Reading Material

NATIONAL LAW UNIVERSITY ODISHA,
CUTTACK
Reading Material
On
Law on Dispute Resolution

in International Trade
B.A. /B.B.A. LL.B

X SEMESTER
2017-2018
COURSE TEACHER
Mr. Kapil Sharma
Table of Contents
1. The WTO and Rules-Based Dispute Settlement: Historical Evolution, Operational

Success, and Future Challenges By: William J. Davey................................................1
2. Applicable Law in the Dispute Settlement Body of the WTO By: Prabhash
Ranjan...........................................................................................................................23
3. WTO Dispute Settlement: Emerging Practice and Procedure By: Debra P. Steger and
Peter Van den Bossche................................................................................................28
4. Why Are WTO Panels Taking Longer? And What Can Be Done About It? By:
Matthew Kennedy.....................................................................................................36
5. The Question of Remand Authority for the Appellate Body By: Fernando
Piérola........................................................................................................................69
6. Third Party Rights and the Concept of Legal Interest in World Trade Organizational
Dispute Settlement: Extended Participatory Rights to Enforcement Rights By:
Ngangjoh H. Yenkong.............................................................................................91
7. The Time to Comply with an Adverse WTO Ruling Promptness within Reason By:
Pierre Monnier........................................................................................................107
8. Seeking Compliance with WTO Rulings: Theory, Practice and Alternatives Brendan
By: P. McGivern........................................................................................................125
9. Compliance Proceedings under Article 21.5 of DSU and Doha Proposed Reform By:
Tsai-Yu Lin................................................................................................................142
10. Peculiarities of retaliation in WTO dispute settlement By: Kym Anderson..............164
11. Compliance Problems in WTO Dispute Settlement By: William J.
Davey.........................................................................................................................176
12. The Challenge of Reforming the WTO Dispute Settlement Understanding By: Heinz
Hauser and Thomas A. Zimmermann........................................................................186
13. The Future of WTO Dispute Settlement By: Andrew W. Shoy................................191
Journal of International Economic Law, 2014, 17, 679–700
doi: 10.1093/jiel/jgu031
Advance Access Publication Date: 4 September 2014
The WTO and Rules-Based Dispute

Settlement: Historical Evolution,
Operational Success, and Future
Downloaded from http://jiel.oxfordjournals.org/ at University of California, San Diego on September 15, 2014
Challenges
William J. Davey
Guy Raymond Jones Chair in Law Emeritus, University of Illinois College of Law.
E-mail: wdavey@illinois.edu
ABSTRACT
The World Trade Organization (WTO) has three main functions—first, negotiations
on trade matters; second, oversight of existing WTO agreements; and third, dispute
settlement. The WTO’s negotiation function has broken down and its oversight func-
tion, while useful and valuable, is the least visible and least significant of the three func-
tions. Thus, the credibility of the WTO as a functioning international organization
essentially depends on ensuring the effectiveness of its dispute settlement function.
This paper briefly traces the historical evolution of the dispute settlement system under
the General Agreement on Tariffs and Trade (GATT) and then considers the current
state of the WTO dispute settlement system—how effectively has it operated to date
and what operational and procedural problems have arisen? It concludes by examining
the challenges the WTO dispute settlement system will face in the coming years, con-
sidering, in particular, whether the system will be able to resolve effectively disputes
between the major WTO powers—the United States, the European Union and China.
Dispute settlement has always been an important function of the multilateral trading
system. It makes no sense to spend years negotiating rules, if there is no effective sys-
tem to clarify the meaning of those rules and to enforce them. But in recent years,
dispute settlement has become even more critically important to the future of the
World Trade Organization (WTO). Under its constitutive agreement, the WTO has
essentially three functions—first, negotiations on trade matters; second, oversight of
existing WTO agreements; and third, dispute settlement. As explained in the intro-
ductory section of this chapter, the WTO’s negotiation function has broken down
and its oversight function, while useful and valuable, is the least visible and least sig-
nificant of the three functions. Thus, in the near-to-medium term, the credibility of
C The Author 2014. Published by Oxford University Press. All rights reserved.
V
679
1
680 The WTO and Rules-Based Dispute Settlement
the WTO as a functioning international organization essentially depends on ensuring

the effectiveness of its dispute settlement function.
The chapter is divided into three sections. The first section briefly traces the his-
torical evolution of dispute settlement under the General Agreement on Tariffs and
Trade (GATT) and John Jackson’s role in its evolution. The second section con-
siders the current state of the WTO dispute settlement system—how effectively has
it operated to date and what operational and procedural problems have arisen? The
last section examines the new challenges the WTO dispute settlement system will
face in the coming years, considering, in particular, whether the system will be able
to resolve effectively disputes between the major powers—the USA, the European
Union, and China. The WTO dispute settlement system must accomplish that at a
minimum if it is to be judged successful in the future. Moreover, as noted above, the
reputation and international standing of the WTO itself may well turn on how the
performance of its dispute settlement function is evaluated. In light of the oper-
ational and procedural problems identified in Section III, the last section concludes
with several reform suggestions.
I. I NT RO D UCTI ON
The WTO Agreement spells out three key functions for the WTO—it is to serve as
a negotiating forum on trade issues; it is to oversee the operation of the existing
WTO agreements; and it is charged with the operation of the WTO dispute settle-
ment system.1 In respect of the negotiation function, it must be recognized that the
Uruguay Round negotiations, which led to the creation of the WTO, were an amaz-
ing success—those negotiations added services and intellectual property to the
GATT remit; brought textiles and agriculture back under GATT disciplines; pro-
vided significant tariff reductions and a large expansion in tariff bindings; updated
the various side agreements; and created a new organization and an effective dispute
settlement system with compulsory jurisdiction and a system of enforcement. Since
the WTO Agreement had to be accepted as a single undertaking, all of these results
applied across the board to all participants in the negotiations. As such, the Uruguay
Round was a singular accomplishment.2 Since its creation in 1995, however, the
WTO’s record in bringing negotiations to a successful finish has not been very good.
The Information Technology Agreement (ITA) was concluded at the Singapore
Ministerial in 1996;3 and, early on in the WTO’s existence, agreements were reached
on two issues left over from the Uruguay Round services negotiations—telecoms
1 WTO Agreement, Article III. Article III actually lists five functions. I consider one of the two other func-
tions—the operation of the Trade Policy Review Mechanism—to be part of the oversight function. The
remaining function—to improve coherence with the International Monetary Fund and the World
Bank—has never been fully operationalized, in part because it is rather vaguely expressed, and I do not
consider it further in this chapter.
2 See generally John Croome, Reshaping the World Trading System: A History of the Uruguay Round, 2nd ed.
(The Hague: Kluwer-WTO, 1999).
3 Ministerial Declaration on Trade in Information Technology Products, First Ministerial Conference,
Singapore, WT/MIN(96)/16 (13 December 1996).
2
The WTO and Rules-Based Dispute Settlement 681
and financial services.4 But it is worth stressing that these three agreements were
concluded on the basis of getting a critical mass of WTO members to accept them.
The key players signed on to the agreements, but many others did not.5 In addition,
a number of major accessions have been completed—most notably China and
Russia.6 However, not much else has been accomplished on the negotiating front.7
The most conspicuous failure of the WTO in the negotiating arena, of course, is
the never-ending trade talks known as the Doha Development Agenda, or more sim-
ply the Doha Round, which was launched in 2001.8 While there may eventually be
some results from the Doha Round, it has been ongoing for a dozen years and the
prospects for significant achievements seem to have diminished over time. Indeed,
at the December 2011 ministerial meeting, the talks were recognized as being at an
‘impasse’.9 At the Ninth Ministerial Conference held in Bali in December 2013 an
agreement was announced on trade facilitation, although as of 1 June 2014 the text
had not been finalized and opened for acceptance.10 A number of minor decisions
were also taken.11 In the first half of 2014, the new WTO Director-General—
Roberto Azevedo—was trying to restart the Doha negotiations, but it is not clear
that much will come of that effort.12 Thus, it was fair to say that as of mid-2014 the
WTOs negotiating mandate is basically not functional.
There are three major reasons for the failure of the Doha negotiations to produce
any significant agreements. First, the existence of the WTOs consensus decision-
making requirement and the aim of having a single undertaking, which requires all of
the diverse WTO members to agree to any outcome, make a broad-based agreement
unlikely. Without prospects for a broad-based agreement, there are limited opportu-
nities for cross-issue trade-offs that are essential to reach compromises in respect of
4 Fourth Protocol to the General Agreement on Trade in Services, S/L/20 (30 April 1996) (telecommuni-
cations); Fifth Protocol to the General Agreement on Trade in Services, S/L/45 (3 December 1997)
(financial services).
5 For example, the WTO website indicates that there are about 50 members who are current participants
in the ITA (counting the 28 EU Member States as one). It also indicates that the financial services proto-
col was signed by 56 members and the telecommunications protocol by 69 members (counting the EU-
15 as one).
6 Accession of the People’s Republic of China, WT/L/432 (23 November 2001); Accession of the Russian
Federation, WT/L/839 (17 December 2011). The WTO website listed 32 completed accessions as of
June 2014, with 24 candidates at various stages in the process.
7 The Uruguay Round agreements contained a number of provisions calling for completion of negotiations
that had not been finalized during the round. For example, the Agreement on Rules of Origin provided
for the negotiation of nonpreferential origin rules by mid-1997. Those negotiations are still ongoing.
Similarly, there were issues that were to be negotiated in the GATS context, such as rules on safeguards
and subsidies. Those negotiations have also not succeeded to date. However, it is worth noting that the
parties to the Government Procurement Agreement—a WTO Annex IV plurilateral agreement to which
most WTO members do not adhere—did negotiate a new version of that agreement in 2012, and it
came into force as of 6 April 2014.
8 Ministerial Declaration adopted 14 November 2001, Fourth Ministerial Conference, Doha, WT/
MIN(01)/DEC/1 (20 November 2001).
9 Chairman’s Concluding Statement, Eighth Ministerial Conference, Geneva, WT/MIN(11)/11 (17
December 2011). There was no ministerial declaration adopted at Geneva.
10 WT/L/911 (11 December 2013).
11 Bali Ministerial Declaration, WT/MIN(13)/DEC (11 December 2013).
12 See WTO website for the Director-General’s speeches on this subject in the January–May period.
Available at http://www.wto.org/english/news_e/spra_e/spra_e.htm (visited 11 August 2014).
3
the more controversial negotiating subjects. While there are advantages to requiring
consensus (for example, it accords legitimacy to negotiated outcomes), its achieve-
ment requires strong leadership. In the Uruguay Round, leadership was supplied by
the so-called Quad countries (Canada, the European Union, Japan, and the USA).13
Such leadership is lacking today,14 and it is hard to see where strong leadership will
come from, if not from the USA and the European Union. Today, none of the major
developing countries seems to be pushing seriously for an agreement in the Doha
Round.15
Second, in a mercantilist world, and I do believe that much of the world operates
with that mindset, many developing countries have a relatively good deal at the mo-
ment in terms of market access to their principal markets and are loathe to commit
to significant concessions or market openings of their own, especially openings of
the sort that the USA and the European Union would like to see. In particular, the
fear of China as a competitor seems to drive some of this unwillingness to make mar-
ket access concessions, since China would benefit from any such concessions under
the most-favored-nation (MFN) clause.16
Finally, the growth in the use of free trade agreements (FTAs) on a bilateral or re-
gional basis has undermined political support for multilateral negotiations in that
those domestic constituencies who want market openings and improved disciplines
in trade matters have an alternative route, one which is more flexible than the WTO
route.17 More particularly, it appears that some industries now view FTAs as the
only way to achieve their negotiating goals in the trade arena. The prime example is
what might be called the intellectual property lobby—that is, the groups that want to
promote intellectual property rights and enforcement. Those groups were prime
Uruguay Round supporters because it brought minimum standards for intellectual
property into the GATT/WTO system and made those obligations subject to WTO
dispute settlement. But today, I think that the TRIPS lobby sees FTAs as the only
way to achieve its current goals, particularly with respect to concerns in regards to
pharmaceutical products. For such groups, what is desired is simply not attainable in
the WTO, but is in FTAs.18
So if a big package deal in the Doha negotiations is out of the question, what does
the near-to-medium term future hold for the WTO in respect of negotiations?
Basically, I anticipate that the WTO will be a bystander to a lot more FTA
13 See generally Croome, above n 2.

