauses were never really flexible and led the
21
See, e.g., the Uruguay Round Agreements on Antidumping Article 13. on Subsidies Article 23. and
TRIPS Article 41.
u
It should be noted that Switzerland led an Initiative halfway through the Uruguay Round to require each
WTO member to give the GATT 94 direct effect or some equivalent status. In their national law. As
Kul|per, supra note 11. at 65. notes this 'would have assured equality between the parties In respect of
"Internal enforcement" of the GAIT. The fact that this was not Included In the final Uruguay Round
Agreements seems to Indicate, however, that the WTO members as a whole still do not desire direct effect
for the GATT 94.
" See. e.g., International Fruit Company, supra note 3. at 1227.
26
See. e.g.. Petersmann (1991). supra note 11. at 86-92. for a general critique of this part of the ECJ's
jurisprudence.
" Kullwtjk. supra note 2. at 342.
634 EJIL 9 (1998), 626-657
contracting parties to adopt grey area trade measures such as voluntary export
restraints (VER) and orderly marketing agreements (OMA). He claims that this was a
constitutional defect ofthe GATT 47 which could have been corrected by giving direct
effect to the agreement Given the new Agreement on Safeguards, he concludes that
there is no longer any excuse for denying direct effect for the GATT 94 because it now
'explicitly prohibits grey area trade restrictions and forms a true example of the new
firm commitment to adherence to the law, subscribed to by the Community'.28
While the ECJ's critics certainly do have valid criticisms, once again the issue is not
nearly as one-sided as presented. For example, an examination of the cases in which
the ECJ considered other international agreements shows that the safeguard clauses
in those agreements differed significantly from those found in the GATT 47 and in the
GAIT 94. In Kupferberg, for instance, the Court stated that the safeguard clauses in the
Portugal Free Trade Agreement applied 'only in specific circumstances and as a
general rule after consideration within the joint committee in the presence of both
parties'.29
In Scvince, the Court agreed with Advocate General Dannon's claim that the
safeguards in the Turkish Association Agreement could be invoked only in cases
where there was a threat of 'disturbances on [the] employment market which might
seriously jeopardise the standard of living or level of employment in a particular
region, branch of activity or occupation'.'0
Therefore, the Court appeared to attach
importance not only to the fact that the safeguards in the other international
agreements were applicable only after political consultation with the other treaty
party, but also that they could be used in specific circumstances.
Neither ofthese limitations was found in the GATT 47, nor is found in the GATT 94.
Article 19 of the GATT 47 only required notification of contracting parties which
would be affected by safeguard measures prior to their implementation in order to
afford those parties an opportunity to ask for consultation over compensation. Articles
12 and 13 of the Agreement on Safeguards do now require notification of the GATT
Committee on Safeguards when initiating an investigation, making a finding of
injury, or imposing a safeguard measure. However, nothing in the GATT 94 allows
the members or the Committee on Safeguards to prevent the implementation of a
safeguard measure by another member." Thus, it appears that GATT 94 safeguards
can be used without the same level of prior consultation as required by the other
international agreements.
GATT 47 Article 19 also required a threat of, or the existence of, 'serious injury'
before a contracting party could impose a safeguard measure. However, this
Ibid. It should be noted that GATT 47 Article 19 remains In force alongside the Agreement on Safeguards
subject to the condition that the Agreement on Safeguards controls In the case of any conflict
Kupferberg, supra note 3, at 3664.
Sevlnce, supra note 16. at 3486.
Agreement on Safeguards Article 13 does provide that the Committee on Safeguards will monitor the use
of safeguard measures and can review the equivalence between the effects of any safeguard measure and
the compensation offered. However, the Committee's findings have no legal status. Thus, these
requirements do not seem to be as strong as those found in Kupferberg where a safeguard measure could
not generally be imposed unless approved by the Joint Committee.
The European Court of Justice and Direct Effect for the GATT 6 3 5
determination was left to the discretion of the government imposing the safeguard
measure itself.32
Agreement on Safeguards Article 2 does now require a member to
conduct an open investigation to prove the existence ofinjury. Article 3, however, still
leaves the determination of 'serious injury' up to the competent authorities of the
member imposing the safeguard. Thus, members can use safeguard measures to
remedy a broader range of injuries and have a larger amount of discretion in
determining the need for safeguard measures than in the other international
agreements. Given these differences, it may not be entirely fair to the Court to claim
that its treatment of safeguards is inconsistent3
'
Because the ECJ could always grant direct effect to the GATT 94 while at the same
time holding that the Community had properly invoked a safeguard clause, one must
move beyond the doctrinal question of the ECJ's case law consistency to consider
whether the political economy ofthe safeguards regime really demands direct effect It
Is correct to state that, prior to the Agreement on Safeguards, the flaw In the GATT 47
safeguard regime was not that it was too flexible but rather that It was not flexible
enough. Contracting parties which did not want to pay compensation for, did not
want to meet the non-discrimination requirement of, and did not want to conduct an
investigation to justify a safeguard measure often adopted grey area trade measures
such as VERs and OMAs to remedy the injuries caused by increasing imports.M
These
measures effectively imposed quota systems on international trade and thus were
more economically l\armful than tariffs. It is not clear, however, how direct effect
could have prevented the use of these measures which were effectively sanctioned
under the GATT 47."
GATT 47 panel reports Indicated that deference was to be given to a contracting party's determination
that a safeguard measure was necessary. See, e.g., the 1951 Hatiers' Fur Cose, Sates No. GATT 1951-3
(Nov. 1951), In which US safeguards applied against Imports of Chechoslovakian hatter's fur were
upheld.
In fact. Advocate General Darmon In Sevince directly contrasted the safeguards In the Turkish
Association Agreement with those In the GATT 47. He noted 'In thefirstplace, that their overall context
Is more restrictive than that of GATT and, secondly, that they are expressly defined as exceptions to the
rule that the provisions in question are to be automatically applied, whereas those of GATT, which
appear to be In harmony with the generally "flexible" overall system of which they form part, could
hardly be classified as exceptions to a rule of automatic application which, In fact, does not form part of
that system'. Therefore, he seemed to acknowledge that the safeguards In the GATT 47 might be
sufficiently moreflexiblethan those in the other international agreements to justify the denial of direct
effect Sevince, supra note 16, at 3488.
For example. In the 1980s both the US and the Community negotiated VERs with Japan to limit auto
imports from Japan. Presumably, both the US and the Community could have Invoked the GATT 47
safeguard clauses but as noted, this would have required an Investigation to prove that their domestic
auto Industries were being harmed by the Japanese Imports and would have required them to
compensate Japan and any other contracting party for the Injury caused by any safeguard measures
Imposed.
Kuilwtjk does provide a useful Insight when he states that 'safeguard clauses In international agreements
on trade Uberallxation always represent a link between two conflicting objectives; first respect by
636 E7ZL 9 (1998), 626-657
The Agreement on Safeguards does represent a real step forward In that it brings
grey area trade measures such as VERs and OMAs within the discipline of the GATT
94 and effectively bans them with only a few exceptions. The fact that these measures
have been reigned in, however, does not automatically mean that direct effect is now
necessary to enforce compliance. In fact, one could say that the adoption of the new
safeguards regime indicates that once the members recognize that there is a problem
with the agreement's structure and muster the political consensus to address the
problem, they will act to do so.56
In this case, it may not be appropriate for the ECJ to
grant direct effect to the GATT 94 and force the Community to adopt unilateral
positions before a world political consensus exists. And as will be shown below,
granting direct effect to the GATT 94 may also deprive the Community of its rights
under the agreement. Thus, one must be very careful indeed not to confuse GATT 94
law with practice.
C The Nature of GATT Dispute Settlement May Not Require Direct
Effect
The dispute settlement mechanism of the GATT 47 was heavily criticized because it
did not require rule-based adjudication to resolve disputes and effectively allowed the
contracting parties to engage in power oriented diplomacy.37
Despite this, however,
the critics of the ECJ's direct effect doctrine claimed that because the GATT 47 dispute
settlement mechanism did not differ appreciably from those found in the other
international agreements to which the Court had granted direct effect, it should not
have been used as a Justification for denying direct effect38
governments for thdr commitments regarding trade liberalization, and, second, their concern to keep a
margin of manoeuvre thus enabling them to protect the domestic market through the Imposition of
restrictive measures when necessary'. Kuilwijk, supra note 2, at 136. However, the fact that the GATT 47
allowed safeguard measures may Indicate that the contracting parties were not able toJointly resolve this
conflict
See, van Themaat The Possibilities for National Measures of Implementation to Strengthen the Legal
Quality of International Rules on Foreign Trade', D. C. Dicke and E.-U. Petersmann (eds), Foreign Trade tn
the Present and a New International Economic Order (1988). at 45—46, for an analysis of why it Is more
politically realistic to undertake efforts to reduce national government discretion on an international
rather than a national level.