14 Although the Obama Administration expresses general support for achieving new trade agreements, it has
generally been viewed as not pushing hard for such agreements, especially not at the multilateral level.
Dan Ikensen, Obama’s Trade Policy Should Be Judged on Its Accomplishments, Not Its Promise, Forbes,
12 March 2013. Available at http://www.forbes.com/sites/danikenson/2013/03/12/obamas-trade-pol
icy-should-be-judged-on-its-accomplishments-not-its-promise/.
15 ‘Business Representatives Explore Ways To Liberalize Trade as Doha Falters’, BNA International Trade
Daily, 29 July 2011.
16 ‘Dead Man Talking’, The Economist, 28 April 2011 (quoting IMF economist Aditya Mattoo).
17 Professor Bhagwati has noted: ‘[W]hen powerful business and labor interests can extract concessions in
those bilateral deals, they have no reason to support a multilateral trade agenda.’ Jagdish N. Bhagwati,
‘The Wrong Way to Free Trade’, New York Times, 24 July 2011.
18 Recent US FTAs usually have what are known as TRIPS-plus provisions that require stricter protection
of certain IP rights than does the WTO TRIPS Agreement. Such provisions would be opposed in the
WTO by major developing countries.
4
negotiating activity—probably around the Pacific and Asia in particular, including de-
veloping country preference agreements under the Enabling Clause. To the extent
that progress in trade liberalization may occur in the near term, the Trans-Pacific
Partnership talks and the FTA negotiations between the USA and the European
Union seem more likely to produce meaningful results than do the stalled Doha
talks.
The second WTO function is the oversight and administration of the various
agreements. This is the mundane business of the WTO secretariat and national gov-
ernment trade specialists, but it has three particular useful aspects. First, the various
WTO agreements require all sorts of notifications to be made in respect of trade
laws of all kinds (from dumping laws to patent laws), regulations affecting trade
(there are hundreds of notifications of technical regulations and sanitary and phyto-
sanitary (SPS) measures every month) and various notifications of actions taken
under trade laws.19 It is common for some of the more problematic notifications to
be discussed in the periodic meetings of the committees in charge of overseeing the
various agreements. In some cases, there is a sort of limited dispute settlement pro-
cess that may settle questions that are raised. The SPS Committee has been particu-
larly active in this regard.20 While I would not want to overstate the significance of
the results of the committee reviews and discussions of questions raised by members
about notifications, they do serve a very useful purpose. They promote transparency
in trade regulation and ensure that WTO members do not take protectionist meas-
ures secretly. Indeed, this oversight function has probably helped forestall the extent
of protectionist reaction that might have otherwise have been expected to occur as a
result of the 2008 financial crisis and its aftermath.21 The WTOs Trade Policy
Review function—which I lump together with the oversight function—has a similar
effect.22
Second, some of the committees are charged with establishing best practices or
adopting interpretations and some activity of this sort has occurred in the
Antidumping Committee and the SPS Committee for example.23 This process has
been less active in the recent past because of the pendency of the Doha negotiations
19 See generally ‘GATT Article X: Transparency and Proper Administration’ in William J. Davey, ‘Enforcing
World Trade Rules: Essays on WTO Dispute Settlement and GATT Obligations 299-313’ (Cameron
May 2006).
20 See Joanne Scott, The WTO Agreement on Sanitary and Phytosanitary Measures: A Commentary (Oxford:
Oxford University Press, 2007).
21 In the WTO’s 17 June 2013, Report on G-20 Trade Measures, it is noted (at page 2) that ‘the WTO
Secretariat has calculated an estimate of the trade impact of import-restrictive measures taken since
October 2008 and which are still in place. The cumulative trade impact of import restrictions imple-
mented by G-20 economies on G-20 trade is estimated to have been around 0.2%. Although the actual
trade impact of all restrictions may be higher, in particular in those specific sectors most heavily affected
by measures, this low aggregate percentage provides an illustration that, on the whole, most countries
have, so far, resisted resorting to widespread protectionism’.
22 WTO Agreement, Annex 3. The trade policies of the major trading countries are reviewed every two to
four years by the Secretariat and the WTO membership discusses them in country specific review
sessions.
23 See, e.g. Committee on Anti-Dumping Practices—Recommendation Concerning the Timing of the
Notification under Article 5.5—Adopted by the Committee on 29 October 1998, G/ADP/5
(3 November 1998).
5
and may not be very active in the future for fear that any such decisions will be used
by the Appellate Body to impose obligations.24 Nonetheless, it provides a mechanism
to make progress even when negotiators otherwise are unable to.
However, a third aspect—what I might call socialization—may indirectly have the
same effect as the second aspect and function at all times. Even if the committees
cannot adopt formal decisions on best practices or interpretations, the informal dis-
cussions that take place between officials with similar administrative interests can be
expected over time to have a standardizing effect on what those officials do. By this I
mean, if someone discovers that they are out of step with everyone else on a particu-
lar approach to a problem, they may well over time change in the direction of the
standard approach to the problem. I think that Secretariat officials thought that the
short-lived trade and competition meetings at the WTO had such an effect and
are hopeful that this can be replicated in other areas. Indeed, there is some reason to
believe it is occurring already.
Thus, the second function—the mundane administration and oversight of agree-
ments and the world trading system—will continue and be useful as it has been in
the past. Unlike negotiations, this function cannot easily migrate elsewhere. Most
FTAs do not establish secretariats with anything more than secretarial functions, so I
see little chance that any competing organization will contest what the WTO does in
this area.25 Thus, simply overseeing and administering what has already been created
ensures an important function for the WTO.
The third principle function of the WTO is dispute settlement. It too is unlikely
to be threatened by the creation of FTAs. While such agreements typically have dis-
pute settlement regimes, often based on the WTO dispute settlement system, those
FTA dispute settlement systems tend not to be used much and the experience under
them is arguably inferior to that of the WTOs system.26 I will not dwell on this, but I
think that several differences give a large advantage to the WTO. The FTA systems
tend to have binational panels without possibility of appellate review and with no
Secretariat support, thus there are more frequent concerns with bias and quality of
decisions. While the WTO system is multilateral, FTAs are often composed of two
or only a few parties. This means that power plays a much more important role in
the implementation of results. One can think of the North America Free Trade
Agreement (NAFTA) and Mexico’s experience in the trucking case to see evidence
24 Article 2.12 of the Agreement on Technical Barriers to Trade requires that there be a reasonable interval
between the publication of a technical regulation and its entry into force. In US – Clove Cigarettes, the
Appellate Body found that a WTO Ministerial Decision that, amongst other things, defined the term ‘rea-
sonable interval’ in Article 2.12 as being at least six months could be considered a subsequent agreement
among the parties for purposes of interpreting the TBT Agreement. As a consequence, in Clove
Cigarettes, the USA was found to have violated Article 2.12 because the interval between its publication of
a ban on clove cigarettes and the ban’s entry into force was only three months. Appellate Body Report,
United States – Measures Affecting the Production and Sale of Clove Cigarettes, WT/DS406/AB/R, paras
237–97 (4 April 2012), adopted by the DSB on 24 April 2012.
25 There is some overlap in this regard in respect of work done in other international organization, such as
the OECD and UNCTAD. However, their activities are far more limited than those of the WTO in this
area.
26 On this subject generally, see William J. Davey, ‘Dispute Settlement in the WTO and RTAs: A
Comment’, in Lorand Bartels and Frederico Ortino (eds), Regional Trade Agreements and the WTO Legal
System (Oxford: Oxford University Press, 2006), at 343–57.
6
of this.27 There is peer pressure to implement WTO Dispute Settlement Body

(DSB) decisions in the WTO, there is nothing comparable in an FTA, particularly
where parties have unequal bargaining power. More to the point, I think that it is un-
likely that any prospective FTA is going to perform what in the end is the basic pur-
pose of the WTO dispute settlement system—to serve as a means of resolving
disputes with and between the principal trading powers, and, in particular, the USA,
the European Union, and China.
So, the remainder of this article will focus on the WTO dispute settlement system
in general and in terms of how effectively it has been able to resolve disputes be-
tween the major trading powers. Particular attention is given to what the future holds
as China becomes a more frequent participant in dispute settlement cases. Given
that the WTOs negotiation function will likely not produce significant results in the
near-to-medium term and that the administration/oversight function is routine, it
seems likely that the WTOs dispute settlement system will have a particularly im-
portant place in the WTO in the near-to-medium term. Its ability to continue to be
viewed relatively favourably as international dispute settlement systems go will be
critical to the maintenance of the overall reputation and credibility of the WTO itself.
If it does not resolve successfully disputes between the USA, the European Union,
and China, it may well be viewed as a failure. After all, if a rules-based institution can-
not effectively negotiate rules and enforce those rules, then its raison d’etre is cer-
tainly called into question and its future would not be very rosy.
II . TH E H I ST OR ICAL EV OL UT IO N O F TH E WT O
D IS P U T E S E T T L E ME N T S YS T E M
It is useful initially to consider the historical evolution of the GATT dispute settle-
ment system. In the beginning, disputes in GATT were considered by working par-
ties and since both disputants were represented, often no clear consensus decisions
were reached. But early in its history, the GATT Contracting Parties started using
panels of experts to consider and report on disputes, with the reports being adopted
by the Contracting Parties or the GATT Council. While a report could be blocked if
a party refused to go along with a consensus decision to adopt it, the system worked
reasonably well, at least with respect to less difficult trade disputes. Over time, the
system gradually evolved from a negotiation-centered approach to dispute settlement
to a more rules-based system and its record in resolving disputes was impressive.28
But it was far from perfect. There was no right to have a panel established and
27 In 2001, Mexico prevailed in a NAFTA dispute settlement proceeding with the USA over the US failure
to open its market to Mexican trucking services as provided by NAFTA. NAFTA Panel Report, In the
Matter of Cross-Border Trucking Services, USA-MEX-98-2008-01 (6 February 2001). The USA has yet
to implement fully the results of the panel report. A pilot program with the aim of doing so was imple-
mented in 2011. It has been challenged in court, although so far not successfully. ‘Teamsters Petition for
Rehearing In Program Allowing Mexican Truck Entry’, BNA International Trade Daily, 5 June 2013.
28 The basic history of GATT dispute settlement is found in Robert E. Hudec, Enforcing International Trade
Law (Salem, NH: Butterworth, 1993). Hudec found that complainants in the GATT system through
1989 achieved complete vindication of their legitimate complaints in about 60% of the cases and partial
vindication in another 29%. Ibid, at 337. Given the lack of real enforcement power in GATT and the fact
that sanctions aimed at achieving compliance with panel decisions were never imposed, this record was
impressive.
7
composed, and any party (usually the losing party) could prevent the formation of a
consensus to adopt the panel report, thereby leaving it in limbo without effect and
the dispute unresolved.29 Interestingly, in a Journal of World Trade Law article from
January 1979, John Jackson stressed the desirability of a rules-based system of dis-
pute settlement in GATT.30 He also cataloged the shortcomings of the then current
GATT system and proposed specific changes. Among those changes were: (i) if con-
sultations or mediation failed to resolve a dispute, the Director would establish a
panel on request,31 (ii) the Director would propose seven panelists from a standing
Panel List and the parties could reject two each, leaving three to serve on the panel,32
(iii) normally the parties would be required to accept the panel’s decisions (absent,
for example, the grant of a waiver),33 and (iv) a surveillance procedure.34 John also
noted that panels should pursue ‘a restrictive method of interpretation and applica-
tion’ of the rules, rather than an ‘expansive’ method.35 None of these ideas was
adopted in the Tokyo Round, but some were in the Uruguay Round and others,
such as the Panel List, are even now under discussion, while the issue of interpret-
ative methods remains controversial. In any event, partly because of ideas put for-
ward by John and others, the Uruguay Round negotiations produced a dispute
settlement system that is unique in the international arena—a rules-based system
covering a broad range of obligations and binding virtually all major nations, one
that has compulsory (and exclusive) jurisdiction.
This major achievement came about when the European Union and Japan, who
tended to favor a less judicial dispute settlement system because they viewed the
GATT process as essentially a negotiating process, decided to accept a more judicial-
like system in return for the US agreement to use the new system in lieu of taking
unilateral action under so-called Section 301 of the 1974 US Trade Act.36 Now in
the WTO system, the process of establishing panels and adopting the results is auto-
matic. It can be blocked only if there is a consensus to do so, which has never
occurred since the prevailing party would have to go along. To guard against the
automatic adoption of defective panel reports, an Appellate Body was created with
29 For a discussion of the problems of the GATT system, see William J. Davey, ‘Dispute Settlement in
GATT’, 11 Fordham International Law Journal 51 (1987).
30 See generally John H. Jackson, ‘Governmental Disputes in International Trade Relations: A Proposal in
the Context of GATT’, 13 Journal of World Trade Law 1 (1979), at 3–4 [Hereinafter ‘Jackson, Proposal’].
See also John H. ‘Jackson, The Crumbling Institutions of the Liberal Trade System’, 12 Journal of World
Trade Law 93 (1978).
31 Jackson, ‘Proposal’, above n 30, point 133. The Director would be in charge of the dispute settlement
function and would be assisted by a staff. Jackson assumed that these positions would be filled by the
GATT Secretariat.
32 Ibid, at 16, point 133.01-.02.
33 Ibid, at 17–18, point 134.2.
34 Ibid, at 20, point 151–153. This procedure allowed for the Director on his own initiative or by request
(including requests by citizens) to investigate possible rule violations. If the Director found good grounds
to believe that there was a rule violation, he could bring that to the attention of countries adversely af-
fected by the violation.
35 Ibid, at 10.
36 William J. Davey, ‘The WTO Dispute Settlement Mechanism’, 1–3 (25 June 2003). Illinois Public Law
Research Paper No. 03-08. Available at SSRN: http://ssrn.com/abstract¼419943 or http://dx.doi.org/
10.2139/ssrn.419943 (visited 11 August 2014).
8
the power to review panel reports. Its reports are also automatically adopted, absent
consensus to the contrary. Thus, the Uruguay Round produced a much more judi-
cial-like dispute settlement system for the WTO.37 So, how has that system worked
out? How effectively has it operated to date? Has it lived up to expectations?
I II . T H E C U R R EN T S T A T E O F TH E W T O
DISPUTE SETTLEMENT SYSTEM
There are a number of factors that suggest that the WTO dispute settlement system
is performing well. As I discuss in more detail below, WTO members have made ex-
tensive use of it since 1995 and the system remains quite busy. Second, the consult-
ations required by the WTO Dispute Settlement Understanding (DSU)—the
nonjudicial step in the process—have worked well in settling many disputes.
Moreover, for those cases that have resulted in panel and Appellate Body reports
finding a measure to be in violation of WTO rules, respondent members have gener-
ally removed or amended the offending measure so as to bring it into conformity
with their WTO obligations. Finally, the Appellate Body has successfully established
itself as the final arbiter of the meaning of the WTO agreements. While there are a
few serious problems, as noted below, overall the system must be given good grades.
A. Usage rates
WTO members have implicitly endorsed the WTO dispute settlement system by
making extensive use of it. There are four particular features of member usage that
warrant comment. First, the pattern of usage among WTO members has varied over
time. Initially, in the first five years of the system’s operation, the overwhelmingly
dominant users of the system were the USA and the European Union.38 These two
members have remained the most frequent users of the system since then, but they
have not been as dominant as they were in the early years. Instead, starting in 2000
or thereabouts, there was a significant expansion of use by developing countries, and
particularly the large developing countries.39 The extensive use by developing coun-
tries has continued to date.40 Second, in the past few years, China has joined the
USA and the European Union as a frequent user, probably in large part in response
to the frequent use of the system against it by those two members.41
Third, in terms of respondents, the USA is by far the WTO member whose meas-
ures are challenged the most frequently in WTO dispute settlement. Indeed, since
2000, the USA has been the respondent in 47% of the panel reports issued (and in
54% of the Appellate Body reports issued).42 In the first five years of WTO dispute
settlement, it was the respondent in only 21% of the panel reports issued.43 The EU
37 See generally Ibid.

38 William J. Davey, ‘The WTO Dispute Settlement System: The First Ten Years’, 8 Journal of International
Economic Law 17 (2005) [The First Ten Years], 17–18.
39 Ibid, at 24.
40 William J. Davey, ‘The WTO Dispute Settlement System at 18: Effective at Controlling the Major
Players?’, EUI Working Paper RSCAS 2013/29, at 5.
41 Ibid.
42 Ibid. at 4 (n.9) & 6 (n.14).
43 Ibid.
9
is a distant second, as it has been the respondent in only 15% of the panel reports
issued since 2000, a figure that was the same in the early years.44 The third most
often challenged country in recent years is China. All other WTO members lag far
behind on this metric.
Fourth, in recent years there has been a noticeable drop in annual consultations
requests since the 1997 peak of 50. The average for the years 2009–11 was 13 con-
sultation requests per year. However, recently there has been an upsurge in dispute
settlement activity. There were 27 such requests in 2012 and 21 in 2013.45 I doubt
this moderation in the number of annual consultation requests signals any dissatisfac-
tion with the system on the part of the WTO membership, rather I tend to think
that it is due mainly to better case selection by the parties—they have a better sense
of when it would be useful to initiate a case—and secondarily to the fact that there
was some pent-up demand to bring cases at the end of the Uruguay Round, which
resulted in a larger than normal number of consultation requests in the early years of
WTO dispute settlement.46 In any event, by any measure the system has been quite
active and can be said to have reached a normal, sustainable level of consultation
requests.
Overall, from the cases that are brought, it appears that the system continues to
serve the function of providing a forum for resolving significant disputes between the
major trading nations—the Airbus-Boeing dispute would be a prime example, but
there are many cases involving the USA, the European Union and China—and the
system continues to serve as a forum for challenges by other nations to actions taken
by those Big Three nations—Brazil’s challenges to EU sugar subsidies and US cotton
subsidies would be examples. Thus, based on its usage, the system seems to be func-
tioning well.
B. Dispute resolution: results of consultations and DSU recommendations

The consultation process has been effective in resolving many disputes. Based on
studies I have done on two separate periods of WTO dispute settlement, it appears
that close to one-half of the cases are disposed of appropriately at the consultation
stage.47 While the parties often do not notify a solution to the DSB as they are sup-
posed to do, further research—such as interviews with the individuals involved in
the case or an examination of the status of the challenged measure—confirms this
high disposition rate. In more recent years, the percentage of settlements through
44 Ibid.
45 These statistics are from the WTO website’s chronological list of consultation requests. Available at
http://www.wto.org/english/tratop_e/dispu_e/dispu_status_e.htm (visited 11 August 2014).
46 Of the first 100 consultation requests, 49 led to the establishment of a panel and 36 resulted in a panel re-
port. During the period from 1 January 2010 to 30 June 2012, there were 38 consultation requests, 29 of
which led to the establishment of a panel (76%) and it appears that a higher percentage of these estab-
lished panels will in fact issue reports than was the case in respect of the first period.
47 William J. Davey, ‘WTO Dispute Settlement: Promise Fulfilled?’, in Inge Govaere, Reinhard Quick and
Marco Bronckers (eds), Trade and Competition Law in the EU and Beyond (Cheltenham: Edward Elgar,
2011), at 194, 197–99 (looking at the consultation requests made in the 2005–07 period); Davey, ‘The
First Ten Years’, above n 38, at 45–46 (looking at the consultation requests made between 1995 and 30
June 2002).
10
consultations has fallen as a higher percentage of cases end up before panels,48 but
the consultation requirement still seems useful as many cases are settled at that
stage.
As to those cases that result in a panel and/or Appellate Body report, the DSU
recommendations in the vast majority of cases are implemented. That is to say, the
noncompliant measure is brought into conformity with WTO obligations or there is
a mutually agreed solution to the case. In the two studies mentioned above, I found
that in general during the periods examined, there was a implementation rate of
80–85%.49 Indeed, if one takes a longer-term perspective and asks in which disputes
implementation has never been achieved, it can be said that for the most part,
there are only a few, relatively insignificant cases that have remained on the DSBs
surveillance agenda. For example, at the DSB meeting of 25 April 2014, there were
10 items on the surveillance part of the agenda and in respect of one of the items the
reasonable period of time for implementation had not expired.50 In three cases, im-
plementation may have been overdue, but only by a year or two. One of those cases
involved a US antidumping measure on shrimp challenged by Vietnam, where some
of the contested implementation issues are soon to be considered by a new panel
sought by Vietnam in respect of related US antidumping measures on shrimp.51 The
second case involved a Filipino challenge to various Thai customs measures involv-
ing Thai treatment of Filipino cigarettes, a matter which has been implemented for
the most part.52 The third case concerned USA claims that China has failed to imple-
ment completely the decision in the China – Electronic Payment Systems case.53
Implementation was long overdue in the remaining six cases. Four of those cases
involved EU and/or Japanese complaints against US measures that overall involve
relatively little trade (Section 211 – Havana Club; Section 1105 – Irish Music)54 or
have been partially implemented or settled (Hot-Rolled Steel and Byrd).55 One
48 See n 46 above.
49 Davey, ‘WTO Dispute Settlement’, above n 47, at 197–200 (looking at the cases started in the 2005–07
period); Davey, ‘The First Ten Years’, above n 38, at 46–48 (looking at the cases started between 1995
and 30 June 2002.
50 WT/DSB/W/523 (23 April 2014). The reasonable period of time had not expired in the Canada –
Renewable Energy case.
51 In the case in question (DS404), the USA reports monthly to the DSB that the Commerce Department
is in the process of implementing the DSB recommendations. See, e.g. WT/DS404/11/Add.23 (13 May
2014). Essentially similar issues are involved in DS429, in which a panel was established on 27 February
2013.
52 WT/DS371/15 & Add.1-24 (13 April 2012–13 May 2014).
53 See, e.g. WT/DSB/M/339, at 11 (10 March 2014). The USA has not commenced a compliance action.
54 The Section 211 – Havana Club case (DS176) essentially arose out of a dispute between two private com-
panies (Pernod Ricard and Bacardi) over the ownership of the Havana Club trademark. In my view, the
violations found are minor and could be easily corrected, but for complications arising from close con-
gressional oversight of all aspects of US policy towards Cuba. The underlying facts of the case and other
litigation involving Bacardi and Pernod Ricard are detailed in ‘Bacardi Has “Cuban Heritage” and Thus
Does Not Mislead With Its Use of “Havana Club”’, BNA International Trade Daily (16 April 2010). The
Section 110(5) – Irish Music case (DS160) was found to involve $1.1 million per year in a DSU Article 25
arbitration. WT/DS160/ARB25/1 (9 November 2001).
55 In Hot-Rolled Steel (DS184), the USA has implemented the DSB recommendations with respect of the
calculation of the antidumping margins at issue in the case, but Congress has failed to enact certain statu-
tory changes that are also required by the decision. WT/DS194/15/Add.127 (14 June 2013). In the Byrd
11
involved a US complaint against the EU (Biotech) that the EU has settled with the
other complainants.56 Finally, there is the Antiguan challenge to US restrictions on
cross-border internet gaming, a matter that the USA has negotiated settlements with
all other interested WTO members.57 Thus, while it would be nice to see these items
finally removed from the DSB’s agenda, they are not major cases.
While this implementation record is very good, two caveats must be mentioned.
First, in many cases, as described below, implementation was not at all achieved in a
timely matter. Second, in some cases, it is likely that the complainants have not pur-
sued the matter—for cost or whatever reasons—even though they believe that full
implementation has not occurred.
C. The performance of the Appellate Body

The Appellate Body has been a notable success of the new WTO dispute settlement
system. It was not clear at the end of the Uruguay Round how the experiment of
adding appellate review would play out, but overall, the Appellate Body has produced
a rich, unified jurisprudence interpreting a broad range of obligations under a wide
variety of WTO agreements. When the WTO came into existence, much of the op-
position to it grew out of concerns that the dispute settlement process would strike
down desirable environmental and health measures. Largely due to Appellate Body
decisions in such cases as US Gasoline (DS2), US Shrimp (DS58), EC Hormones
(DS26), and Brazil Tyres (DS332), I think that those fears have lessened over
time.58 The concerns have not disappeared, however.59 This is largely due to the fact
that seems to be an inexhaustible supply of potentially controversial cases. As of early
2013, one such case was the US Clove Cigarettes case (DS406), in which the
Appellate Body ruled against a US law that served a clear health purpose. It is not
case, Congress repealed the offending provision in February 2006, effective as of 1 October 2007, but
some antidumping duties collected before that date are still awaiting distribution. ‘Customs Announces
Intent to Distribute Byrd Amendment Funds for Fiscal Year 2013’, BNA International Trade Daily, 31
May 2013. As time passes, the amount of funds that might still be distributed (and thus be subject to EU
or Japanese sanctions) is gradually declining to insignificance.
56 The USA has not pursued its request to suspend concessions in this case (DS291). While the USA and
the EU have significant disagreements over what should be the appropriate policy on GMOs, it is not
clear to me that the DSB recommendations in this case are still particularly relevant at this point in time.
Indeed, the EU’s settlements with Canada and Argentina simply established a regular bilateral dialog on
the application of biotechnology to agriculture. WT/DS292/40 (17 July 2009) (Canada); WT/DS293/
41 (23 March 2010) (Argentina).
57 For a discussion of the outcome of this case, see John H. Jackson, William J. Davey and Alan O. Sykes,
Legal Problems of International Economic Relations, 6th ed. (St. Paul, MN: West, 2013), at 1099–1101.
58 The WTO ministerial meeting in Seattle in 1999 produced violent protests that have not occurred at its
subsequent ministerial meetings. Among the marchers were demonstrators dressed as turtles in protest of
the decision in US Shrimp, which they feared would allow imports into the United States of shrimp
caught by methods that harmed endangered turtle species. However, the USA was able to implement that
decision without really changing its turtle protection rules. In the recent US Tuna II (Mexico) case
(DS381), US rules on dolphin-safe labeling of tuna were successfully challenged. Indeed, environmental
groups first became concerned with GATT as a result of GATT dispute settlement cases involving those
labeling rules. The GATT Tuna-Dolphin cases did not lead to any change in US rules and it seems likely
that the USA will be able to implement the more recent case without compromising its basic policy on
tuna labeling.
59 See, e.g. Public Citizen’s Global Trade Watch website, available at http://www.citizen.org/
Page.aspx?pid¼1328 (visited 11 August 2014).
12
yet clear whether the Appellate Body decision will present serious implementation
problems for the USA. However, to the extent that the decision in Clove Cigarettes
means that the USA cannot ban flavored cigarettes unless it also bans menthol cigar-
ettes (which have a large market share—in excess of 25%—so as to potentially make
a ban politically infeasible), there will be very harsh criticism of the Appellate Body
for striking down a very good, politically feasible health measure because it was not a
perfect health measure.
So far, the Appellate Body has been fortunate in that the cases where many obser-
vers think it has misinterpreted the agreements involve trade remedy issues of little
concern to the public at large. Indeed, the issues are so technical that even trade ex-
perts may have difficulty becoming exercised over them. For example, it was com-
monly thought that one achievement of the Uruguay Round was to bring an end to
certain gray-area measures by promoting the use of safeguards regulated by the new
Agreement on Safeguards. If this was the goal of negotiators it is probably fair to say
that the Appellate Body’s jurisprudence on the need to establish ‘unforeseen circum-
stances’—a requirement of GATT Article XIX not found in the Safeguards
Agreement—was a surprise to them. Moreover, its interpretation of the causation/
nonattribution requirement has made it very difficult to defend safeguard measures
when challenged. The effect of all this is unclear, however, since safeguards are typic-
ally imposed for only about three years and the process of challenging such a meas-
ure in WTO dispute settlement takes almost that long in a typical case, at least if the
reasonable period of time for implementation is included in that time frame. Thus,
the overly restrictive interpretations may have had little practical effect on preventing
use of safeguards.60 The same lack of concern and effect may also be found in respect
of the Zeroing cases, which involve a complex and somewhat obscure issue as to how
to calculate average levels of dumping. While panels resisted the Appellate Body’s
rulings as badly reasoned and some commentators questioned its decisions as well,
the fact that antidumping laws are not well liked by many muted the criticism.
Moreover, the actual impact of the decisions remains unclear as there is an excep-
tion—the scope of which is yet to be explored in WTO dispute settlement—that
would appear to allow use of zeroing in certain circumstances. Further, given the
broad discretion of national antidumping authorities in calculation matters, any im-
pact of the ‘zeroing’ decisions may be blunted in many cases.61
All in all, the Appellate Body has generally received good reviews, even though
most of its decisions are criticized by some commentators on at least some issues.
D. Cautionary notes: lack of timeliness, limited results, and gap-filling

It would be disingenuous, however, to suggest that the WTO dispute settlement sys-
tem has not had problems. First and foremost is timeliness. The process takes far
longer than initially envisaged. Most significantly, panels tend to take much longer to
circulate their reports than called for in the Dispute Settlement Understanding.
Indeed, over time the timeliness record of panels has gradually worsened. Through
60 For a discussion of the difficulties in using the Safeguards Agreement, see Jackson, Davey and Sykes,
above n 57, at 771–820.
61 See generally ibid, at 838–43, 871–86.
13
September 2006, the median time for initial panel reports from panel establishment
to report circulation was twelve and half months, with a post-2002 median time of al-
most fifteen months.62 In the 15 cases where initial panel reports were circulated in
2011 and 2012, the median time between establishment and circulation had
increased to seventeen months, almost twice as long as the nine months specified in
the DSU.63 While the Appellate Body had tended to meet its ninety-day deadline for
many years,64 it has not done so in the past couple of years in many cases. For ex-
ample, the median time for the eight Appellate Body reports issued in 2012 was
seven months from date of the panel report to date of the Appellate Body report.65
This is compared to a maximum of five months permitted under the DSU.66 In the
case of compliance proceedings under DSU Article 21.5, panels have typically taken
much longer than the ninety days suggested in the DSU.67
Timeliness of implementation is also a major concern. I noted that implementa-
tion almost always occurs, but in my study of implementation results, it was not com-
pleted within the reasonable period of time in roughly 60% of the cases.68 Indeed,
only in December 2012 did the WTO announce the formal end of the Bananas case,
about twenty years after the first case led to a blocked GATT panel report and al-
most fourteen years since the expiration of the reasonable period of time for imple-
mentation in the WTO case.69
Second, it is important to realize that not infrequently complainants do not
achieve much at the end of the day. By that I mean that implementation, even if
timely does not always have much practical benefit for the prevailing party. For ex-
ample, in safeguard cases—where a country imposes a temporary import restriction
to allow an injured domestic industry to adjust to import competition, implementa-
tion has usually been timely, but the overall time taken from the commencement of
the case until the expiration of the reasonable period of time has meant that the chal-
lenged safeguard has been in place without compensation having been paid for all or
most of its initially intended term.70 In dumping cases, which involve claims that im-
ported products are being sold below their normal price or at a loss and have injured
62 William J. Davey, ‘Expediting the Panel Process in WTO Dispute Settlement’, in Merit E. Janow, Victoria
Donaldson and Alan Yanovich (eds), The WTO: Governance, Dispute Settlement and Developing Countries
(Huntington, NY: Juris Publishing, 2008), at 409, 415–18.
63 Based on information from the WTO website under Dispute Settlement.
64 Davey, ‘Expediting the Panel Process’, above n 62, at 418. The normal deadline is actually sixty days, but
there is a possibility of a thirty-day extension. DSU Article 17.5.
65 Based on information from the WTO website under Dispute Settlement.
66 The five-month period is a maximum in that the appeal is to take place not later than sixty days after cir-
culation of the panel report and the Appellate Body is to issue its report no later than ninety days after
the appeal.
67 Davey, ‘Expediting the Panel Process’, above n 62, at 420–21.
68 William J. Davey, ‘Evaluating WTO Dispute Settlement: What Results Have Been Achieved Through
Consultations and Implementation of Panel Reports?’, in Y. Taniguchi, A. Yanovich and J. Bohanes
(eds), The WTO in the Twenty-First Century: Dispute Settlement, Negotiations and Regionalism in Asia 98
(Cambridge: Cambridge University Press, 2007), at 112–13.
69 ‘Historic signing ends 20 years of EU-Latin American banana disputes’, WTO News Item, 8 November
2012, available at http://www.wto.org/english/news_e/news12_e/disp_08nov12_e.htm (visited 11
August 2014).
70 Davey, ‘Evaluating WTO Dispute Settlement’, above n 68, at 110.
14
the domestic industry, implementation has often been effected through a reconsider-
ation of the issue challenged, which means a new dumping or injury analysis is con-
ducted. During one period I studied about 40% of the implemented cases did not
result in a significant change in the level of antidumping duties because the new ana-
lyses essentially confirmed the initial results.71 Thus, the positive implementation
data must be taken with a grain of salt.
Third, although the overall quality of decisions has been high, there are certainly
instances when the Appellate Body has engaged in gap-filling, ignoring John
Jackson’s advice noted earlier that a restrictive—not an expansive—interpretative ap-
proach should be used.72 Of course, gap-filling is inevitable in the judicial process, es-
pecially when the charge to the system is to clarify the meaning of the agreements in
light of the customary rules of treaty interpretation. Nonetheless, it raises problems
in the WTO system, where the legislative branch—that is to say, the negotiating
function—is not working. The result is that dispute settlement decisions may be
made that the WTO membership generally disagrees with, but about which it cannot
effectively take action. So far, I would not say that this has been a major problem,
but it lurks. I would note that this was a major concern of the Sutherland Report, in
a section that undoubtedly was heavily influenced by John. The Report’s proposal,
which was to have a small number of cases analyzed in depth by a reasonably impar-
tial special expert group of the DSB, so as to provide a measured report of construct-
ive criticism for the information of panels and Appellate Body, is not one that I
would endorse, but there is a need for some way for the political system to ensure
that the negotiated rules are appropriately interpreted.73
Notwithstanding these cautionary notes, I am basically positive overall on the cur-
rent state of WTO dispute settlement. It has defused and settled many difficult cases,
including many involving the major players.
I V . T HE F UT UR E P R O S P E C T S OF TH E W T O
DISPUTE SETTLEMENT SYSTEM
What are the future prospects and challenges for the WTO dispute settlement sys-
tem? Simply put, by far the major challenge will be to ensure the maintenance of the
system’s current level of effectiveness. Indeed, as I noted earlier, given the problems
on the negotiating front, the focus of attention on the WTO will be on its dispute
settlement system in the near-to-medium term and a key component of how effect-
ive the dispute settlement function is viewed will be how it handles disputes involv-
ing China as a complainant and a respondent, particularly in cases involving one or
more of the other major powers. While so far there have not been serious implemen-
tation problems in such cases, as noted below, there have been too few cases and it is
accordingly too soon to draw any firm conclusions about how effective the system
will be in dealing with such cases. A lot is riding on this unknown. Moreover, given
71 Ibid, at 110–11.
72 See text accompanying n 35.
73 The so-called Sutherland Report was formally entitled ‘The Future of the WTO: Addressing Institutional
Challenges in the New Millennium’ and published by the WTO in 2004. It is available at http://www.
wto.org/english/res_e/publications_e/future_wto_e.htm (visited 11 August 2014). The proposal men-
tioned in the text is at page 56.
15
the shortcomings in dispute settlement that I just identified—in particular, those re-
garding timeliness and implementation—there are some specific changes that could
be made to the system to improve its performance and ensure that it remains the
crown jewel of the WTO, as one Director-General apparently described it. I would
suggest three areas where reforms would be useful.
A. The USA, the European Union, and China: do