Of the 195 disputes brought to the attention of the contracting parties under GATT 47 Article 23
between 1948 and March 1995, only 98 resulted in the circulation ofpanel reports and only 81 of these
panel reports were actually adopted. Komuro, supra note 11.
See, R. E. Hudec, Enforcing International trade Law. The Evolution ofthe Modem GATTLegal System (1993).
lor a detailed description of the operation of the GATT 47 dispute settlement See Jackson, The Legal
Meaning of a GATT Dispute Settlement Report Some Reflections', In N. Blokker and S. Muller (eds),
Towards More Effective Supervision by International Organizations (1994) for general criticisms ofthe GATT
47 dispute settlement See van Themaat 'Strengthening the International Legal Framework of the
GATT-MTN System; Reform Proposals for the New GATT Round: A Comment', In Hllf and Petersmann
(1991), supra note 11, for an argument that direct effect should await the establishment of an
International trade court
The European Court ofJustice and Direct Effectfor the GATT 637
While the GATT 47 dispute settlement mechanism was rightly criticized for being
weak, there still may have been an important distinction between the GATT 47 and
the other international agreements which justified the ECJ's determination that it was
part of the reason for different direct effect doctrines. The GATT 47 was an agreement
designed to establish a system of multilateral rounds of reciprocal trade negotiations
across multiple sectors. Thus, the GATT 47 effectively created a continuing
relationship of negotiations between the contracting parties designed to propel the
process oftrade liberalization and resolve lingering disputes by means of cross-sectoral
bargaining.
This was not the case with the other international agreements. These agreements
created static systems that resulted from one-time negotiations over the rights and
obligations of the parties. Thus, direct effect may not have been appropriate for the
GATT 47 because it would have effectively allowed an individual to interfere with a
process of ongoing negotiations between the contracting parties: that is, direct effect
might have prevented contracting parties from blocking panel reports that settled
disputes in ways with which they disagreed and hoped to modify through negotiation.
While one may not approve of blocking panel reports in order to force negotiations to
resolve policy differences, it cannot be denied that this was a significant right which
each contracting party possessed under the GATT 47.
The Uruguay Round DSU significantly strengthened the GATT 47 dispute
settlement mechanism. The DSU did so by providing deadlines for all procedures, by
creating an Appeals Body, and, most importantly, by reversing the consensus-toadopt-panel-reports
approach of the GATT 47 to a consensus-to-block-panel-reports
approach.39
Given the changes introduced by the DSU, Kuilwijk argues that the
dispute settlement mechanism can no longer be characterized as flexible. He
concludes that because 'the dispute settlement system matured and now closely
resembles an actual judicial system [with a]firmcommitment to legalism' itshould no
longer be used by the ECJ as a justification for denying direct effect to the GATT 94.40
One must carefully examine the actual operation of dispute settlement under the
GATT 94, however, before concluding that the changes introduced by the DSU mean
that direct effect is now appropriate. Dispute settlement systems do not fall into one of
two polar categories, i.e. hard legal or soft political systems, but rather lie along a
continuum between the two. Although the DSU has shifted the GATT 94 system
towards the legal pole, it is not yet clear where it falls on the continuum. While it
certainly has made GATT 94 dispute settlement more legalistic, the DSU does not
appear to have replaced all intergovernmental negotiation with strict precedent-based
See Komuro, supra note 11, for a detailed description of the procedures of the DSU.
KuUwlJk. supra note 2, at 342.
638 E/IL9(1998), 62fr-657
application of GAIT 94 law.41
Thus, it may be premature to state, as Kullwijk does,
that the GATT 94 should be given direct effect because the DSU is now a true judicial
system.
In fact, the language of the DSU shows that negotiation still plays a large role in the
panel procedure. For example, DSU Article 3(3)'s call for the 'prompt settlement of
situations in which a Member considers that any benefits accruing to it directly or
indirectly under the covered agreements are being impaired'42
indicates that the
adoption of panel reports is not the only means of resolving a dispute. DSU Article 4
requires consultations between the disputants before one may request the establishment
of a panel to hear the dispute and Article 5 offers the use of good offices,
conciliation and mediation as an alternative means of resolving disputes.
DSU Article 12(7) provides that 'where the parties to the dispute have failed to
develop a mutually satisfactory solution, the panel shall submit itsfindingsin the form
of a written report to the DSB',43
This implies that negotiations between the disputants
may take place on a parallel track with a panel procedure. After the issuance ofa panel
report, DSU Article 16(1) provides for a 20-day waiting period before the report can be
considered for adoption by the DSB. This provides the disputing parties with another
opportunity to settle their dispute by negotiated agreement. Finally, even after a panel
report has been appealed and the Appeals Body has issued an appellate report, there is
still a 30-day waiting period before that report can be adopted by the DSB. This again
provides time for the disputing parties to negotiate a settlement of their dispute
without the adoption of a panel report
One should also realize that the DSU procedures are not the only means of settling
disputes within the GATT 94. It has been noted that there are over 30 different ways
to resolve disputes within the GATT 4:7.** Many GATT 47 provisions, such as Article
8(1) which requires a member to review the way it values goods for customs purposes
when requested to do so by another member, call for consultations between members.
Other GATT 47 provisions are framed to avoid disputes in the first place, such as
Article 25(5) which allows a member to seek a waiver from its obligations upon
approval of the other members. Thus, one should be careful not to focus solely on the
DSU when considering the nature of GATT 94 dispute settlement.
The current dispute over the US Helms-Burton Act illustrates some aspects ofGATT
94 dispute settlement that may caution against automatically assuming that direct
effect is now appropriate. After this Act took effect on 12 March 1996, pressure from
the Community and Canada forced President Clinton to delay enforcement of its
Sec e.g., Lee and Kennedy, supra note 11, at 81, for an analysis of thefirstdispute to come under the DSU,
a dispute between the US and Japan over cars and car parts. They conclude that the resolution of this
dispute by means of a negotiated agreement between the countries 'seems to suggest that, far from
becoming a strictly legal device, the dispute resolution system retains Its flexible nature and that
negotiation, rather than formal legal complaint, will remain the most common form of conflict
resolution'.
Emphasis added
Emphasis added.
M. J. Trebllcock and R. Howse, The Regulation ofInternational Trade (1995), at 386.
Pie European Court ofJustice and Direct Effectfor the GATT 639
provisions which granted US nationals a private right of action against foreign
companies that 'traffic' in property confiscated by the Cuban government from US
nationals.45
When negotiations to resolve the dispute broke down on 16 October
1996, the Community filed a WTO complaint against the USA claiming that the Act
violated both the MFN principle and the prohibition against quantitative restrictions.
The US argued in. defence that the Act qualified for the Article 21 national security
exemption from GATT 94 violations. Although negotiations between the two sides
continued, no compromise was reached by the time a panel was formed. At that time,
the USA stated it would boycott any panel because it did not 'believe anything the
WTO says or does can force the U.S. to change its laws'.46
The US stance, which quickly prompted a renewed effort to reach a compromise,
illustrates the strengths and limitations of the DSU. Both the US and the Community
realized that a panel resolution of the dispute was a 'lose-lose' proposition: that is, a
ruling for the US would threaten the WTO by encouraging other countries to defend
questionable trade-related policies under the national security exemption, while a
ruling for the Community would slow momentum for further trade liberalization by
weakening domestic political support in the US.47
This merely intensified the pressure
for a political settlement of the dispute.48
Thus, the DSU may actually encourage,
rather than inhibit, the use of political negotiation to resolve disputes in cases where
there is not strong support for the adoption of a panel report to settle the dispute.
In fact, negotiations eventually did lead the Community to suspend its WTO
complaint for six months on 11 April 1997 and again on 15 October 1997. The
Community agreed to do so in order to allow the Clinton administration time to seek a
Congressional change to the Helms-Burton Act. The Community, however, reserved
its right to revive the complaint ifno satisfactory result was achieved. Stuart Eizenstat,
the US Under Secretary of Commerce and chief US negotiator, has stated that 'this
understanding represents a beginning and not an end. There is absolutely no
President Clinton did not suspend the provisions which deny visas to executives and major shareholders
of companies which 'traffic' in property confiscated by the Cuban government from US nationals.
Sanger. 'U.S. Won't Offer Trade Testimony on Cuba Embargo'. NY Times. 21 February, 1997. Section A,
at 1.
Lawrence. 'A Cuban Dilemma: Competing Pressures from Home. Abroad Confront Clinton as He Nears
Deadline on Helms-Burton'. Journal of Commerce, 6 January 1997, at 32C.