they implement adverse decisions?
As noted above,74 the overall implementation rate for WTO dispute settlement is
good, with only a few cases having not been implemented or otherwise settled at this
point in time. As also noted, however, implementation often is not timely.75 In this
part, I look closely at how the USA, the European Union, and China have imple-
mented adverse WTO dispute settlement decisions.
In the case of the USA, its overall implementation record is good, although not
great. In a recent study of US implementation, I reached the following conclusions.76
The USA generally complies with adverse WTO rulings, but not promptly.
Compliance has been overdue by more than a year in approximately one-half of the
cases that the USA has lost (17 out of 43 cases were more than one year overdue; 20
of 43 were implemented within the reasonable period of time). This record undoubt-
edly undermines the WTO system by making tardy compliance a standard practice,
but the fact that compliance usually occurs in the significant cases is positive in the
sense that the world’s leading economic power eventually accepts rules-based dispute
settlement. Moreover, the principal cases of non-compliance are due to a failure of
Congress to act (all except the zeroing cases), which suggests the US administrations
have continued to act responsibly towards the WTO. As noted above,77 the US cases
due for action that remain unimplemented and unsettled as of June 2014 are gener-
ally not of much economic significance. Thus, I conclude that the US record is good,
but not great.
In the case of the European Union, at the time of the study, implementation was
due in 17 cases and was timely in 13 cases.78 While full compliance has been ques-
tioned in a number of these cases, no further proceedings have been initiated, and
only in the Sugar case have the complainants continued to raise compliance ques-
tions over time.79 There have been only four problematic EU cases—Bananas
(DS27), Bed Linen (DS141), Hormones (DS26), and GMOs (DS291). The first three
are now resolved, although it took many years to resolve Bananas and Hormones.
Only the GMOs case is still on the DSB surveillance agenda and that case has been
settled with the two non-US complainants. Thus, the EU record can be described as
very good.
74 See text accompanying n 49 above.

76 These statistics are taken from Davey, ‘The WTO Dispute Settlement System’, above n 40, at 18, at 9–11
and 15–17 (Annex II).
77 See text accompanying nn 51–57 above.
78 These statistics are taken from Davey, ‘The WTO Dispute Settlement System’ above n 40, at 18, at
11–12 and 18 (Annex III).
79 See, e.g. WT/DSB/M/325, at 10 (15 February 2013) (DSB meeting of 19 November 2012).
16
For the future, I think that it is clear that the key issue is how China behaves in
WTO dispute settlement. In the past ten years, it has become quite active as a party
in dispute settlement. In the past ten years, 31 consultation requests (involving 19
matters) were made to China, and 11 of those matters went to a panel. During the
same period, China initiated 11 consultation requests, nine of which led to the estab-
lishment of a panel. So far, China’s implementation record has been very good. Of
the seven cases where implementation was due as of June 2014, implementation was
announced in a timely manner and seems to have been accepted in four cases. Of
the remaining three cases, one case is in compliance proceedings, and in the other
two cases, the USA has questioned compliance, but has not pursued compliance pro-
ceedings.80 However, since China has not lost many cases to date, it is difficult to
predict on the basis of past experience whether in the future it will faithfully imple-
ment cases it has lost.
One factor in how willing China may be to accept losses in dispute settlement and
implement adverse panel/Appellate Body decisions will be the record of implementa-
tion in cases that China has initiated against others. So far, China has prevailed in four
cases against the USA and two against the European Union, and the times for imple-
mentation have passed in all six cases. The USA implemented two of the cases in a
timely fashion.81 The other four cases involved antidumping and countervailing duties.
In each case, the USA or the European Union claimed that it had implemented the de-
cision and, in each case, China questioned whether implementation was complete.
China has initiated compliance proceedings in one case (EC – Fasteners (China))
(DS371) and has brought a new case that revisits the issues in another case (US –
Anti-Dumping Methodologies (China))(DS471). Since implementation in antidumping
and countervailing duty cases usually involves re-determinations by national author-
ities, there are often disagreements over whether the new determination is WTO con-
sistent, and complainants often do not obtain much relief from the challenged
duties.82 Thus, the Chinese experience is typical. So far, I would say that China’s own
experience in WTO dispute settlement as a complainant would not give China an ex-
cuse not to implement adverse WTO dispute settlement decisions as a respondent.
Accordingly, as noted at the outset, the ability of the WTO dispute settlement
system to resolve disputes involving, and especially between, the major players will
be critical to maintaining the WTOs overall reputation. To date, it appears that the
system will be able to do this. However, the experience so far with China as both a
complainant and respondent is too limited to draw any firm conclusions in this re-
gard. Moreover, it is worth noting that the recent accession of Russia could eventu-
ally present challenges similar to those currently presented by China.
80 In one case – China Publications and Audiovisual Products (DS363) – implementation appears to have
been somewhat delayed, but seems to have largely occurred. Among other things, the case involved access
of films to the Chinese market, and the USA has recently asserted that the US–China agreement on that
subject did not completely resolve the case in the US view, although initially the USA had seemed to ac-
cept Chinese implementation actions. Davey, ‘The WTO Dispute Settlement System’ above n 40, at 18,
19 (Annex IV); WT/DSB/M/328, at 9–11 (23 March 2013) (DSB meeting of 28 January 2013).
81 In the Steel Safeguards case (DS252), the US promptly withdrew the safeguard measure. In the Poultry
(China) case (DS392), the challenged US measure expired during the panel process.
17
B. Useful reforms to improve operation of

the WTO dispute settlement system
In the review of the operation of the WTO dispute settlement system in Part II, sev-
eral problems were noted. In particular, the system has serious problems in terms
of timeliness. The panel process takes much longer than it is supposed to take and
losing respondents too often fail to implement DSB recommendations within the
reasonable period of time that they are given to do so. While there are many useful
reforms that might be made to the WTO dispute settlement rules,83 I will focus on
several that potentially would speed-up the panel process84 and encourage prompter
implementation of DSB recommendations.85 Without such reforms there is a possi-
bility that government and businesses will come to view the WTO system as subopti-
mal and look for other venues to resolve disputes. If that happened, it would be a
serious blow to the standing of the WTO, especially given its problems noted earlier
in effectively fulfilling its negotiation function.
1. Improving timeliness
i. Changing the panel timeline At the panel level, there are two reforms that would
significantly reduce the time taken by the panel process. First, the standard panel
timeline in DSU Appendix 3 needs to be modified in three ways. At present, the
standard panel working procedures allow the complainant up to six weeks to file its
first written submission after the panel is composed. This is unnecessary since the
complainant should know what its case is about and should be able to file its first
written submission as soon as the panel is composed. Moreover, it would be fairer to
respondent, who is given only up to three weeks to file its submission in response,
even though the respondent does not know what precise claims the complainant will
make. Such a change would also be consistent with the practice of the Appellate
Body, which requires the appellant to file its written submission at the same time it
files its notice of appeal.86 This would save six weeks in the typical case.
83 There have been negotiations on revising the WTO Dispute Settlement Understanding almost constantly
since 1998. Generally speaking, these negotiations have not made a lot of progress because the WTO dis-
pute settlement system functions sufficiently well, such that there is no particular pressure to implement
reforms. Moreover, since the DSU negotiations are part of a broader set of negotiations, there is a ten-
dency for some WTO members to withhold agreement on DSU reforms in hopes of gaining some advan-
tage in the other negotiations. For a discussion of the history of the reform negotiations and a discussion
of many of the proposed reforms, see William J. Davey, ‘Reforming WTO Dispute Settlement’ (29
January 2004). University of Illinois, Public Law & Legal Theory Research Paper No. 04-01. Available at
SSRN: http://ssrn.com/abstract¼495386 or http://dx.doi.org/10.2139/ssrn.495386 (visited 11 August
2014). For the most recent (as of this writing in June 2013) summary of the reform proposals currently
under discussion, see Dispute Settlement Body – Special Session, ‘Compilation of Recent Draft Legal
Text’, JOB/DS/14 (28 May 2013) and ‘Summary of Issued Raised in Consultations’, JOB/DS/15 (24
June 2013).
84 See generally Davey, ‘Expediting the Panel Process’ above n 62.
85 See generally William J. Davey, ‘Sanctions in the WTO: Problems and Solutions’, in Chad P. Bown and
Joost Pauwelyn (eds), The Law, Economics and Politics of Retaliation in WTO Dispute Settlement
(Cambridge: Cambridge University Press, 2010), at 360–72.
86 Appellate Body Working Procedures 21(1).
18
Second, the interim review process provided in DSU Article 15 should be elimi-
nated. Under the current timeline, that process takes five weeks at a minimum. The
interim review process is unnecessary in a system with appellate review. As a practical
matter, panels do not make significant changes in their legal reasoning or results as a
consequence of the interim review process. While the interim review process may
help ensure that the panel has the facts straight in its report in some cases, that func-
tion could be performed by allowing parties to request the panel to correct misstate-
ments of facts or argument following the issuance of the final report to the parties.
Since it takes several weeks at a minimum for the report to be translated, there would
be time for such a process without adding to the overall timeline. This would save
five weeks at a minimum in the typical case.
Third, instead of allowing the parties to take sixty days to decide whether or not
to appeal as is now permitted by the DSU, parties should be required to appeal
within thirty days of the circulation of the panel report. Given the time taken by
translation, this would essentially give the parties about two months to decide
whether to appeal, which seems to me to be plenty of time, even in governments
with a complicated inter-agency review process of trade policy decisions.
These three changes could be made without affecting fairness or the quality of
panel reports, as they would not change the time now given to panels to decide
cases. The changes would, however, eliminate roughly thirteen or fourteen weeks, i.e.
three months, from the typical panel time line.
ii. Permanent panelists A second desirable reform to the panel process would be to
adopt a system of permanent panelists, which would mean that a specified group of
individuals would be appointed to a standing body and all panels would be selected
from that group. Under such a system, there would be considerable time-savings in
the composition process. With a standing body from which panelists would always
be chosen, the panel could be composed shortly after its establishment. This would
save some 10 weeks in the average case according to Steger.87 Additional time sav-
ings would likely occur because meetings of the panelists could be scheduled more
easily. Moreover, a more experienced group of panelists would be able to develop
fairer and more efficient procedures for handling cases and produce better reports. It
would also facilitate the introduction of such procedural improvements as remand.
Other advantages mentioned by Steger include greater diversity in the nationality of
panelists (currently the USA, the EU, and China are grossly underrepresented on
panels considering their involvement in the system as parties), an improved percep-
tion of independence and impartiality (assuming that the group had its own dedi-
cated staff) and the possibility of more effective training in judging skills.88
87 Debra Steger, ‘Establishment of a Dispute Tribunal in the WTO, in Trade and Development
Symposium: Perspectives on the Multilateral Trading System’ (ICTSD 2012).
88 Ibid. In strong support of a permanent panel body, see also William J. Davey, ‘The Case for a WTO
Permanent Panel Body’, 6 Journal of International Economic Law 175 (2003) and William J. Davey, ‘A
Permanent Panel Body for WTO Dispute Settlement: Desirable or Practical?’, in Daniel L.M. Kennedy
and James D. Southwick (eds), The Political Economy of International Trade Law: Essays in Honor of
Professor Robert E. Hudec (Cambridge: Cambridge University Press, 2002), at 496–527. For a collection
19
2. Encouraging prompt implementation

While the WTOs experience with implementation of DSB recommendations is posi-
tive, as noted above, implementation often occurs only after considerable delay.89
There are two reforms that would likely improve the speed of implementation. First,
the time by which implementation is due could be tightened. Second, the remedies
available in the event of non-timely implementation could be strengthened.
i. Setting the reasonable period of time Under the DSU, implementation is due at the
end of the reasonable period of time, which is usually set by agreement of the parties
or by arbitration.90 However, in the case of the Subsidies Agreement, the reasonable
period of time for implementation is specified by the panel if a prohibited subsidy
has been found and is specified by the Subsidies Agreement as six months if an ac-
tionable subsidy has been found.91 In contrast, the DSU states that as a guideline the
reasonable period of time ‘should not exceed 15 months’. As a consequence, the rea-
sonable periods of time for implementation in nonsubsidy cases tend to be consider-
ably longer than six months. Since I am not convinced that implementation is more
difficult in subsidy cases, I think that a general six-month reasonable period of time
for implementation should be specified in the DSU. While it is true that such a short
period would make it more difficult for some countries to achieve timely implemen-
tation than at present, I see no reason to cater to the peculiarities of the legislative/
regulatory processes in defaulting members. The end of the reasonable period of
time acts as the starting point at which the defaulting member may suffer some eco-
nomic consequences for its violation of the agreements. Since the DSU does not pro-
vide for reparations, but only forward-looking ‘compensation’, the time that
compensation becomes available should not be unduly delayed past the point at
which a violation has been finally determined by the DSB adoption of panel/
Appellate Body recommendations. I would argue that six months is a reasonable
time period that appropriately balances the need of the defaulting member to have
some grace period to bring its measures into conformity with the WTO Agreement
and the right of the prevailing member to have timely redress of the continuing inju-
ries caused by the violation. By moving the deadline for implementation (and poten-
tial retaliation) forward, more serious efforts at compliance could be expected. The
result would be prompter implementation of DSB recommendations.
3. Strengthening WTO remedies for nonimplementation

The WTO remedy in the event of nonimplementation is suspension of concessions,
or, more simply, retaliation.92 There are two key factors to keep in mind in respect
of diverse views on the issue, see ‘Mini-Symposium on the Desirability of a WTO Permanent Panel
Body’, 6 Journal of International Economic Law 175 (2003).
89 See text accompanying nn 68–69 above.
90 DSU, Article 21.
91 Subsidies Agreement, Articles 4.7 (prohibited subsidy), 7.9 (actionable subsidy). Since prohibited subsi-
dies are viewed, at a minimum, as no less serious breaches than actionable subsidies, panels have tended
to set the reasonable period of time for implementation in prohibited subsidy cases at six months or less,
except in the US FSC case (DS108).
92 DSU, Article 22. See generally Davey, ‘Sanctions in the WTO’,above n 85, at 360–72.
20
of retaliation under WTO rules. First, any authorized retaliation is prospective in na-
ture. It can be authorized only after expiration of the reasonable period of time and
after the level of retaliation has been accepted or arbitrated. Second, retaliation has
two goals—first, to restore the balance of concessions that was upset when one
member violated its obligations (a temporary aim since compliance is the preferred
result—indeed, I think that John Jackson has persuasively established that it is the
mandatory result93); and second, to give that member an incentive to comply.
The structure of this ultimate WTO remedy has two unfortunate consequences.
First, although the WTO is designed to be a force for trade liberalization, the reba-
lancing that takes place with retaliation occurs at a lower level of trade liberalization
than had been agreed to and, even more disturbing, usually has adverse effects on
the country imposing the retaliatory measure since its consumers of the targeted
products will face higher prices or short supplies as a result of the higher tariffs typic-
ally imposed and the domestic importers and sellers of those products will lose busi-
ness.94 Moreover, when a small or developing country is the prevailing party and the
defaulting party is a major power, it is likely that retaliation by the prevailing party
have a quite adverse effects on its own economy and questionable whether it will
have any noticeable effect on the defaulting party’s economy. This disparity in the
ability to use the WTO’s ultimate remedy is a problem in itself. Second, with a sys-
tem of prospective remedies such as the WTO offers, there is an incentive for the de-
faulting party to delay the time at which point they might be imposed, such as by
seeking a long reasonable period of time for implementation and then taking ques-
tionable implementation measures, thereby forcing the prevailing party to go through
a DSU Article 21.5 panel (and Appellate Body) compliance proceeding.95
The first-mentioned problem could be mitigated in the arbitration setting the
level of nullification/retaliation by giving the prevailing party the right to receive
monetary compensation (in lieu of authority to retaliate) if it convinces the arbitra-
tors that retaliation would be ineffective.96 The level of monetary compensation
would be calculated in the same way that the level of retaliation is now calculated, as
modified by the proposal in the next paragraph. The use of monetary compensation
in such circumstances would solve the above-mentioned disparity problem, would
not be self-defeating and would not result in further anti-trade liberalization meas-
ures, all problems associated with retaliation.
Assuming that WTO members will want to retain a system of prospective
remedies,97 it would be possible to modify the way that the level of retaliation is
93 John H. Jackson, ‘International Law Status of WTO Dispute Settlement Reports: Obligation to Comply
or Option to “Buy Out”?’, 98 American Journal of International Law 109 (2004).
94 See, e.g. BNA International Trade, ‘U.S. Announces Final List of EU Imports Targeted for Higher Duties
in Beef Dispute’ (20 July 1999).
95 There have been 155 panel reports adopted and 27 compliance (21.5) panel reports adopted, so disputed
compliance measures are not uncommon. WTO Appellate Body, Annual Report 2012, at 104, WT/AB/
18 (9 April 2013).
96 Davey, ‘Sanctions in the WTO’, above n 85, at 364.
97 In considering the proposals made in this paragraph, it is worth remembering that the standard remedy
in international law for breach of obligations is reparations. International Law Commission, Article on
State Responsibility, Article 31(1). The proposals do not attempt to introduce that concept into WTO
rules, but only to calculate and adjust levels of retaliation or compensation so as to encourage prompter
21
calculated so as to create incentives for early compliance. In this regard, one could
imagine a system providing that the amount of retaliation (or compensation) will be
calculated from a date prior to the date set for implementation (such as the date of
adoption of the panel/Appellate Body reports) and/or by providing for annual in-
creases in the level of retaliation (or compensation) over time in the event that
noncompliance continues for more than a year. Either of these changes would raise
the cost of noncompliance and thereby encourage prompter implementation.98
V . C O NC L U S IO N
The current state of the WTO dispute settlement system is good. It has performed
quite well notwithstanding the controversial cases with which it has been presented.
While the challenges of dealing with disputes involving China (and eventually
Russia) will undoubtedly test the system, its experience in dealing with disputes
involving the USA and the European Union bodes well. Nonetheless, improvements
in timeliness—both in the panel process and in the implementation of results—are
imperative if the system is to continue to be relevant to business and governments
facing trade problems.
compliance with DSB recommendations. A WTO member that complies promptly would continue to
face no penalty for its default, notwithstanding the damage caused by its violation of WTO rules.
98 Davey, ‘Sanctions in the WTO’, above n 85, at 365; William J. Davey, ‘Remedies in WTO Dispute
Settlement’, in John Jackson, Davey and Sykes, above n 57 at 396–98.
22
=-
1 Pauwelyn's Argument
App This section evaluates Pauwelyn's argument
in light of the relevant dsu provisions.
Sett
1.1 Analysis of Article 3
of the DSU
PRABHASH RANJAN The purpose of the dsu is given in Article

3.2 of dsu - the dispute settlement system
Since the
of of the wto is form
the a central element un
in "provid-
Uruguay
ing security and predictability to the mul- r
Dispute Settle
that led to the establishment of tilateral trading system". The panel in us
the World Tra
the World Trade Organisation (wto) was - Section 3014 held that providing security
an interesting
the formation of the Dispute Settle- and predictability to the multilateral trad-
context of
ment Body (dsb) with compulsory juris- ing set
regime is the central object and pur-
in diction to settle trade disputes

the wtoamong pose of the system.5 Further, the
is dsb has
wto member countries. This marked to "preserve the rights and obligations of
process is lim
a shift from the diplomatic, power-
members under the covered agreements"
or, can
oriented approach to a rule-basedthe
regime en
of the wto and clarify the existing provi-
international sions of those agreements in accordance
to settle trade disputes.1 An interesting l
question in the context of
article settling
endeavo with customary rules of interpretation of
trade disputes in the wto is - "whatpublic
is international law. Article 3.2 also
question by pr
the applicable law for the dsb?". Isimposes
it an important obligation on the
of certain
limited to wto law enshrined in the issu
panel not to make such recommendations
covered agreements2 or, can the entire
Pauwelyn, and rulings to the dsb thatwh
will "add to or
body of public international law be diminish
used the rights and obligations pro-
entire body of
in settling disputes among the videdwto in the covered agreements".6 In oth-
law is
member countries? applicab
er words, the panel cannot create new
This article endeavours to answer norms for the member countries, since
this question by providing a critique
this of
is the job of the member countries. It
certain issues raised by the well-known
is important to bear in mind that the wto
legal expert Joost Pauwelyn inis his
a member-driven organisation, and only
paper, which argues that the entire
the member countries have the power to
body of public international law is appli-
make new rules or modify existing rules.
cable law for the dsb (Pauwelyn 2001,
An important consequence of Pauwe-
Bartels 2001). lyn's argument will be to undermine the
Pauwelyn has argued that the sub-
security and predictability of the wto.7 It
stantive jurisdiction of wto panels3 is
is important to understand that the wto
limited to claims under the wto covered as a trade treaty reflects an attempt by
agreements, but the applicable law is member countries that are at different lev-
the entire gamut of public international els of development, to strike a fine balance
law. Pauwelyn distinguishes between between their rights and obligations so
claims based on non-WTO law from that the entire system brings mutual ben-
efits. The trade rules that come into exist-
defence made on non-WTO law (Tracht-
man 2004a). In other words, according
ence are the result of years of hard negoti-
to Pauwelyn, a wto member country ations and involve huge economic and
cannot bring a case in the dsb against
political bargains (Trachtman 2004b: 861).
another member country for not This delicate balance, if disturbed, will
honouring a non-WTO public inter-
result in damaging consequences for the
The author expresse
national law obligation, but can use overall multilateral trade system, especially
Trachtman for his c
topic public
for international law as aadefence differen
for developing countries and least deve-
to an to anonymous
deviate from its wto obligations. loped countries. This can be understood re
Views and
This article errors
offers the critique by analys- with the help of the following example. if
responsibility.
ing the Dispute Settlement Under- Assume a case where country "A"
standing (dsu) provisions and theRanjan
Prabhash wto (developed country) starts imposing a tar- (p
com) is
jurisprudence with
that do not support iff rate on the WB
product X coming from country
Juridical Sciences, K
Pauwelyn's argument. "B" (developing country) that is higher
Economic & Political weekly 13321 april 11, 2009 vol xliv no 15 23
This content downloaded from 14.139.213.70 on Thu, 04 Jan 2018 09:21:46 UTC
All use subject to http://about.jstor.org/terms
23
COMMENTARY =: =
than the most

panel To examine,
adding favo
in the light of the relevant pro-
visions in (name of the covered agreement(s)
(the countries,
increase in the w
cited by the parties to the dispute), the mat-
dsucoming
products of the w
fr
ter referred to the dsb by (name of party) in
challenge
Whenthis
document ...a actio
and to make coun
such findings as
lation the
of wto,
Article it
will assist the dsb in making the I ex
recommen- o
Tariffs anddations orTrade
countries in giving the rulings provided
wil for
mfn rule.
ple of in that/those
Suppose
non- agreement^). di
states form
that of
this
Further, Article 7.2 of the hig
dsu states: "Pan- m
imposed
out because
discrimin
els shall address the relevant provisions in
particular human
tions, is
any covered agreement orthe
agreements cit- r
both A try. and Allowin ed B
by the parties are
to the dispute". pa
that fence on non-WTO
it law will disturb the
follows According to Article 7, the the
panel has to
honour and
balance of rights and respect
obligations of mem- examine the "matter"9 in the light of the
courage exports
ber countries by creating fro
new obligations, relevant provisions of the covered agree- .
it thinks do
not contained in the wto treaty, for some ments. Subsequent
not to this examination,
ho
such a
of defence
them. The be
wto treaty does not allowthe panel has to make "such findings" that a
Applying
the modification of the provisions Pauwelyn
of thewill assist the dsb in making the recom-
of A isin thisacceptable.
agreement manner. mendations or rulings provided in the cov- H
such a defence^
Pauwelyn has argued that different ap- ered agreements. Here, Pauwelyn arguesba
jus cogens,
plication of wto rules by different wtothat referenceas its
to "such findings", in Article m
the mfn
members (dependent on whetherrulenon-WTO 7.1 implies that the panel does
may have to re-
the law has been complied with) is an unavoid- sort to international law or non- wto
exceptions givlaw
the defence of
able consequence (Pauwelyn 2001: 567). (Pauwelyn 2001: A
562). This argument is i
crease in
This argument the
goes against the very basis untenable. Themfn
words "such findings" t
this would
of the wto agreement given in the pream-needmean
to be read with the preceding words t
only if it
ble that countries an in- and in the overall context of Article 7.the
fulfils
resolve to develop
In other words,
tegrated, more viable and durable multilat- The panel, after examining the "mat- th
would recommend
eral trading regime. Pauwelyn's argument ter" (i e, the specific "measure" [actions of t
fulfilwould create
the a system that has different
human a country] and the "claims" [complaints r
on it rules
by for different countries.
virtueIf a non- wtoagainst that action] related to of
that meas- t
exists
law is made a between
condition for a particularure) brought before it, may arrive at either he
panel would
wto member add
country to do international of the following two conclusions. First, the t
tions of
trade B
without the same non-WTO lawto
be- trade
specific actions (measure) of one country b
take ing made the condition for other
any such wto is in accordance with the covered
oblig agree-
Thus,member the panel
countries, it will result in creat- ments and hence the "claims" or the com- wi
to ing a multilateral trading system, which
Articles 3.2 plaint of the other country
and is dismissed. 1
this will of common Second, the specific
is not integrated on the basis also action (measure) is
imply
for different
set of rules and obligations barring those wto inconsistent, and coun
hence, the com-
instance,
exceptions which the system itselfif there
has plaint of the other country is upheld. In
that created.
is Such a system
not will also not be both the cases
partythe panel will judge the
durable because the economic gains that "measure" and
convention, it the "claim" related
may to it, in
the the wto member countries
mfn expect to accordance with the covered agreements
tariff rate
related point
flow from the multilateral trading regime and not in accordance is
with non-WTO law.
th
defence, the
will get frustrated because the non-WTO After this examination, the panel pamay
examining wheth
law obligations will be imposed on them. make either of the following two findings.
withIt is neither
a non-WTO
integrated nor durable. In the former case, the defending country
human rights).
may continue with its actions (measure) S
1.2 Terms of Reference
implies of the
that because it is consistentthe
with the wto. In p
Panel:
jurisdictionArticle 7 the latter case,
andthe action (measure) be b
of We now turn to Article 7 of the dsu,
non-WTO removed or modified
disput as per the relevant
gest that wto
which describes the terms of reference
honourinof provisions. "Such findings" in
wto panels and hence
gations is is more not Article 7.1 needs to be understood
directly impo in this
related to the issue of applicable
superior to law.
context. human
Pauwelyn's argument will lead to
is used
Article 7.1 states: as a thirda publ
kind of situation, where a country's
example action
of (measure) is not in accordance
Panels shall have the followingintende
terms
Pauwelyn's
reference...: with thethesis
covered agreements and yet the
24 April 11, 2009 vol xliv no 15 Q32I Economic & Political weekly
24
== - _ COMMENTARY
panels have to examine