After the US announcement Sir Leon Brittan. the Community's top foreign trade official, stated that the
WTO 'panel procedure can be halted at any time ifthe parties reach agreement'. WTO chairman Renato
Rugglero also stated that he hoped 'the two sides canfinda solution to the problem... [and] deal with the
dispute settlement procedure in a way which will facilitate agreement by the two sides'. Lawrence, 'U.S.
Will Snub WTO Panel on Anti-Cuba Law'. Journal of Commerce, 21 February 1997. at 1A.
640 E/H, 9 (1998), 626-657
certainty we will reach an agreement . . . . There is going to be hard bargaining
ahead.'49
As of the time of this writing the resolution of this dispute is still uncertain.
This dispute shows that a shift towards a more rule-based adjudication of disputes
requires a change in the attitude of the members' governments, not just a change in
the DSU.50
A more legalistic DSU does not change the fact that dispute settlement in
the GATT 94 still may involve political negotiations between members.51
Where the
GATT 94 has well-defined rules and disputes are narrowly focused, the DSU will
indeed work well. However, when there is no consensus on either the substance or the
application of the rules, the more legalistic DSU may not be able to save dispute
settlement from political negotiation.52
Thus, one should carefully consider the actual
operation of the DSU before claiming that dispute settlement would benefit from the
existence of direct effect.
3 Motive Analysis
A second line of criticism of the ECJ's direct effect doctrine has attempted to impugn
the Court's motive for denying direct effect. This criticism has largely been based upon
an analysis of the ECJ's treatment of the second element of reciprocity found in
international agreements, i.e., reciprocity in the enforcement of obligations. The ECJ's
critics have argued that it was improper for the Court to use the lack of absolute
reciprocity in the GATT 47 as a reason for denying direct effect because unilaterally
granting direct effect would have ensured that at least the Community strictly
adhered to its GATT 47 obligations.53
The critics have claimed that this would be
beneficial because it would have prevented the Community institutions from acting
arbitrarily in the context of foreign trade policy.54
Kuilwijk attempts to extend this critique by endowing it with moral significance. He
states that the ECJ's doctrine shows that it maintains a 'persistent adherence to the
de Jonquierres and Dunne. 'Helms-Burton Antagonists Keep Fingers Crossed', financial Times, 14 April
1997. at 6. While President Clinton agreed to continue to delay enforcement of the provision allowing
suits againstforeigncompanies which traffic In seized property, there was no movement on the provision
allowing the US to block the entry of officers and shareholders of companies which traffic in such
confiscated property. President Clinton stated that he would ask Congress to repeal this measure if the
Community concluded and Implemented a bilateral agreement to constrain the ability of companies to
Invest In illegally expropriated foreign assets in the future.
See. e.g.. Hudec supra note 38, at 363.
As of 9 May 1997, of the 24 cases brought by the UStothe new DSU, four had been decided by panels in
favour of the US. five had been settled, seven were still under consideration by panels, and dght were In
the early stages of negotiation. Ofthe 10 cases brought against the US. three had been withdrawn, three
had been decided by panels against the US, one had been settled, one was still under consideration by a
panel, and two were in the early stages of negotiations. Sanger. 'Qlnton Grants, Then Suspends. Right to
Sue Foreigners', NY Times. 9 May 1997, Section A. at 1.
See Trebllcock and Howse, supra note 44, at 392.
See. e.g.. Hllf, Jacobs and Petersmann. supra note 11.
See, e.g.. Hilf and Petersmann (1993). supra note 11.
The European Court of Justice and Direct Effect for the GATT 641
traditional mercantilist, negative and defensive view of foreign trade policy'.55
He
argues that the ECJ focuses on the reciprocity of enforcement in order to disguise a
desire to allow the Community institutions maximum flexibility in creating foreign
trade policy. He concludes that this harms the Community because 'the true reason
that the political institutions do not want to have their hands tied is that they wish to
retain the freedom to confer special favors on specific industries, not because they are
convinced that they are capable of better serving the public interest by retaining such
freedom'.56
A Possible Non-Mercantilist Motivations for the ECJ's Doctrine
Before accusing the Court ofundue protectionism, one must determine ifthere are any
valid legal reasons for denying direct effect First, it must be acknowledged that the
reciprocity of enforcement in the GAIT 94 is much different from that in the
Community or the other international agreements to which the Court has granted
direct effect57
The US has explicitly stated that the GATT 94 will not have direct effect
in US courts.58
It also appears that direct effect does not exist in Japan.59
Because both
the US and Japan, the two other large players in the dispute settlement mechanism
" Kullwijt supra note 2, at 340. One should recognize, however, that the language and structure of the
GATT 47 Itself Is mercantilist Within GATT 47 parlance, reductions oftariffs and other trade barriers are
termed 'concessions'. Because these concessions are made from the current levels of protection without
regard for harmonliatlon, It cannot be said that the GATT 47 Is intended to create a level playing field.
Therefore, It may be wrong to assume that the agreement's provisions always maximize economic
welfare and that the ECJ must have mercantilist motives for denying direct effect The relationship
between economic theory and direct effect will be examined In more detail In Section 4A.
" Ibid, at 341.
" In the Community legal setting, the ECJ has held that a Member State cannot escape a breach ofits Treaty
obligations by claiming that either another Member State or a Community institution has breached Its
Treaty obligations. See. e.g.. Case 232/78. Commission v. France, [1979] ECR 2729. and. Cases
90-91/63, Commission v. Luxembourg and Belgium, [1964] ECR 625. The Court dealt explicitly with the
issue ofreciprocityofenforcement In regard to the other international agreements only In Kupferberg. The
reciprocity Issue arose In this case because a Swiss court had denied direct effect to a free trade area
agreement similar to the EC-Portugal one. The Court concluded that the structure of the EC-Portugal
free trade area agreement in which a Joint Committee engaged In political resolution of all disputes, did
not contain a lack of reciprocity. Kupferberg, supra note 3. at 3650.
M
The US implementing legislation, 19 U.S.C. §102(c). provides the following:
No person other than the United States (A) shall have any cause ofaction or defense under any of
the Uruguay Round Agreements or by virtue ofcongressional approval of such an agreement or
(B) may challenge. In any action brought under any provision of law, any action or Inaction by
any department agency, or other instrumentality of the United States, any State or any political
subdivision of a State on the ground that such action or Inaction is inconsistent with such
agreement
" While Article 98(2) of the Japanese Constitution states that International agreements must be respected,
a 1984 Kyoto District Court held that a GATT 47 violation by the Japanese government merely triggers
the right of other countries to demand negotiations and does not Invalidate domestic law. This decision
was affirmed by theJapanese Supreme Court which held that 'a violation of a provision ofGATT pressures
the country In default to rectify the violation by being confronted with a request from another member
country for consultation and possible retaliatory measures. However, it cannot be interpreted to have
more effect than this. Therefore. It cannot be held that the legislation In question Is contrary to the GATT
and null and void.' J. H. Jackson, W. J. Davey and A. 0. Sykes. Jr.. legal Problems ofInternational Economic
Relations (3rd ed.. 1995). at 224-226.
642 E/ZL 9 (1998), 626-657
besides the Community,60
explicitly deny direct effect, this position may represent the
common understanding of the members about the nature of the GATT 94 agreement
Second, it should be acknowledged that the dispute settlement mechanism actually
utilizes the lack of direct effect to help resolve disputes. As previously noted, the lack of
direct effect allows members to settle a dispute by means of a negotiated agreement
prior to a formal ruling on the validity of trade measures. This is useful because it
allows member-to-member trade-offs and cross-sectoral bargaining over concessions
which would not occur if the dispute were resolved by means of a panel report or by
national courts applying GATT 94 law." Thus, direct effect might actually inhibit the
operation ofthe GATT 94 by denying the members some of the flexibility contained in
the DSU.
Third, one must recognize that direct effect may actually deprive the Community of
rights which it possesses under the GATT 94 after the issuance of an adverse panel
report. Ifthe ECJ required the withdrawal ofa Community measure because it violated
the GATT 94 before a panel had even decided that the measure actually did violate the
agreement, then the Community would be deprived of the right to choose the method
of remedying the injury caused by the measure: that is, a Court ruling by means of
direct effect would mandate specific performance as the only remedy for all GATT 94
violations and deny the Community the ability to offer compensation as a remedy for
such violations. This limitation on the policy discretion of the Community might be
particularly harmful in the context of the GATT 94 because the agreement has such
broad economic consequences.62
This possible asymmetry between Community
enforcement of the GATT 94 and individual enforcement is reinforced by DSU Article
23(1), which requires members to resort to the DSU to redress any violations of the
GATT 94.
There is some debate over whether a member which has been found to be in
violation of a GATT 94 obligation actually may choose the method of remedying its
Ofall the complaints brought between 1948 and 1989, the US. Japan and the Community, or Its Member
States, were Involved as plaintiffs 62% ofthe time and as defendants 76% of the time. Hudec. supra note
38.
Lee and Kennedy, supra note 11, at 88. argue that allowing an Individual Importer to successfully
challenge the validity of an agreement settling a dispute such as the US-Japan agreement on cars and car
parts could lead to a case of double Jeopardy because 'a state having already made concessions In other
fields may then also be forced to withdraw the measure Impugned by Its national court'. They conclude
that this would then decrease the Incentive to settle disputes through negotiated agreements and might
threaten the functioning of the GATT. a. HUf and Petersmann (1991), supra note 11. for an argument
that reciprocity Is a bilateral concept which should not be used In a multilateral treaty such as the GATT.
See Kuijper. supra note 11. at 64. for a discussion of why a country which specifically excludes direct
effect when others do grant direct effect 'places Itself In such a favorable position that It becomes
fundamentally unfair to Its trading partners'.
The European Court of Justice and Direct Effect for the GATT 6 4 3
violation.63
Judith H. Bello has argued that the GAIT 94 preserves the GATT 47
principle that
the only sacred. Inviolable aspect of the GATT [Is] the overall balance of rights and obligations
of benefits and burdens, achieved among members through negotiations ... [so that] a
government could renege on Its negotiated commitment not to exceed a specified tariff on an
Item, provided it restored the overall balance of GATT concessions through compensatory
reductions in tariffs on other Items.64
Therefore, she concludes that a member found to be violating the GATT 94 may do
one of the following: i) withdraw the offending measure; ii) maintain the measure but
provide compensatory benefits; or ill) decline to withdraw the measure or provide
compensation and instead allow the injured member to retaliate.65
In response, John H. Jackson has argued that the withdrawal of the offending
measure is the proper method of remedying GATT 94 violations.66
He notes the
historical practice of 'the GATT contracting parties [to treat] the results of an adopted
panel report as legally binding' and argues that 11 clauses of the DSU support the
notion that a panel report Imposes an international law obligation on the member to
Petersmann has argued extensively that International law requires the specific performance of all
International obligations, including WTO obligations. In good faith. See e.g., E.-U. Petersmann,
Constitutional Functions and Constitutional Problems of International Economic Law. Foreign Trade Law and
Policy In the USA, the EC, and Switzerland (1993); Petersmann. 'Proposals for a New Constitution for the
European Dnion', 32 CMLR (1995) 1123; Petersmann. 'Constitutionalism and International Organizations'.
17 Northwestern Journal ofInternational Law and Business (1996/97) 398.
However, it should be acknowledged that compensation has been recognized as a valid remedy in some
International law contexts. This has occurred particularly In cases involving the breach of a concession
agreement For example, dictum of the Permanent Court ofJustice In Chorzow Factory, 1927 PQJ, Series
A, No. 9, first suggested that while restitution was the preferred remedy, compensation would serve as a
valid secondary remedy ifrestitution were not possible. Although the arbitrator In Texaco v. Libya, 53ILR
(1974) 389. equated restitution with specific performance, the eventual remedy in this case was the
payment of compensation. And In BP v. Libya, 53 ILR (1974) 296, the arbitrator rejected the primacy of
specific performance and ordered the payment of compensation.
Furthermore, Derek Bowett has argued that the US-Iranian Claims Tribunal decisions explicitly
establish the Internationa] law principle that the remedy for the breach of a state contract with an alien Is
compensation. Bowettclaims that there Is a distinction between the remedies for unlawful takings, lawful
ad hoc takings, and lawful takings that are part of a general programme of nattonalliation. In the case of
an unlawful taking, Bowett states that the appropriate remedy Includes all potential profits up to the date
of the judgment The remedy for ad hoc takings Is the value of the lost going concern, Le. the value of all
tangible and Intangible assets, but does not include lost profits. Finally, the remedy for a taking that Is part
of a nationalization programme Is the value of the tangible assets lost Bowett 'State Contracts with
Aliens; Contemporary Developments on Compensation for Termination or Breach', 59 BYbIL (1988) 49.
These principles might be relevant Ifone viewed a GATT 94 violation as equivalent to the taking of one
nation's going concerns by another.
BeDo. The WTO Dispute Settlement Understanding: Less Is More', 90 AJJL (1996) 416, at 417.
Ibid. See also Roessler, 'Diverging Domestic Policies and Multilateral Trade Integration'. In J. Bhagwatl
and R. E. Hudec (eds), Fair Trade and Harmonization. Vol 2, Legal Analysis (1996), at 47. who argues that
an integrated dispute settlement system under which violations of commitments in one sector can be
remedied by retaliatory action In another helps preserve the entire negotiated package of concessions.
Jackson, 'Editorial Comment The WTO Dispute Settlement Understanding — Misunderstandings on the
Nature of Legal Obligation', 91 AJ1L(1997) 60.
644 EJ1L 9 (1998), 626-657
which it is directed to perform the recommendations of the paneL67
Thus, he
concludes that the GATT 94 'clearly establishes a preference for an obligation to
perform the recommendation ... and provides that in nonviolation cases, there is no
obligation to withdraw an offending measure, which strongly implies that in violation
cases there is an obligation to perform1
.68
While the language of the DSU does establish a preference for specific performance,
it also supports the notion that compensation is a possible remedy. DSU Article 22(1)
states that 'compensation and the suspension of concessions or other obligations are
temporary measures available in the event that the recommendations and rulings [of
the panel] are not implemented within a reasonable time' and that 'neither
compensation nor the suspension of concessions or other obligations is preferred to
full implementation of a [panel] recommendation'. DSU Article 21(3)—(4) establishes
the reasonable period of time for implementation as 15 months from the date of
adoption of a panel or Appellate Body report.69
Thus, full compliance is indeed the
preferred remedy for a GATT 94 violation.
But, DSU Article 22(2) states that if a member
falls to bring the measure found to be Inconsistent with a covered agreement into compliance
... within the reasonable period of time ... such member shall, if so requested, and no later
than the expiry ofthe reasonable period oftime, enter into negotiations with any party having
Invoked the dispute settlement procedures, with a view to developing mutually acceptable
compensation.
Thus, the DSU clearly states that compensation is a possible, even if not a preferred,
remedy for a GATT 94 violation. The fact that the members may agree on
compensation during the reasonable period of time given for the implementation of a
panel decision suggests that compensation may be used as the final means of
remedying a GATT 94 violation.
This idea is further supported by the DSU language pertaining to retaliation. DSU
Article 22(2) also states that if an agreement on compensation is not reached, the
invoking member may request authorization from the DSB to suspend GATT 94
concessions granted to the violating member. DSU Article 22(8) states, however, that
Ibid, at 62.
IWd, at61.
DSU Article 2 l(3)(c) states that the reasonable period of time may be shortened or lengthened In binding
arbitration depending on the particular circumstances of the case. The reasonable period of time for
Implementation has been arbitrated In two cases: Japan and Taxes on Alcoholic Beverages and the European
Communities and the Banana Regime. In the Alcoholic Beverages case, WT/DS8/15 (14 February 1997). the
arbitrator rejected the Japanese argument that political difficulties In passing tax reform legislation and
the adverse effects on consumers and producers demanded an Implementation period longer than 15
months. In the Banana Regime case. WT/DS27/15 (7 January 1998). the arbitrator did give credence to
the Community arguments that the complexity ofthe banana regime legislation and the need to consult
the ACP countries mandated an extension of the reasonable period oftimebut only gave the Community
one week beyond the 15-month period to Implement the panel decision. Neither of these arbitration
rulings touched on the Issue of compensation as an alternative remedy.
The European Court ofJustice and Direct Effectfor the GATT 645
the suspension of concessions... shall only be applied until such time as the measure found to
be inconsistent with a covered agreement has been removed, or the Member that must
implement recommendations or rulings provides a solution to the nullification or impairment
of benefits.
Under the language ofDSU Article 22(2) cited above, one possible solution would be
compensation. Thus, while retaliation is truly temporary, it may be the case that
compensation can be a permanent remedy to a violation. This may make sense within
the context of the GATT 94 because retaliation would in effect counter one violation
with another violation, while compensation counters a violation with monetary
payments.
Thus, the GATT 94 may actually be a unique international agreement in that it
attempts to utilize the 'realpolitik' ofthe environment in which it operates to its benefit
by providing methods of remedying violations which areflexibleenough to maintain
the functioning of the agreement that is, the GATT 94 may provide a choice of
remedy precisely because allowing such a choice helps preserve momentum for the
trade liberalization process which is the ultimate purpose of the agreement.