complaint claims and de- assist the dsb in making the recommenda-
(claims) of th
related to that measure cannot succeed fences in accordance with the covered tions or in giving the rulings provided for in
the covered agreements.
because the "measure" is invoked as a de- agreements, which form the basis of
fence based onnon-wTO law to which claims and defences in wto disputes. According to Article 11, the panel has
both the parties are bound.10 the overall function to assist the dsb in
Had the parties intended to allow the
In other words, this will be a strange
panel to use law other than that givendischarging
in its responsibilities under the
situation where the panel will make thea covered agreements, Article 7 should
dsu and other covered agreements. Fur-
finding that the action of a country have
is reflected such intention. In that ther, under this overall function, the panel
case Article 7.1 could have been drafted
inconsistent with the covered agreements has to do three things:
as follows:
but is permissible since it is consistent (1) To carry out the objective assessment
with the non-WTO law. This, from the un- of the matter and facts of the case.
To examine in the light of the relevant provi-
sions in (name of the covered agreement®
derstanding of Article 7, is something that (2) To examine the applicability and con-
and other agreements cited by the partiesformity
the panel cannot do. Nothing in the dsu to of the matter and facts with the
the dispute) the matter referred to the dsb
provides for a wto inconsistent action to relevant covered agreements.
by (name of party) in document. ... However,
be maintained due to its consistency with (3) To make findings that will assist the
Article 7.1 does not contain the words 'other
non-WTO law. It is also important to em-agreements'. Therefore, Professor Pauwe- in making recommendation or rulings
dsb
phasise here that Pauwelyn has argued as provided in the covered agreements.
lyn's analysis is based on words that are not
written in Article 7.1. 12 Hence, it is the duty of the panel to
that there is no presumption that in case
of conflict between non-WTO law and wtoMoreover, there is little textual support
apply the law given in the covered agree-
for the difference between claims based
law, the latter would prevail (Pauwelyn ments to the facts and circumstances of
on non-WTO law and the applicable de-the case and to find out their applicabi-
2001: 564). Which rule would prevail will
fence based on non-WTO law (Trachtmanlity and conformity with the covered
depend on the conflict rules of interna-
2004a: 857), the distinction made by Pau-agreements. Article 11 nowhere mentions
tional law. This argument further rein-
welyn. The argument that non-WTO lawthat the objective assessment of the
forces the possibility of the third situation
discussed above (Bartels 2001: 507-09). can be used as defence because there is matter and the facts of a case are to be
assessed according to non-WTO law. Had
nothing in the wto treaty that prohibits
Pauwelyn has argued that despite the
the application of non-WTO law to settle
obligation to address and possibly apply this been the intention of the parties, the
the wto rules, there is nothing in the disputes
wto panel should have been put under an
among member countries is un-
tenable. If "express prohibition" is the
(wto has not contracted out from the ap- obligation by the member countries
yardstick,
plication of public international law) that then there is also nothing under
that Article 11 to assess the facts in
precludes panels from addressing and accordance with the covered agreements
prohibits claims based on non-WTO law to
and other sources of law. Pauwelyn
be brought to the dsb. However, according
applying non-WTO law to decide on wto
to Pauwelyn, this is not possible. Then,
claims (ibid). It has also been argued that argues that the reference in Article 11 to
applying the same logic, application "other
a judicial system aimed at enforcing cer- of findings", also supports the accept-
tain rules of international law need not non-WTO law as defence in the dsb ability of defence based on non-WTO
should also not be allowed. How is it that
refer explicitly or conform to the applica- law (Pauwelyn 2001: 562). However,
reading
the "automatic confirmation", referred
tion of potentially relevant rules of inter- to the phrase "other findings" in
Article 11 with the words preceding it
national law (Pauwelyn 2001: 561-62). Heby Pauwelyn, of the rules of international
does not support this argument. The
goes on to argue that such reference orlaw happens only in cases where inter-
national law is used as a defence and not
confirmation occurs automatically as a re- phrase other findings is preceded hy the
sult of the simple fact that the dsu wasas a claim? Thus, Pauwelyn draws anfollowing
un- words: "....the applicability of
tenable distinction between claims and
created and continues to exist in the wider and conformity with the relevant agree-
applicable defence based on non-WTO
context of international law, and hence, ment and make such other findings..."
other rules of international law apply au-law (ibid). According to the ordinary and plain
tomatically (ibid). This is not supported by meaning of the words this means that the
1.3
the language of Article 7 or of any other Analysis of Article 11 panel first has to determine whether a
Similarly, Article 11 of the dsu also
provision of the dsu. The dsb may exist in particular measure is applicable as per
the wider context of international law, but
provides: the covered agreement and conforms to
its basic function is to settle trade disputes
The function of panels is to assist the dsb
the relevant covered agreement. If it
among wto member countries that may in discharging its responsibilities under does not, then the panel has to make
arise from their rights and obligations this Understanding and the covered agree- findings that will bring the measure in
given in the covered agreements and not ments. Accordingly, a panel should make an conformity with the covered agreements.
objective assessment of the matter before
from their rights and obligations outside "Other findings" does not mean that the
it, including an objective assessment of the
the wto treaty. The panels or the Appellate panel is required to find out whether the
facts of the case and the applicability of and
Body (ab)11 are not courts of general conformity with the relevant covered agree- particular "measure" is in conformity
jurisdiction. It is clear from Article 7 that ments, and make such other findings as will with non-WTO law.
Economic & Political weekly D3S3 april 11, 2009 vol xliv no 15 25
25
COMMENTARY _JZ- - ^î^^îî:
A combined international law or non-WTO law the ba-

for the dsb extends beyond the covered
reading of
sions reveals sis of substantive claims or
agreements. It only supports the factthat
that defences in the
the p
is to wto does not exist in isolation
find from the
whether wto disputes. There is evidence to sup-
c a
has been
larger body of publicviolated.
international law. Inport the former but there is nothingSub
to
finding, the
fact, this point was made by the ab in the support the latter argument. Eachwill
panel time
very first
dations to dispute in us-Gasoline.14
the Subse-the abdsb
refers to the wto as not
toexisting in as
try to quently, the ab and the panels have re-splendid isolation
modify its from the greater
mea body
obligations.
ferred to many other non-WTO laws In arrivin
inof public international law, it supports the
sion, the panel
various disputes such as in us-Shrimp,15former argument andwillnot the latter. a
wto law
the Convention inand
International Trade on not
The ab has ruled out thenon-possibility of
2002: 759).
Endangered Species of Wild Flora andusing non-WTO law as a defence to violate
Fauna was referred to understand the the wto obligations. In EC-Hormones,17
2 Argument on WTO term "exhaustible natural resources" where the European Commission used the
Jurisprudence occurring in Article xx (g) of gatt (ibid,
precautionary principle to impose stricter
Pauwelyn has cited the panel's decision on para 25-28). Likewise, in us-Shirts sanitary
and and phytosanitary (sps) meas-
Korea-Government Procurement13 case in ures,
Blouses,16 the ab adopted the practice of the ab held that this principle had
numerous international tribunals includ-
support of his thesis that Article 7 does not not been written into the sps agreement
restrict the law applicable to the wto dis-ing the International Court of Justice (icj) as a ground for justifying measures that
pute resolution to the covered agreements that the "burden of proof" is on the party are otherwise inconsistent with the
that makes the assertion.
(Pauwelyn, op cit, p 562). In this case, the agreement (ibid, para 124). This clearly
panel said that Article 7.1 of the dsu In the debate on applicable law in thesuggests that only exceptions given in the
wto it is fundamental to draw the distinc- sps agreement should be used to main-
does not restrict the ability of the panel to
tion between the use of international law
refer to the broader rules of customary tain sps measures that are otherwise
international law. However, this does not
or non-WTO law in clarifying the wto inconsistent with other provisions of the
covered agreements and in making the sps agreement. Stricter sps measures
support the assertion that applicable law
Ééfs EndGdvour ^.scs^mS

^^^T ) Payment by DD in \
"W (RGNIYD's Journal of Youth
^^ iSriperumbudui^(
RGNIYD brings out a half-yearly Jou
to the diverse voices of those concern
from various disciplines and profession
and provides a forum for exchange of
planners and executives concerned wi
Articles/papers for the Journal 'Endea

policy makers, thinkers, planners, exe
will be provided on request. Broad th
■ Leadership ■ Personality Develo

■ Adolescents ■ Entrepreneurship ■
Development ■ Social Harmony and
Rajiv Gandhi National Institute of

University u/s 3 of UGC Act 1956 and
Sports, Government of India, functio
Extension and Outreach besides Docum
Contributions/Articles/Sub
The Faculty Head
READ Division
RGNIYD, Sriperumbudur - 602 105
Phone: 044-27162401 ■ Fax: (091) 044-27163227
Website: www.rgniyd.gov.in
- E-mail: rgniyd@yahoo.com/fhread@rgniyd.gov.in
26 April 11, 2009 vol xliv no 15 E33S3 Economic & Political weekly
26
===
cannot Report of the Appellate

from wto obligations to secure compli- be Body, WT/DS60/AB/R, i
2 November 1998.
ance by another country under ani
rules, inter- e,
10 Pauwelyn has argued that defence based on non- n
cautionary WTO can be invoked only when both the coun-
national agreement (non-WTO law, in this p
tries are bound by the concerned non-WTO law.
In case nafta).23
EC-Poultr See Pauwelyn (2001), p 566.
This discussion illustrates that while
oilseeds 11 The permanentagre
judicial body of the WTO.
12 Also see Palmeter and Mavroidis (1998),
dispute
the wto panels may use non-WTO law to PP 398-413. be
(Trachtman
interpret or clarify the provisions given 13
inKorea - Measures Affecting Government Procure-
ment, Report of the Panel, WT/DS163/R, 19 June
the covered agreements, this certainly 2000.
oilseeds agre
cannot lead us to the conclusion
basis of that coun-
14 US - Standards forthe
Reformulated and Conventio
Gasoline, Report of the Appellate Body, WT/DS
such as
tries start using non-WTO laws as defence the
AB/R, 29 April 1996.
ment
for violating wto agreements. Shrimpn was
15 US - Import Prohibition of Certain
Shrimp Products, Report of the Appellate Bod
within the w
WT/DS58/AB/R, 12 October 1998 (hereinafte
3 Conclusions
agreement US-Shrimp). w
The dsb of the wto exists to settle trade 16 United States - Measures Affecting Import
lateral oblig
Woven Shirts and Blouses from India, Report
Brazil
disputes between countries that arise as a purs
the Appellate Body, WT/DS33/AB/R, 25 Apr
1997.
(ibid). Furth
result of the interplay between their rights
17 European Community - Measures Concerning Meat
oilseeds
and obligations contained in the covered and Meat Products ag
(Hormones), Report of the
agreements. The panels or the ab in the Appellate Body, WT/DS26/AB/R, WT/DS48/
applied pur
AB/R, 13 February 1998.
Marrakesh
dsb have not been given the mandate to 18 European Communities - Measures Affecting the
function like courts of general jurisdiction
wto.20 Importation of Certain Poultry Products, Appellate
Pauw
Body Report, WT/DS69/AB/R, 13 July 1998
in by the member countries of the wto.
this case
(hereinafter EC - Poultry).
Therefore, not just the claims that coun- 19 EC Poultry, para 79.
agreement w
20 EC Poultry, para 80.
cause
tries bring to the dsb but it form
also defence 21 EC Poultry, para 83.
should be based on wto law. There is no
fence (Pauw
22 Mexico - Tax Measures on Soft Drinks, Appellate
Body Report, WT/DS308/AB/R, 6 March 2006
oilseeds
legal basis to disallow claims but allow de- (hereinafter
agr
Mexico - Soft Drinks).
fence based on non-WTO international the
fence, 23 Mexico - Soft Drinks, para 72. ou
rent.
law. Both are not allowed, although Howe
public
international law can be used to interpret
supported REFERENCES
by
the wto treaty.
The ab Any change in this man- made
Bartels, Lorand ( 2001): "Applicable Law in WTO D
pute Settlement Proceedings", Journal of Worl
date of the dsb will imply adding to the Trade fJWTJ, 35,
applying P499. th
it existing set of obligations on countries
was not in
Cameroon, James and Kevin R Gray (2001):a "Prin
ples of International Law in the WTO Dispute Set
it the wto and hence, modifying the exist- tlement Body", International
would hav and Comparative
ing obligations and rights.
even if This, in turn, Law Quarterly
the (ICLQ), 50-2, pp 248-98.
Marceau, Gabrielle (2002): "WTO Dispute Settleme
invoked
will undermine the security and predicta- as
and Human Rights", European Journal of Inter
held
bility of the wto. that
national Law (EJIL), 13(4), pp 753-801. b
Palmeter, David and Petros Mavroidis (1998): "T
oilseeds agr
wto Legal System: Sources of Law", AJIL, 92(3
supplementPP 398-413-
notes
Pauwelyn, Joost (2001): "The Role of Public Inter
pursuant
1 Petersm
to
national Law in the WTO: How Far Can We Go?
ventionAmerican Journal on
(2001). Al
of International Law (AJIL), 95
- Understanding on Rules and Procedures (3), PP 535-78.
Given
Governing the Settlement of Disputes
the a
(hereinaf- Petersmann, E U (1997): The G ATT/WTO Dispu
that
ter DSU). the ab
Settlement System w
(London: Kluwer Law Inter
2 All the agreements given in the different annexes
national), pp 63-64.
oilseeds
of the WTO agreement are covered agreements.
agr
Trachtman, P Joel (1999): "Domain of WTO Dispu
differently 3 A panel is a temporary body established by the Resolution", Harvard International Law Journa
dsb of the WTO at the request of the complain- (H. INT. L.J), 40(2), pp 333-77.
defence, if one accepts Pauwelyn's ing party to settle a dispute between two - (2004a): "Review of Conflict of Norms in Publ
argument that it was used as a claim and member countries of the WTO (See Article 6 of International Law", AJIL, 98, pp 855-62.
the DSU).
not as a defence. - (2004b) : 'Jurisdiction in WTO Dispute Settlement
4 United States - Sections 301-310 of the Trade Act of in Rufus Yerxa and Bruce Wilson (éd.), Key Issues in
Again, in the Mexico - Soft Drinks22 1974, Report of the Panel, WT/DS152/R, 27 Janu- WTO Dispute Settlement: The First Ten Years (Cam
ary 2000 (hereinafter US-Section 301).
case, one of the issues before the ab was bridge: Cambridge University Press), pp 132-43.
5 US-Section 301, para 7.75.
whether Mexico could use Article xx (d) 6 Also see Article 19.2 of the DSU that also states
of gatt to deviate from the National that the panel cannot add or diminish the rights
and obligations of the countries given in the Subscription Numbers
Treatment obligation in Article m of gatt
covered agreements.
Subscribers are requested to note their
to secure compliance by the us for its 7ob-
Also see Trachtman (1999), (2004b), and Marceau
(2002). Subscription Numbers mentioned on
ligation under North American Free 8 Mexico - Soft Drinks, para 78, also see Marceau
the wrappers and quote these numbers
Trade Agreement (nafta) (non-WTO (2002) op cit.
9 For interpretation of "matter" occurring in Arti- when corresponding with the circulation
law). The ab decided that Article xx (d)
cle 7.1 see Guatemala -Anti Dumping Investiga- department.
could not be used by a country to deviate
tion Regarding Portland Cement from Mexico,
Economic & Political weekly B3Q april n, 2009 vol xliv no 15 27
27
WTO Dispute Settlement 79
submissions available to interested parties on a "real-time" basis. Business confidential

information has generally not played a role in WTO proceedings, but parties should clearly
retain the right to protect such information from disclosure.
wto Dispute Settlement: Emerging Practice and Procedure

by Debra P. Steger and Peter Van den Bossche*
Introduction
In the more than fifty years of General Agreement on Tariffs and Trade/World Trade
Organization (GATT/WTO) history, the dispute settlement system has been remarkable for
its pragmatism and effectiveness. The dispute settlement system of the GATT 1947 evolved
over time from the very sparse foundation provided by Articles XXII and XXIII. Although
procedural rules were codified in certain Decisions and Understandings agreed to by the
contracting parties in 1966, 1979 and 1989,* the Understanding on Rules and Procedures
Governing the Settlement of Disputes (DSU), which emerged from the Uruguay Round
negotiations, represents the first extensive, negotiated agreement reforming and revitalizing
the GATT dispute settlement system. It has been argued that some of the detailed
innovations in the DSU will drive the system ever more into a judicialized model of dispute
settlement. Foremost among these innovations was the establishment of the Appellate Body
in late 1995.
Since that time, the Appellate Body has heard and decided ten appeals. The Appellate
Body consists of seven members, selected from different legal systems. The members serve
for fixed terms, in most cases of four years, which are renewable for another four years. The
Appellate Body sits in divisions of three members, selected on a rotational basis, to hear and
decide individual appeals. The Appellate Body has developed a practice of collegiality
whereby all seven Appellate Body members come to Geneva in the deliberation phase of
each appeal to exchange views on the matters at issue in the appeal. This practice has
promoted consistency in the early decisions of the Appellate Body, which has been
especially important for issues of a systemic or general nature.
It is perhaps somewhat anomalous that the WTO dispute settlement system is now
characterized by, on the one hand, an Appellate Body, which is judicial?indeed, it is much
like an appellate court?and, on the other, panels that function more on the arbitral model.
Appellate Body proceedings are governed by the detailed Working Proceduresfor Appellate
Review,2 which were drawn up by the Appellate Body in early 1996, and which resemble,
in their complexity and specificity, rules of procedure of some courts. An Appellate Body
hearing is a formal judicial hearing, much of which is taken up with penetrating questioning
of the participants and the third participants by the Appellate Body members. In contrast,
panels do not have detailed, standard rules of procedure; rather, there are some very general
provisions set out in Appendix 3 of the DSU. However, panels are free to develop their own
working procedures in individual cases, after consulting with the parties. Panels have
'Debra Steger is Director, and Peter Van den Bossche is Counselor, with the Appellate Body
Secretariat, World Trade Organization, Geneva. The views expressed in this article are the views of
the authors and do not represent the views of the Appellate Body or the World Trade Organization.
1 Procedures under Article XXIII (decision of 5 April 1966) GATT Basic Instruments and Selected
Documents (BISD) 14S/18; Understanding Regarding Notification, Consultation, Dispute Settlement
and Surveillance (adopted on 28 November 1979) BISD 26S/210; Improvements to the GATT Dispute
Settlement Rules and Procedures (decision of 12 April 1989) BISD 36S/61.
2 WT/AB/WP/3,28 Feb. 1997. This is a consolidated, revised version ofthe Working Procedures
for Appellate Review, and replaces WT/AB/WP/1, dated 15 Feb. 1996.
28
80 ASIL Proceedings, 1998
sometimes resolved difficult procedural issues during their proceedings by consulting with,
and trying to accommodate the views of, the parties to the dispute. However, this "process
of accommodation" in individual panel proceedings does not always lead to consistent
results or to a shared perception of fairness and due process by all parties and third parties
to the dispute.
In almost every appeal to date, issues relating to systemic matters or panel procedures
have been raised by the parties. As a result, the Appellate Body has been called upon to rule
on several points of general or procedural law. This has led to a growing "jurisprudence" on
matters such as treaty interpretation, burden of proof, standard of review, status of prior
GATT panel reports and panel procedures. To facilitate a discussion, we have grouped these
general rulings of the Appellate Body under four headings: rights of parties, jurisdiction of
a panel, panel procedures, and treaty interpretation and sources of law.
Rights of Parties
Standing. With regard to the right of a WTO member to bring a dispute settlement claim
under the GATT 1994, the Appellate Body stated in European Communities?Regime for
the Importation, Sale and Distribution of Bananas (EC?Bananas)? that there is no explicit
provision in the DSU requiring that a member have a "legal interest" in order to request a
panel, nor is such a requirement implied in the DSU or in any other provision of the WTO
Agreement. Furthermore, the Appellate Body did not read any of the judgments of the
International Court of Justice and the Permanent Court of International Justice, referred to
by the participants in that appeal, as establishing a general rule that in all international
litigation a complaining party must have a "legal interest" in order to bring a case. According
to the Appellate Body, Article XXIII:1 of the GATT 1994 and Article 3.7 of the DSU
clearly indicate that "a Member has broad discretion in deciding whether to bring a case
against another Member under the DSU," and that "a Member is expected to be largely self
regulating in deciding whether any such action would be 'fruitful'"4
Representation by private counsel. Also in the EC?Bananas appeal, the Appellate
Body was asked whether a certain third participant could be represented in the Appellate
Body oral hearing by private counsel, who were not government employees. In that case, the
panel's working procedures, agreed by the parties to the dispute, provided that only
government officials could appear before the panel. The Appellate Body ruled that "nothing
in the WTO Agreement, the DSU or the Working Procedures for Appellate Review, nor in
customary international law or in the prevailing practice of international tribunals, prevents
a WTO Member from determining the composition of its delegation in Appellate Body
proceedings."5 Thus, the Appellate Body allowed private legal advisers, who were not
government employees, to participate in its oral hearing. The Appellate Body also stated in
EC?Bananas that
representation by counsel of a government's own choice may well be a matter of

particular significance?especially for developing-country Members?to enable them
to participate fully in dispute settlement proceedings. Moreover, given the Appellate
Body's mandate to review only issues of law or legal interpretation in panel reports, it
is particularly important that governments be represented by qualified counsel in
Appellate Body proceedings.6
3Adopted 25 Sept. 1997, WT/DS27/AB/R, paras. 132 and 133.

4Id. para. 135.
57tf.para. 10.
6Id. para. 12.
29
Jurisdiction of a Panel
Specificity ofa requestfor the establishment ofa panel In EC?Bananas, the Appellate
Body stated that it is important that, under Article 6.2 of the DSU, a request for the
establishment of a panel be sufficiently precise, for two reasons: first, it often forms the
basis for the terms of reference of the panel pursuant to Article 7 of the DSU, and, second,
it informs the defending party and the third parties of the legal basis of the complaint.7 In the
same case, the Appellate Body ruled that it is sufficient for a complaining party to list in its
panel request the specific provisions of the particular agreements alleged to have been
violated, without setting out detailed arguments as to which specific aspects of the measures
at issue relate to which specific provisions of those agreements.8 However, in India?Patent
Protection for Pharmaceutical and Agricultural Chemical Products {India?Patents), the
Appellate Body ruled that the phrase "including but not necessarily limited to," used by the
United States in its request for the establishment of the panel, was "simply not adequate to
'identify the specific measures at issue and provide a brief summary of the legal basis of the
complaint sufficient to present the problem clearly' as required by Article 6.2 of the DSU."9
The Appellate Body, therefore, found that the request for the establishment of the panel in
this case fell below the "minimum standards" that it was willing to accept in EC?Bananas.10
The Appellate Body has also ruled that there is a significant difference between the
claims identified in the request for the establishment of a panel, and the arguments
supporting those claims, which are set out and progressively clarified in the first written
submissions, the rebuttal submissions and the first and second panel meetings with the
parties.11 According to the Appellate Body, Article 6.2 of the DSU requires that the claims,
but not all the arguments, must be specified sufficiently in the request for the establishment
of a panel in order to allow the defending party and any third parties to know the legal basis
of the complaint. A faulty request for the establishment of a panel, furthermore, cannot be
"cured" by the subsequent filing of a more detailed written submission.12
Terms ofreference. With regard to the terms of reference of panels, the Appellate Body
has observed that a panel's terms of reference are important for two reasons:
First, the terms of reference fulfil an important due process objective?they give the
parties and the third parties sufficient information concerning the claims at issue in the
dispute in order to allow them an opportunity to respond to the complainant's case.
Second, they establish the jurisdiction of the panel by defining the precise claims at
issue in the dispute.13
Furthermore, the Appellate Body has repeatedly stated that a panel may only consider those
claims that it has authority to consider under its terms of reference, which are often based
on the request for establishment of the panel. To the extent that a claim is not referred to in
the terms of reference, the panel has no authority to consider this claim.14
7 Id. para. 142.

'A/.para. 141.
"Adopted 16 Jan. 1998, WT/DS50/AB/R, at para. 90.
"Y*/. para. 91.
11 EC?Bananas, at para. 141.
l2Id. para. 143.
n Brazil?Measures Affecting Desiccated Coconut (Brazil?Coconut), adopted 20 Mar. 1997,
WT/DS22/AB/R,at22.
uSee, e.g., Brazil?-Coconut, supra note 13, at 22, and India?Patents, at para. 92. The Appellate
Body found in India?Patents that the panel's ruling at the outset of the first substantive meeting, that
all legal claims would be considered if they were made prior to the end of that meeting, was
30
Burden of proof No provision in the WTO Agreement explicitly addresses which party
has the burden of proof in a WTO dispute settlement proceeding. In addressing this issue,
the Appellate Body observed, however, that "various international tribunals, including the
International Court of Justice, have generally and consistently accepted and applied the rule
that the party who asserts a fact, whether the claimant or the respondent, is responsible for
providing proof thereof."15 Furthermore, the Appellate Body has noted that "it is a generally
accepted canon of evidence in civil law, common law and, in fact, most jurisdictions, that
the burden of proof rests upon the party, whether complainant or defending, who asserts the
affirmative of a particular claim or defence."16 The Appellate Body has also ruled:
If that party adduces evidence sufficient to raise a presumption that what is claimed is
true, the burden then shifts to the other party* who will fail unless it adduces sufficient
evidence to rebut the presumption.17
The Appellate Body has also ruled that
a prima facie case is one which, in the absence of effective refutation by the defending
party, requires a panel, as a matter of law, to rule in favor of the complaining party
presenting the prima facie case.18
Regarding precisely how much and precisely what kind of evidence will be required to
establish such a presumption in the context of the GATT 1994 and the WTO Agreement, the
Appellate Body has stated that "this will necessarily vary from measure to measure,
provision to provision, and case to case."19
The Appellate Body has also ruled that this general rule in a dispute settlement
proceeding, requiring that a complaining party establish a prima facie case of inconsistency
before the burden of showing consistency with that provision is taken on by the defending
party, is not avoided by simply describing that same provision as an "exception."20
Standard of review?objective assessment of the matter. There are no provisions in the
DSU or in any of the agreements annexed to the WTO Agreement, other than the Agreement
on Implementation of Article VI of the General Agreement on Tariffs and Trade 1994 (Anti
Dumping Agreement), that explicitly prescribe a standard of review for panels.21 The
Appellate Body has stated, however, that Article 11 of the DSU "articulates with great
succinctness but with sufficient clarity the appropriate standard of review for panels in
respect of both the ascertainment of facts and the legal characterization of such facts under
inconsistent with the letter and spirit of the DSU. Although panels enjoy some discretion in
establishing their own working procedures, this discretion does not extend to modifying the substantive
provisions of the DSU. Nothing in the DSU gives a panel the authority either to disregard or to modify
the explicit provisions of the Understanding, such as Article 7 thereof, pursuant to which the
jurisdiction of a panel is established by that panel's terms of reference.
15 United States?Measure Affecting Imports of Woven Wool Shirts and Blouses (US?Shirts and
Blouses), adopted 23 May 1997, WT7DS33/AB/R, at 14.
l6Id
"Id at 14.
,8EC Measures Concerning Meat and Meat Products (EC Hormones), adopted 13 Feb. 1998,
WT/DS26/AB/R, WT/DS48/AB/R, at para. 104.
l9US?Shirts and Blouses, supra note 15, at 14; see also India?Patents, supra note 14, at para.
74.
20?C Hormones, supra note 18, at para. 104.
2,Only Article 17.6(i) of the Anti-Dumping Agreement has language on the "standard of review"
to be employed by panels, but this provision is specific to the Anti-Dumping Agreement. See Anti
Dumping Agreement, Article 17.6(i).
31
the relevant agreements."22 Article 11 of the DSU stipulates, in part, that it is the function
of a panel to make "an objective assessment of the matter before it, including an objective
assessment of the facts of the case and the applicability of and conformity with the relevant
covered agreements." The Appellate Body has stated that, insofar as fact-finding by panels
is concerned, the applicable standard is neither de novo review as such, nor "total
deference," but, rather, the "objective assessment of the facts."23 Insofar as legal questions
are concerned?that is, the consistency or inconsistency of a member's measure with the
provisions of the applicable agreement?the Appellate Body has also considered Article 11
of the DSU directly on point, requiring a panel to "make an objective assessment of... the
applicability of and conformity with the relevant covered agreements."24
In assessing whether a panel has made "an objective assessment of the fa?,ts of the case"
as required under Article 11, the Appellate Body has posited that the deliberate disregard of
evidence, or the refusal to consider it, as well as the willful distortion or misrepresentation
of the evidence presented to a panel, is incompatible with a panel's duty to make an
objective assessment of the facts.25 The Appellate Body has been careful to stress, however,
that "'disregard' and 'distortion' and 'misrepresentation' of the evidence, in their ordinary
signification injudicial and quasi-judicial processes, imply not simply an error of judgement
in the appreciation of evidence but rather an egregious error that calls into question the good
faith of a panel."26 In no case to date has the Appellate Body found that a panel made such
an "egregious error" in assessing the facts.
Judicial economy. With regard to the principle of judicial economy, the Appellate Body
has stated, on more than one occasion, that nothing in Article 11 of the DSU or in previous
GATT practice requires a panel to examine all of the legal claims made by a complaining
party,27 and that, on the contrary, a panel should address only those claims that must be
addressed in order to resolve the matter in issue in the dispute.28 Given that the basic aim of
dispute settlement in the WTO is to resolve disputes, the Appellate Body did "not consider
that Article 3.2 of the DSU is meant to encourage either panels or the Appellate Body to
'make law' by clarifying existing provisions of the WTO Agreement outside the context of
resolving a particular dispute."29
Panel Procedures
Right to seek information or expert advice. With regard to a panel's right to seek
information and technical advice, the Appellate Body has held that Article 13 of the DSU
provides for "a grant of discretionary authority: a panel is not duty-bound to seek
information in each and every case or to consult particular experts under this provision," but,
on the contrary, has "the discretion to determine whether to seek information or expert
22EC Hormones, supra note 18, at para. 116.