Permitting members to preserve domestic legislation that arguably violates GATT 94
provisions, as long as they are willing to pay for them in the form of compensation,
may help convince those members to accept other restrictions upon their foreign trade
policy autonomy.
This possible function of the dispute settlement mechanism can be illustrated using
the recent Beef Hormones decisions. In this case, both a GATT panel and a WTO
Appeals Body have ruled that the Community's non-discriminatory ban on the sale of
hormone-treated beef violated the GATT 94 because it was not supported by
sufficiently strong scientific evidence. However, both of these bodies upheld the right
of the Community to set the level ofconsumer protection it deemed appropriate under
the 1994 Sanitary and Phytosanitary Agreement Thus, the Community was allowed
to keep the ban in place until April 1999 while it attempted to produce stronger
evidence for the scientific validity of its ban.
Rather than forcing the Community to produce evidence for the scientific validity of
its ban, perhaps it would be best to allow the Community merely to pay compensation
to the US producers of hormone-treated beef in order to keep its ban in place. This
would avoid the need for a ruling on debatable scientific evidence and give expression
to the desires of Community consumers. In fact, during the dispute, Community
officials admitted that the beef hormones ban was specifically designed to placate
consumer suspicions about chemical additives in food.70
After the initial panel ruling,
some Community officials also suggested that they would prefer to pay compensation
than to remove the ban.71
Thus, it may be the case that the welfare which consumers
in the Community derive from knowing that their beef has not been treated with
70
Andrews. 'In Victory for U.S.. Europe Ban on Treated Beef Is Ruled Illegal', NY Times. 9 May 1997.
Section A. at 1.
7I
. French Agriculture Minister Vasseur went so far as to state that 'France Is entirely prepared to pay
penalties Ifthat Is what U needed to prevent hormone-treated American meat from gaining entry to our
territory.1
Reuters, 'France May Ban US. Meat Over Hormones', LA Times, 12 May 1997, Part D, at 2.
646 E/IL 9 (1998), 626-657
hormones exceeds the compensation which the Community will have to pay to
maintain the beef ban in the face of the panel ruling.72
By allowing the Community the flexibility to pay compensation, the GAIT 94
would help to ensure its own survival and reconcile domestic political interests with
the requirements of the world trade regime. For example, one should consider what
would happen ifit were not the US, but instead a single small country that was the sole
producer of hormone-treated beef. In this case, it seems entirely disproportionate and
unfair to require the Community to abolish a complex domestic policy regime when it
could easily pay compensation instead. Direct effect becomes a very troublesome
proposition in such a situation because the ECJ could not realistically do anything
other than order specific performance for a GAIT 94 violation. Thus, direct effect
would deny the Community the ability to choose the compensation remedy under the
GATT 94.73
Finally, one should consider the possible outcomes for the Community if the ECJ
actually did begin to rule on the GATT 94 compatibility of Community measures. If
the ECJ correctly held that a Community measure violated the GATT 94, then the
Community could be deprived of its ability to choose the method of remedying the
violation. Thus, the Community would be worse off than If a panel had decided that
the Community measure violated the GATT 94. If the ECJ incorrectly held that a
Community measure violated the GATT 94, then the ECJ would also be worse off
because it would be forced to repeal a measure which did not really violate the
agreement.
If the ECJ held that a Community measure was compatible with the GATT 94, then
the Community would remain in the same position as when direct effect did not exist
because either the ECJ's decision would be correct in which case there would never be
an adverse panel decision, or the ECJ's decision would be incorrect, in which case it
could be overturned by a future panel. Therefore, it is not clear that the Community
would ever be placed in a better position within the GATT 94 framework by granting
direct effect to the agreement.74
Given all of the foregoing, it may be the case that the
Court has only denied direct effect to the GATT out of a desire to preserve the best
Estimates state that the US beef Industry lost $250 million dollars of sales In 1996 because of the
Community measure. Clark, 'U.S. Halls EU Beef Imports Ruling by WTO'. Financial Times, 10 May 1997.
at 2.
See Roessler. The Constitutional Function of the Multilateral Trade Order'. In Hllf and Petersmann
(1993). supra note 11. for an argument that the GATT is not really Intended to resolve conflicts of interest
between nations but rather within nations. Roessler concludes, however, that this calls for the
constttuttonalhation of the GATT In domestic legal settings.
These considerations point to another possible motive for the ECJ's decision not to grant direct effect. If
one looks at the ECJ as a domestic court within the GATT 94 context the Court may not want to grant
direct effect because It does not want to make Incorrect decisions. That Is. the ECJ may not be very
different from national courts in the Community context or lower courts In a domestic context. In that It
does not want to operate outside its area of expertise and risk having its decisions reversed.
The European Court of Justice and Direct Effect for the GATT 6 4 7
possible legal position for the Community within the GAIT system rather than to
satisfy a mercantilist agenda for the benefit of specific traders.75
B Implications of the ECJ's Method of Analysis
Having recognized the possibility of non-mercantilist motives for denying direct effect
to the GATT 47 and GATT 94, it is necessary to examine the ECJ's method of analysis
to determine if it proves that the Court's direct effect doctrine is motivated by a
protectionist agenda. Kuilwijk attempts to do this by noting that the ECJ's method of
analysis has been inconsistent. He argues that while the Court has considered the
direct effect of individual provisions ofthe other international agreements, 'in the case
of GATT the Court has unequivocally denied direct effect of the whole agreement as
such'.76
He claims that the Court has done this in order to cloak its true protectionist
motives for denying direct effect77
Thus, he concludes that the Court's inconsistency
is proof of its mercantilist agenda.
This criticism misconstrues the ECJ's method of analysis. The Court has considered
the direct effect of all international agreements, including the GATT 47, by means ofa
two-level inquiry. First, the Court determines if the nature and purpose of the
agreement precludes all provisions from having direct effect. If that threshold inquiry
is met, then the Court examines a particular provision ofthe agreement to determine if
it has direct effect. This second-level inquiry necessarily involves a second consideration
of the nature and purpose of the agreement as the Court must establish the role
of the particular provision within the framework of the agreement.
A close examination of the ECJ case law clearly shows consistency in the Court's
method of analysis. In Recital 21 of Kupferberg, the Court stated that 'neither the
nature nor the structure of the Agreement concluded with Portugal may prevent a
trader from relying on the provisions of the said Agreement before a court in the
Community'.78
Thus, direct effect is possible. In Recital 22, the ECJ stated that the
'question whether such a stipulation is unconditional and sufficiently precise to have
direct effect must be considered in the context of the Agreement ofwhich it forms part
... [and] it is necessary to analyse the provision in the light of both the object and
purpose of the Agreement and of its context'.79
Thus, the Court must examine the
specific provision in the context of treaty to determine if it has direct effect. In Recital
26, the Court held that the particular provision did have direct effect. Therefore, the
references to the nature and purpose of the agreement not only establish that the
The fact that Kulhvljk fails to acknowledge this argument Is telling as much of bis critique is based on the
underlying assumption that the GATT 94 is designed to protect the individual rights of consumers. The
effect of making such an assumption on the direct effect question wiD be explored further In Section 4A.
Kuilwijk. supra note 2. at 105.
BM, at 158.
Kupferberg, supra note 3. at 3664.
Kupferberg, supra note 3. at 3665.
648 EJJL 9 (1998), 626-657
agreement passes the first-level inquiry but also relate to the second-level inquiry as
welL
International Fruit Company shows the ECJ using this same method of analysis with
respect to the GATT 47. In its decision, the Court stated that in order to determine
'whether the provisions of the General Agreement confer rights on citizens of the
Community on which they can rely before the courts in contesting the validity of a
Community measure ... the spirit, the general scheme and the terms of the General
Agreement must be considered'.80
Thus, the Court must engage in the first-level
inquiry. The ECJ never got to the second-level inquiry, however, because 'those factors
[spirit, general scheme and terms of the GATT 47] are sufficient to show that, when
examined in such a context, Article 11 of the General Agreement is not capable of
conferring on citizens of the Community rights which they can invoke before the
courts'.81
Therefore, it held that Article 11, and all other provisions of the GATT 47,
could not have direct effect. Thus, the ECJ's different direct effect doctrines may be due
to differences between the other international agreements and the GATT 47 rather
than to a deliberate policy of trade protectionism.82
The Court's treatment of direct effect for the GATT 47 may have been not only
consistent with its prior jurisprudence but also logical, given the structure of the
GATT 47. The GATT 47 did not contain any references to the possibility ofdirect effect
for its provisions and, as a result the US and Japan denied direct effect This left the
Court with a policy choice. It could hold, as it had with the other international
agreements, that because the GATT 47 did not preclude direct effect and because it
was desirable to enhance enforcement of the agreement in the Community, the GATT
47 was directly effective. Alternatively, the Court could hold that direct effect was not
appropriate for the GATT 47.
It is true that the ECJ deferred to the wishes of the Community institutions in
making this policy choice. For example, in FEDIOL m. Advocate General van Gerven
noted that while Articles 1 and 2 of the New Commercial Policy Instrument 'confine
themselves to making a general reference to international law and to generally
accepted rules ... the background to Regulation 2641/84 leaves no doubt that the
reference to "international law" is a reference to GATT'.83
Likewise, in Nakajima, the
ECJ stated that the New Basic Anti-Dumping Regulation 'was adopted in accordance
with existing international obligations, in particular those arising from Article VI of
the General Agreement and from the Anti-Dumping Code'.84
Thus, the Court did allow
International Fruit Company, supra note 3. at 1227.
Ibid.
Kullwijk actually Implicitly recognizes the possibility ofconsistency In the Court's direct effect analysis by
virtue of the fact that he examines the same four criteria in analysing both the other International
agreements and the GATT 94. Kullwljk. supra note 2, at 106. However, the necessity to prove his thesis
that the Court Is deliberately protectionist prevents him from fairly acknowledging the differences in the
two types of agreements and noting the consistency of the Court's analysis.
FEDJOL ZII. supra note 3. at 1805.
Nakajlma, supra note 9. at 2178.
The European Court of Justice and Direct Effect for the GATT 649
the Community institutions to determine when they wished to be explicitly bound to
the GATT 47 obligations in the Community legal setting.
This position may have been entirely appropriate, however, given the structure of
the GATT 47. If the GATT 47 contracting parties had truly wished to have private
rights of action for the agreement then they would have negotiated provisions
guaranteeing such rights or at least would have provided greater specificity with
respect to the domestic implementation of the agreement's provisions.8
' As Hudec has
noted, while courts can certainly aid in the furtherance of an existing trade policy,
governments must actually fully commit themselves to a policy before courts can
perform the substantive function of judicial review.86
Because the GATT 47
contracting parties had not committed themselves to direct effect, the ECJ's hands
were not tied. In this situation, it may have been entirely appropriate for the Court not
to require direct effect but instead to leave the policy choice to the Community
institutions.87
This situation has not changed with respect to the GATT 94. The GATT 94 still does
not require direct effect and the US and Japan continue to deny the agreement direct
effect. The Community institutions also continue to make their opposition to direct
effect clear. The Commission proposal concerning the implementation of the Uruguay
Round agreements stated that
It Is Important for the WTO Agreement and Its annexes not to have direct effect [because]
without an express stipulation of such exclusion In the Community Instrument of adoption, a
major Imbalance would arise in the actual management of the obligations of the Community
and other countries."
The Council endorsed this proposal in its decision approving the conclusion of the
WTO agreement and its annexes.89
Thus, the Community institutions continue to
express a desire for the GATT 94 not to have direct effect except in the specific
instances in which they declare that they do desire direct effect
The Court's policy position does not necessarily lead to a situation in which the
Community institutions have completely unfettered discretion in the exercise of
foreign trade policy.90
The ECJ can still hold that direct effect does exist after a panel
" See. e.g., Hllfand Petersmann (1993), supra note 11. and Hauser, 'Foreign Trade Policy and the Function
of Rules for Trade Policy Making', in Dicke and Petersmann, supra note 36, for discussions of this
possibility.
84
Hudec, The Role ofJudicial Review in Preserving Liberal Foreign Trade Policies'. In Hllf and Petersmann
(1993), supra note 11.
" That is. the direct effect decision may fall into-an area of law and politics where it Is perhaps appropriate
for the ECJ to carefully reflect on the political acceptability of Its decision. See. e.g.. Everllng. "The Law In
the External Economic Relations ofthe European Community'. In Hilf. Jacobs and Petersmann. supra note
11. at 98, and Lee and Kennedy, supra note 11, at 88. for general discussions.
11
COM(94) 414 final.
" Council Decision 94(800) EC. OJ 1994 L336 (22 December 1994).
90
See Petersmann. Constitutional Functions, supra note 63, for the argument that the Court's position does
lead to a situation in which the rule of law has been abandoned. He claims that It Is a mistake to reject
stricter constitutional and judicial controls on foreign trade policy powers on the grounds that a
constitutional democracy needs unfettered powers to survive In a world of International power politics.
He concludes that the lack ofan effective foreign policy constitution can even undermine the effectiveness
of the domestic policy constitution.
650 Ejm 9 (1998), 626-657
and WTO Appeals Body ruling. This would allow individuals and Member States to
enforce the Community's GATT 94 obligations as determined by the WTO, while
preventing them from denying the Community of its rights under the agreement
Thus, the Court would be responsible for ensuring that the Community does either
alter its law, or provide compensation if permissible, after the WTO ruled that the
Community was in violation of GATT 94 law.
In fact, granting direct effect to the GATT 94 could lead to a serious problem of
interpretive coherence under the agreement. If individuals were allowed to litigate
GATT 94 provisions in national courts this could lead to interpretations of the
agreement which are valid only within a certain member's national courts. For
example, in the absence of a definitive panel ruling addressing a certain issue, two
different members' national courts could reach contradicting rulings on identical
issues. This is especially likely given the fact that the DSU does not contain provisions
allowing national courts to mak& references to a WTO Appeals Body to answer
interpretive questions.'1
Thus, it may be best for the consistency of GATT 94 law to
leave all interpretations of the agreement to panel and Appeals Body rulings.
Such a system would not necessarily weaken compliance with the agreement.
Germany v. Council, which led to the reconsideration of the Court's denial of direct
effect in the first place, demonstrates the Community's desire to comply with its
obligations once they are firmly established by panel and Appeals Body rulings. On 5
September 1997, the Community lost its appeal of the panel ruling which held that
the Community's banana regime violated the GATT 47. At this time, Community
officials stated that the Community would abide by the Appeals Body ruling.92
However, the Community was not able to develop a new banana regime which was
GATT 94 compatible by the 16 October 1997 deadline for submitting a proposal to
implement the WTO ruling. At this time, the Community cited the complexity of the
banana regime for its failure to Implement the ruling and stated that it might decide to
pay compensation instead." hi either case, though, the Community clearly expressed
a desire to remedy its violation.
On 14 January 1998, the Community did propose a new banana regime, which it
claimed satisfied all of its GATT 94 obligations. On 13 February 1998, however, the
six original plaintiffs in the panel ruling issued a declaration stating that this proposed
regime still violated the GATT 94.94
Given the complexity of implementing panel
rulings by means of specific performance, as illustrated by the banana dispute, it may
well be wise for the Court to deny direct effect for the GATT 94 until panel and Appeals
" This problem may become even more troublesome when one considers the fact that Individuals do not
have Intervenor rights In GATT 94 disputes. The lack of a right for an Individual affected by a trade
dispute to be heard by a panel deciding the dispute could be used to bolster the argument that a national
court should not be bound by a panel report This could then lead to further Interpretive divergence if
individuals litigated GATT 94 provisions in national courts via direct effect Of course, one must admit
that Individuals do have the right to petition thdr member governments and have domestic means by
which they can express their positions in trade disputes.
" 'Bananas: EU Agrees to Abide by WTO Ruling', European Report, 29 September 1997.
" 'EU Requests "Time to Think" over WTO Banana Ruling', European Report, 18 October 1997.
M
"EU Agriculture Ministers Seek Second Opinion on Bananas'. European Report, 18 February 1997.
The European Court ofJustice and Direct Effectfor the GATT 651
Body rulings have been made. This would not necessarily prevent individuals or
Member States from suing the Community to enforce compliance with specific rulings.
It would only appropriately leave the possible choice of remedy up to the Community.
Thus, it Is not clear that the Court's denial of direct effect prior to a WTO ruling
necessarily proves that the Court acts with protectionist motives.
4 Political Economy Analysis
According to the GATT 47 preamble, the members agreed to the formation of
'reciprocal and mutually advantageous arrangements directed to the substantial
reduction of tariffs and other barriers to trade and to the elimination of discriminatory
treatment in international commerce'. The principles of reciprocity and nondiscrimination
are operationallzed by means of the Most Favoured Nation (MFN) and
the National Treatment (NT) requirements of GATT 47 Articles 1 and 3. Substantial
reductions in tariffs and other trade barriers are achieved through the periodic rounds
of negotiation required by Article 28 bis and are recorded in binding schedules under
Article 2. Finally, WTO Agreement Article 2(1) provides a 'common institutional
framework1
for the management of trade and commercial relations among the GATT
members. Thus, the Uruguay Round Agreements created the structure necessary to
facilitate and operate a liberalizing multilateral world trading regime.
Given the GATT 47's purposes, the ECJ's critics have complained that denying
direct effect both prevented the Community from obtaining the full economic welfare
benefits possible under the agreement and denied individuals the protection of the
fundamental rights provided by the agreement.95
Kuilwijk extends this critique by
arguing that the Court 'for the sake of the sacred political balance, has failed to strike
an acceptable balance between protection of individual rights and protection of the
public interest' in the area of foreign trade policy, while it has not done so with respect
to the Community's internal market96
Thus, Kuilwijk concludes that direct effect is
necessary 'not only [as] a matter of sound economics but also, and first and foremost,
[as] a matter of justice'.97
A Differences between GATT and Community Fundamental Rights
One must carefully consider the nature of individual rights In the GATT 94 before
drawing comparisons with the Community internal market. First, whether the GATT
94 protects individual rights depends on the extent to which national implementing
legislation provides individuals with the ability to claim rights under the agreement.98
Since the agreement does not provide for direct effect Itself and most members deny
direct effect, it may be assuming what one wants to prove to claim that direct effect is
" See. e.g.. Hllf and Petersmann (1993). supra note 11, and Hllf and Petcrsmann (1991). supra note 11.
w
As wtth his motive analysis. Kuilwijk argues that this occurs because the Court does not wish to Inhibit
the Community Institutions' ability to conduct a protectionist foreign trade policy. Kuilwijk. supra note 2.
at 257.
" Kuilwijk. supra note 2, at 258.
" See, e.g., Tumllr, supra note 11.
652 E/IL 9 (1998), 626-657
necessary to protect individual rights. It is the very existence of direct effect itself
which determines whether individual rights actually exist within the GAIT 94. Thus,
the proper inquiry is to determine whether the members' mutual promises to respect
the agreement's rules is Intended to create rights for individuals in the manner of the
EC Treaty.
Second, it is not entirely clear that the GATT 94 embodies Individual rights in the
same form as they exist within the Community's internal market For example, the
Community right to non-discrimination requires the equal treatment of all traders
unless discriminatory treatment is objectively justified." Although GATT 47 Article
13 prohibits discriminatory treatment of like products, there are several exceptions to
this general principle. Most noticeable are the anti-dumping and countervailing duty
regimes.100
Article 24 also permits 'an exception to the MFN principle for regional
trade blocs. Because these exceptions effectively allow members to distinguish
between traders, the agreement may sanction instances of discrimination which
would not be permitted within the Community.
The principle of proportionality is used within the Community to analyse the
efficiency and effects of Community regulations when determining their validity
under the EC Treaty.101
While the GATT 47 generally imposes stricter limitations on
the use oftrade policy instruments as they become less economically efficient, it is not
clear that this is based on as rigid a principle of proportionality as is observed in the
Community. For example, although Article 11 outlaws the quota form of protectionism
in favour of the less distorting and more transparent tariff form, there are also
" See, eg.. Case C-177/90, Rolf-Herbert Kuhn v. Landwirtschafiskammer Weser-Ems, [1992] IECR 35. The
Court has stated that discrimination results from either 'treating differently situations which are
identical or treating In the same way situations which are different'. Case 8/82, Hans-Otto Wagner v.
Bundesanstalt fOr Umdwlrtschaftlichc Mnrklordnwg, [1983] ECR 371, at 387.
100
Not only do these regimes permit discretionary treatment, but they also cannot be Justified on efficiency
grounds. For example, anti-dumping duties may be Imposed on imported products merely upon a
showing of price discrimination. If anti-dumping duties were designed strictly on economic efficiency
grounds, they would only be allowed upon a showing that imports were being priced at less than their
average variable cost of production because this would Indicate that foreign exporters were engaged In
predation. The strongest rationale for the current anti-dumping duty regime Is that it serves as a safety
valve to vent protectionist pressures. Trebllcock and Howse, supra note 44, at 115-119. If the members
really want to deal with predation, then they should negotiate competition rules. Michael Leldy,
"Antidumping: Solution or Problem In the 1990's'. In IMF, International Trade Polities: The Uruguay Round
and Beyond (1994).
likewise, a country Imposing countervailing duties Is not required to prove that a foreign
government's subsidies cost more than the benefits they ultimately produce. Instead, duties may be
Imposed merely on a showing that the subsldlied Imports have caused material Injury to similar situated
domestic producers In the Importing country. Thus.countervalllng duties serve as an Inefficient means of
dealing with the distributive effects caused by liberalized trade. Trebllcock and Howse, supra note 44, at
152-153. Instead of countervailing duties, direct transfer programmes such as labour adjustment
assistance payments would be a more efficient means of dealing with these distributive effects.
101
See, e.g., Case 114/76, Bela-MuhleJosefBergmann v. Grows-Farm, [1977] ECR 1211. The Court has used
this principle to determine 'In the first place, whether the means [a Community law] employed to achieve
Its aim correspond to the Importance of the aim and. In the second place, whether they are necessary for
Its achievement'. Case 66/82. Fromancals v. Fonds d'orientation a de regularisaUon des marches agricoles
(FORMA), [1983] ECR 395. at 404.
The European Court of Justice and Direct Effect for the GATT 6 5 3
several exceptions to this requirement102
Article 12 permits quantitative restrictions
in response to a serious balance of payment problem and Article 18 allows LDCs to
impose quantitative restrictions for infant industry reasons. Thus, it appears that the
agreement permits Inefficiencies which would not be allowed within the Community.
Finally, the Community and the GATT 47 principles of undistorted competition
may not be equivalent While the Community is intended to create a single unified
market the GATT 47 merely attempts to create a multilateral trading system based on
the principle of reciprocity.10
' The GATT 94 members do not have to adopt
harmonized trade rules but merely must exchange what they view to be equal trade
benefits. As noted above, there are several instances in which unilateral and
discriminatory trade actions are permitted. And one of the largest problems with the
GATT 94 is the lack of a competition policy regime.104
Thus, the members retain a
measure of national control which allows them to affect competitive conditions to a
greater extent than the Community Member States can.
The foregoing arguments are not meant to deny that the GATT 94 has significant
effects on individuals. They are merely intended to show that the nature of individual
rights in the GATT 94 may differ from that in the EC Treaty. After all, the ECJ has never
held that the GATT 47 'is more than an agreement which creates mutual obligations
between the contracting states'105
as it has for the EC Treaty. Thus, one should be
careful about assuming that the Community concept of direct effect should
automatically be extended to the GATT 94 in order to protect individual rights.
Instead, one should engage in a thorough analysis of the actual nature of the
obligations contained within the GATT 94 to determine the appropriateness of direct
effect.
B The Place of Individuals within the GATT
When considering the nature of the GATT 94 obligations in relation to direct effect, it
is important to make explicit the political economy view of international trade from
which one is operating. Kuilwijk and the Court's other critics generally work from the
classical trade theory view which emphasizes the benefits of unilateral trade
liberalization.106
Under this view, unilaterally granting direct effect may indeed be
102
It should be noted that the Article 11 prohibition against quotas was historically incapable of preventing
the rise of negotiated bilateral agreements such as VERs and OMAs which effectively imposed quota
systems. Although these measures have been disciplined under the Agreement on Safeguards, the
settlement of the US-^apan dispute over cars and car parts under the DSU by means of a bilateral
agreement shows that negotiated solutions which result in quota-like regimes may still occur under the
GATT 94.
101
Kuilwijk claims that the Community Is Intended 'directly or Indirectly, at the creation of a level playing
field'. Kuilwi|k, supra note 2, at 231. As anyone familiar with trade theory will recognize, this Is a very
problematic statement because a truly level playing field in the sense of equal conditions across all
countries, results In a situation where trade is no longer beneficial to any country. Thus, It is not clear
what to make of Kuirwtjk's claim.
1W
See Trebilcock and Howse. supra note 44. at 122-124.
105
Case 26/62. MgtmeneTransport- en ExpedltieOndememlng van Gend& Loos v. Nederlandseadmlnistratieda
belastlngen, [1963] ECR 1. at 12.
10
* See Trebilcock and Howse. supra note 44. at 1-6, for a concise yet thorough explanation of this theory.
654 EJ1L 9 (1998), 626-657
beneficial and necessary because it would produce welfare benefits for individuals in
the Community and help protect the individual rights of both Community consumers
and traders.
The GATT 94, however, is ultimately an agreement regulating the rights and
obligations ofits members, not ofindividuals. The preamble to the GATT 47 states that
the members, as sovereign states, recognize
that their relations In thefieldof trade and economic endeavor should be conducted with a
view to raising standards ofliving, ensuring full employment and a large and steadily growing
volume of real Income and effective demand, developing the full use of the resources of the
world and expanding the production and exchange of goods.
These are goals which can only be achieved and evaluated on a macroeconomic basis
with respect to the members themselves and not on a microeconomic basis with
respect to individuals.
The concluding remarks contained in the second panel decision concerning the
Community banana regime are revealing. The panel stated that it
wishes to .point out, however, that the Contracting Parties have at their disposal other
procedures under the General Agreement... that are designed to allow Contracting Parties to
take Into account. In the view of the Panel economic and social considerations. The adoption
ofthis report would not prevent the Contracting Partiesfromtaking action under any ofthese
Articles.
Thus, the panel made its ruling on the validity of the Community's banana regime
without regard to the economic and social consequences on individuals.
This emphasis on the members as opposed to individuals is inherent in several of the
agreement's founding principles. For example, non-discrimination within the GATT
47 means non-discrimination between members and not non-discrimination between
individual traders in different member countries. In fact, it would be impossible
to ensure that discrimination did not occur between individual traders across
members because natural differences in the factor endowments and technology of
different members automatically produce differences in the competitive positions of
their traders. And without these differences in factor endowments and technology,
international trade would not produce economic welfare gains at all.
The Uruguay Round Agreements confirmed that reciprocal and multilateral
negotiations between sovereign members form the basis for all trade agreements. This
emphasis on reciprocity has been severely criticized for the negative effects it may
have on trade liberalization efforts.107
One should realize, however, that the GATT 47
107
Kullwijk states that reciprocal trade negotiations are a 'costly farce' designed to 'win domestic political
support from important export industries'. He concludes that 'reciprocity never may be viewed as an
exchange of economic gains, because this would suggest that liberalization entails economic costs' and
that reciprocity Is only an 'exchange of domestic political support among governments'. Kullwijk. supra
note 2, at 128-129. Kuilwijk's 'costly farce' characterization depends, however, on the assumption that
the GATT members would actually unilaterally liberalize their trade policy. As the history of
international trade suggests, this is not necessarily guaranteed. Because countries have historically acted
in a self-interested and protectionist manner, the reciprocity requirement provides built-in pressures for
multilateral trade liberalization. Thus, reciprocity may serve a useful economic purpose by helping to
achieve further trade liberalization.
The European Court of Justice and Direct Effectfor the GATT 6 5 5
does attempt to mitigate these effects with the MFN principle: that is, the combination
ofreciprocity and MFN provides an incentive for the exporters in any given member to
lobby their government for further tariff reductions on imports. The exporters have an
incentive to do this because they know that any reductions in their domestic import
tariffs will be matched on a reciprocal basis by not just some, but all the other
members. Thus, exporters are harnessed as a trade liberalizing lobbying force to
counteract the protectionist lobbying of domestic import-competing firms.108
The reciprocal method of negotiating agreements is used, even though nonreciprocal
trade liberalization is in the self-interest ofeach member, because individual
members may not be willing to liberalize trade on a non-reciprocal basis.109
That is to
say, it may be necessary for each member to provide some tangible benefit to the
others in order for each member to actually accept the entire package of trade
liberalization measures. The reciprocity principle helps ensure that these beneficial
trades actually do occur by requiring each member to grant a trade concession in
order to attain a concession. Given this design, direct effect might actually inhibit
further worldwide trade liberalization by allowing an individual to require a member
to unilaterally liberalize its trade measures without forcing other members to do the
same: that is, direct effect may result in concessions which do not have to be matched
by the other members.
In fact, granting direct effect may even allow individuals within a member to
directly harm others. The welfare gains which result from free trade occur on an
aggregate level for a member as a whole. However, distributive effects occur on an
individual level due to labour dislocations resulting from free trade.110
While it is
"* Cf. Hauser, supra note 85, at 3-4, for an argument that Importers are Inherently better organized than
exporters as a lobbying force because a government cannot secure the benefits which exporters desire,
access to foreign markets, without the help of other foreign governments.
"" In general, unilateral trade liberalization will produce welfare gains for the liberalizing country.
Although there are two cases where unilateral trade liberalization Is economically Inefficient, the optimal
tariff and infant industry arguments, these cases require the assumption of a large country with
significant market power and are therefore probably unrealistic. Thus, they should not be used to claim
that national Interests should prevail over Individual rights within trade policy. There Is a case where
limited protectionism can lead to worldwide, not just national, welfare gains. This may occur when
external scale production economies are present In a national economy which would allow a national
producer to become the most efficient worldwide producer but which cannot be realized currently due to
the fact that the producer's Initial production level Is inefficient compared to the current world
production. In this case, limited protection may allow the national producer to achieve the economies of
scale necessary to compete In the world market Once the producer reaches the competitive scale. It will
supply the world market at a price which Is lower than the current world price. Thus, the world achieves
a welfare gain. This argument depends, however, on two limiting assumptions. First, the protected
domestic market must be large enough for the producer to achieve the scale needed to compete on the
world market Otherwise, once protectionism Is lifted, the producer will not be able to expand to achieve
the scale of the most efficient worldwide producer. Second, protectionism must be limited in time and
extent to the minimum necessary to allow the producer to achieve the required scale of production.
Deciding when and how to end protectionism Is much easier In theory than practice. Thus, once again
this should not be used to argue that national interests are superior to individual rights.
110
That Is, workers trapped In Import competing Industries will be hurt by free trade because they cannot
shift to the sectors which enjoy additional export opportunities. These trapped workers will suffer a
decrease In their real wage and a decline In their standard of living.
656 EJIL 9 (1998), 626-657
theoretically possible to offset these effects with a portion ofthe overall welfare benefits
resulting from a shift to free trade, this Is a political decision. Since there is no
guarantee that this decision will actually be made, it is possible that direct effect would
allow individuals to use courts to force their government to adopt trade policies which
harm other individuals without a consideration ofthe domestic distributive effects.111
Given these considerations,, it is clear that one should be cautious about
automatically assuming that the GAIT 94 should be viewed as a source of individual
rights. Much of the language and structure of the agreement is designed with the
intent of regulating trade relations between sovereign members which retain the
freedom to control their domestic economies. Therefore, it may be appropriate for the
Court to refuse to grant legal effect to the GATT 94 until the Community institutions
have expressly declared a desire for such effect or until the Community's obligations
have been determined by panel and Appeals Body rulings. In the end, the Court has
left the question of enforceability of the GATT 94 to the institutions capable of dealing
with their distributive effects and reconciling individual rights with the public interest.
It is far from obvious that this constitutes a rejection of sound economics or principles
of justice.
5 Conclusion
The question ofwhether the GATT 94 should have direct effect is complex. The GATT
94 is a constitutional agreement like the EC Treaty. However, it adopts a different set
of economic goals than the EC Treaty does. The GATT 94 is also an international
agreement of the Community. However, it establishes a regulatory structure which is
different from that established by the other international agreements into which the
Community has entered.
It is also far from obvious that the GATT 94 is intended to protect individual rights.
The agreement is specifically designed to manage trade relations between sovereign
entities. Thus, perhaps the concept of direct effect which exists for the EC Treaty and
the other international agreements should not apply to the GATT 94. After all, one
may still respect the rule of GATT 94 law and Individual rights domestically without
granting direct effect to the agreement.
On the other hand, it must be acknowledged that the agreement does contain clear
and precise obligations to which all the members have agreed. The GATT 94 has
introduced a dispute resolution system which mandates the resolution of conflicts
between members in a rule-based manner. And, in the end, it is true that the effects of
the world trade rules are felt by individuals. However, the fact that Individuals are
ultimately affected by seemingly clear rules should not be enough, by itself, to
111
See Trebllcock and Howse. supra note 44, at 178-183, for a discussion of possible labour adjustment
policies and wealth transfer schemes designed to compensate those who lose under fixe trade. Ofcourse,
one could always grant direct effect and then use redistributlve measures to mitigate any undesirable
effects. Even In this case, however, it Is still the case that direct effect Introduces a level of complexity into
dispute settlement that may not be desirable given the Intended operation and current structure of the
world trading system.
The European Court of Justice and Direct Effect for the GATT 6 5 7
conclude that there should be individual rights of enforcement As argued, there are
numerous doctrinal, pragmatic and philosophical reasons to refuse direct effect until
the members explicitly agree on such a right. Until that time, it certainly is possible,
and perhaps logical, to grant a limited form of direct effect which would allow for
individual enforcement of any final panel or Appeals Body rulings. But in the end,
while the principle of direct effect has made significant contributions to the
development of the Community legal system, it may not be appropriate for the GATT
94 and the world trading system.
Comments on this article are invited on the EJIL's web site: <www.ejil.org>.