2'Id. para. 117.
27</.para. 118.
25A/.para. 133.
uId.
27The Appellate Body observed that previous GATT 1947 and WTO panels have frequently
addressed only those issues that such panels considered necessary for the resolution of the matter
between the parties, and have declined to decide other issues. See US?Shirts and Blouses, supra note
15, at 18.
2tSee, e.g., US?Shirts and Blouses, supra note 15, at 18, and India?Patents, supra note 14, at
para. 87. In EC Hormones, supra note 18, at paras. 250 and 252, the Appellate Body agreed with the
panel's application of judicial economy, and held that the panel did not err in refraining from making
findings on Article 2.2 and 5.6 of the Agreement on the Application of Sanitary and Phytosanitary
Measures.
29US?Shirts and Blouses, supra note 15, at 19.
32
advice at all."30 The Appellate Body has also held that Article 13.2 ofthe DSU, and Article
11.2 of the Agreement on the Application of Sanitary and Phytosanitary Measures, allow a
panel to seek information and advice in the manner it deems appropriate in a particular case,
and "leaves to the sound discretion of a panel the determination of whether the establishment
of an expert review group is necessary or appropriate."31
Evidence. With regard to the submission of evidence to a panel, the Appellate Body has
called upon all parties engaged in dispute settlement under the DSU to be "folly forthcoming
from the very beginning both as to the claims involved in the dispute and as to the facts
relating to those claims."32 Claims must be stated clearly, and facts must be disclosed freely.
According to the Appellate Body, a party that, in the aftermath of consultations, believes that
all the pertinent facts relating to a claim are, for any reason, not before the panel, should ask
the panel to engage in additional fact-finding. The Appellate Body has noted that "with
respect to fact-finding, the dictates of due process could better be served if panels had
standard working procedures that provide for appropriate factual discovery at an early stage
in panel proceedings."33 With respect to when evidence must be submitted to a panel, the
Appellate Body has noted that at present there are no "hard and fast rules on deadlines for
submitting evidence."34 Neither Article 11 of the DSU nor the Working Procedures set out
in Appendix 3 of the DSU establish any precise time limits for the presentation of evidence
by a party to the dispute.35 However, the Appellate Body has pointed out recently that the
Working Procedures in Appendix 3 do contemplate two distinguishable stages in a
proceeding before a panel: a first stage including the written submissions to the panel and
the first substantive meeting with the parties, and a second stage including the rebuttal
submissions and the second substantive meeting with the parties. Under the Working
Procedures in Appendix 3 of the DSU, the complaining party should "set out its case in
chief during the first stage, while the second stage of a panel proceeding is "generally
designed to permit 'rebuttals'."36
Sources of Law and Principles of Treaty Interpretation

Principles oftreaty interpretation. With regard to the principles of treaty interpretation
that are to be applied when interpreting the WTO Agreement, the Appellate Body has ruled
that the general rule of treaty interpretation, set out in Article 31.1 of the Vienna Convention
on the Law of Treaties (Vienna Convention),37 "has attained the status of a rule of customary
or general international law."38 This rule, therefore, "forms part of the 'customary rules of
interpretation of public international law' which the Appellate Body has been directed, by
Article 3.2 of the DSU, to apply in seeking to clarify the provisions" of the WTO
Ârgentina?Measures Affecting Imports of Footwear, Textiles, Apparel and Other Items

{Argentina?Import Measures), WT/DS56/AB/R, at para. 84, With respect to seeking the opinion of
a specialized agency, such as the International Monet Fund, the Appellate Body has said that this
is within a panel's discretion under Article 13 of the DSU, except where there are specific provisions
in the WTO Agreement requiring such consultation (e.g., Article XV:2 of the GATT 1994).
31?C Hormones, supra note 18, at para. 147.
nIndia?Patents, supra note 14, at para. 94.
33/rf. para. 95; see also Argentina?Import Measures, supra note 30, at n.68.
Ârgentina?Import Measures, supra note 30, at para. 80.
3SId. para. 79.
"Id.
3723 May 1969,1155 U.N.T.S. 331.
38United States?Standards for Reformulated and Conventional Gasoline (US?Gasoline),
adopted 20 May 1996, WT/DS2/AB/R, at 17. Referred to in Japan?Taxes on Alcoholic Beverages
(Japan?Alcohol), adopted 1 Nov. 1996, WT/DS8/AB/R, WT/DS10/AB/R, WT/DS11/AB/R, at 10.
33
Agreement*9 The Appellate Body has repeatedly stressed that pursuant to the principles of
treaty interpretation provided for in Article 31.1 of the Vienna Convention, "the words of
a treaty form the foundation for the interpretative process," and that these words are to be
given their ordinary meaning, in their context and in the light of the treaty's object and
purpose.40 The duty of a treaty interpreter is "to examine the words of the treaty to determine
the intentions of the parties," and this in accordance with the principles of treaty
interpretation set out in Article 31 of the Vienna Convention.41 The Appellate Body has
ruled that "these principles of interpretation neither require nor condone the imputation into
a treaty of words that are not there or the importation into a treaty of concepts that were not
intended."42
The Appellate Body also noted that one of the corollaries of Article 31.1 of the Vienna
Convention is that interpretation must give meaning and effect to all terms of a treaty. The
Appellate Body has stressed that "an interpreter is not free to adopt a reading that would
result in reducing whole clauses or paragraphs of a treaty to redundancy or inutility."43
According to the Appellate Body, the principle of effectiveness is a "fundamental tenet of
treaty interpretation."44
Finally, the Appellate Body has noted that merely characterizing a treaty provision as
an "exception" does not by itself justify a "stricter" or "narrower" interpretation of that
provision than would be warranted by examination of the ordinary meaning of the actual
treaty words, viewed in the context and in the light of the treaty's object and purpose, or, in
other words, by applying the normal rules of treaty interpretation.45
Treatment of municipal law. With regard to the question how a panel should treat
municipal law, the Appellate Body has referred to the judgment of the Permanent Court of
International Justice in Certain German Interests in Polish Upper Silesia, and stated that a
panel should not interpret municipal law as such, but that it is quite appropriate for a panel
to examine the municipal law for the purpose of determining whether a member has met its
obligations under the WTO Agreement.46
Status ofprior GA TTpanel reports. With regard to the legal status of prior GATT panel
reports, the Appellate Body has repeatedly noted that adopted panel reports are an important
part of the GATT acquis. They are often considered by subsequent panels. Adopted panel
reports create "legitimate expectations" among WTO members and, therefore, should be
taken into account where they are relevant to any dispute, although they are not binding,
except with respect to resolving the particular dispute between the parties in that dispute.47
With regard to unadopted panel reports, the Appellate Body stated that such reports have no
legal status in the GATT or WTO system, but noted that a panel could nevertheless "find
useful guidance in the reasoning of an unadopted panel report that it considered to be
relevant."48
"US?Gasoline, supra note 38, at 17.

40Japan?Alcohol, supra note 38, at 11-12.
41India?Patents, supra note 14, at para. 45.
42Id.
43US?Gasoline, supra note 38, at 23; Japan?Alcohol, supra note 38, at 12.
44Japan?Alcohol, supra note 38, at 12.
45EC Hormones, supra note 18, at para. 104.
46India?Patents, supra note 14, at para. 66.
41 Japan?Alcohol, supra note 38, at 14; see also US?Shirts and Blouses, supra note 15, at 19-20.
4%Japan?Alcohol, supra note 38, at 14-15.
34
Conclusion
It has often been said that the beauty of the GATT dispute settlement system has been
its ability to adapt and evolve on a case-by-case basis to meet the needs of the contracting
parties. In this pragmatic, incremental manner, the most effective international dispute
settlement system was built. The DSU represents an extensive, negotiated effort to reform
the GATT dispute settlement rules and procedures and to establish a fully integrated system
for the resolution of disputes under the Uruguay Round agreements. It is interesting to note
that there is already a small but growing corpus of WTO "jurisprudence" on matters of
dispute settlement practice and procedure in the nine Appellate Body and panel reports
adopted by the Dispute Settlement Body to date.
The WTO Dispute Settlement System: New Horizons

by Thomas Cottier
Introduction
The WTO dispute settlement system is one of the three core functions of the World
Trade Organization: negotiations (rule-making), adjudication and implementation of rights
and obligations of members. It is at the heart of the international trade system; particularly
between multilateral rounds of trade negotiations. It assumes a key role and responsibility
in the advancement of the rule of law. And it moves along with the expanding role assigned
to the WTO: from an multilateral trade agreement to what we may call constitutional
functions. Since the days of GATT 1947, the scope and task of the WTO have been
widening. Important trade-related matters were attached to the WTO. The TRIPs agreement
introduced the age of standard setting. The GATS Agreement essentially introduced core
elements of investment protection. Interfacing legitimate trade and environmental concerns
moved to the center of the agenda. Additional areas, in particular rules concerning private
party conduct (beyond anti-dumping), are in the pipeline, and first links to social policy will
continue to be sought. Regulation moved from border measures to key domestic policies.
Foreign and domestic economic regulation no longer can be separated as we no longer can
separate foreign and domestic policies in a meaningful manner.
How does the dispute settlement system, the life and law of which has been experience
and not logic (as Justice Holmes observed), adjust and cope with these constitutional
transmutations? Ever since the basic concept was enshrined in GATT Article XXIII with its
peculiar concept of nullification and impairment, it evolved case by case, adjusting and
finding its most recent expression in the DSU of the Uruguay Round. Indeed, let me start
with concurring that, overall, the system has fared remarkably well since it started to operate
with the Appellate stage in 1995. Looking at it from a multilateral perspective, it
successfully assisted in curbing unilateral trade sanctions, specifically unilateral measures
under Section 301 of the U.S. Trade Act of 1974. The goals of DSU Article
23?Strengthening of the Multilateral System?have been reinforced as trust and confidence
in the mechanism are, it would seem, increasing. Industrialized and developing countries
alike perceive it as a useful tool to further their interest. The rule of law seems to have
gained terrain, as power-oriented attitudes have receded.
Debra Steger and Peter van der Bosche rightly conclude their paper emphasizing that
"in this pragmatic, incremental manner, the most effective international dispute settlement
system was built." It has come a long way. Yet, the most important advances and changes
'Professor of Law, University of Bern, Switzerland (thomas.cottier@iew.unibe.ch). I am indebted

to Debra Steger, Steve Charnovitz, Frieder Roessler and Andy Shoyer for reactions and comments.
35
Why Are WTO Panels Taking Longer?
And What Can Be Done About It?
Matthew KENNEDY
World Trade Organization (WTO) dispute settlement panels have steadily taken longer to complete their
work since the establishment of the WTO, but since 2005, the delays have sharply increased. The Dispute
Settlement Understanding (DSU) time frames for panel proceedings are problematic in light of the time and
number of parties’ submissions. However, unexpected delays in panel work have become more frequent,
which, if they continue to increase at their present rate, will offset reform of DSU time frames at other stages
of the procedures. This article examines the reasons why delays have occurred and considers some possible
solutions.
1. INTRODUCTION
The performance of any system of justice depends not only on the fairness of its proceedings,
the soundness of its rulings, and the effectiveness of its remedies but also on the speed at which
it operates.1 This is often summed up in the maxim that ‘justice delayed is justice denied’.
Nowhere is that maxim more apposite than in a system that lacks both interlocutory and
retrospective relief, such as the World Trade Organization (WTO) dispute settlement system.
Significant delays are occurring at almost all stages of the WTO dispute settlement
procedures.2 Among these, the lengthiest stage in many cases is the panel stage, which is
beyond the exclusive control of the parties. Five panels have now taken more than two years to
issue final reports to the parties.3 These panels comprise a trio of Agreement on the Application
of Sanitary and Phytosanitary Measures (SPS Agreement) disputes involving scientific evidence
and the Airbus/Boeing civil aircraft cases. The time taken by the Boeing panel has attracted
particular criticism.4 These five panel proceedings have demonstrated that the WTO dispute
settlement system lacks any upper limit on the time that it takes to deliver a ruling.

The author was formerly a senior lawyer in the WTO Secretariat. Email: <Matthew.Kennedy2007
@gmail.com>.
1
This was recognized as long ago as the thirteenth century in the Magna Carta, Clause XXIX in fine: ‘we will sell
to no man, we will not deny or defer to any man either justice or right’. (emphasis added).
2
See William Davey, ‘Expediting the Panel Process in WTO Dispute Settlement’, in The WTO: Governance,
Dispute Settlement and Developing Countries, ed. M. Janow et al. (New York: Juris Publishing, 2008), 409–440, 440.
3
Time counted from date of panel composition: EC – Biotech (GMOs), WT/DS291, DS292, DS293; US/Canada –
Continued Suspension of Obligations in the EC-Harmony Dispute, WT/DS320, DS321; EC and certain Member States – Large
Civil Aircraft, WT/DS316; US – Large Civil Aircraft (2nd complaint), WT/DS353; and Australia – Apples, WT/DS367.
4
For example, Pilita Clark, ‘Airbus Fears WTO Delay on Boeing Report Could Hurt Contract Hopes’, Financial
Times, 20 Dec. 2009; Europa Press Release, ‘European Commission Disappointed by Delay to WTO Confidential
Interim Panel Report on Boeing (DS 353), 8 Jul. 2010’, <http://europa.eu/rapid/pressReleasesAction.do?reference¼IP/
10/912>, 16 Sep. 2010. Part of the concern relates to the time taken relative to the Airbus panel.
Kennedy, Matthew. ‘Why Are WTO Panels Taking Longer? And What Can Be Done about It?’.
Journal of World Trade 45, no. 1 (2011): 221–253
Ó 2011 Kluwer Law International BV, The Netherlands
36
222 JOURNAL OF WORLD TRADE
While the five lengthiest panel proceedings draw attention to the problem of delays at
the panel stage, they divert attention from the underlying trend in WTO practice in less
extraordinary cases. The average time taken to issue panel reports has lengthened inex-
orably since the establishment of the WTO, but in recent years, it has arrowed upwards,
even excluding the big five cases. This sharp increase offsets any time that might be saved
by reforming other stages of the Dispute Settlement Understanding (DSU) procedures and,
if it continues at its present rate, could undermine the effectiveness of the system.
This article analyses the trends in the time taken by panels to complete their work,
compares them with the DSU time frames, discusses the reasons that panels are taking
the time that they do, and considers possible solutions. The focus of this article on the
panel stage does not understate the importance of delays at other stages of the WTO
dispute settlement procedures, such as panel composition and implementation of
recommendations.
2. TIME TAKEN TO ISSUE PANEL REPORTS
At the tenth anniversary of the establishment of the WTO, differing assessments were made
of the performance of WTO panels, in terms of timing. One assessment was that ‘a major
weakness of the DSU panel process and appellate review is that disputes still take too long
to resolve’.5 Another was that: ‘[t]he whole process is, for all but a very few cases,
remarkably efficient’.6 The difference between these assessments is easily explained. Each
reflected the particular benchmark that was selected to assess WTO panels’ performance:
the former referred to time frames established in the DSU whereas the latter did not.7
The efficiency of WTO panels can be evaluated first by identifying the time that
panels actually take to carry out their function and then comparing that with the
applicable benchmarks in the DSU and other relevant features of WTO dispute settlement
practice. A standard benchmark comprises the time elapsed between the date of composi-
tion of a panel and the date on which the panel issues its final report to the parties.8 This
benchmark is shorter than the length of an entire panel proceeding as it excludes panel
5
Andrew L. Stoler, ‘Enhancing the Operation of the WTO Panel Process and Appellate Review: Lessons from
Experience and a Focus on Transparency’, Conference on the WTO (New York City, 6 Apr. 2006), 10, later published
in, The WTO: Governance, Dispute Settlement and Developing Countries, ed. M. Janow et al. above n. 2, 525–542, 530. He
added that ‘Members seem to be generally understanding when a panel chairperson or the Appellate Body announces that
due to the complex nature of the arguments needing to be analysed, more time will be required to produce the panel or
Appellate Body reports. There is not much we can do about those cases – but this should be the exception and not the
rule. Clearly, in complex cases, most parties would prefer for the panel to spend enough time on the case to get it right’.
6
Valerie Hughes, ‘Initiating Proceedings to Ensuring Implementation’, in The WTO at Ten: The Contribution of the
Dispute Settlement System, ed. G. Sacerdoti et al. (Cambridge: Cambridge University Press, 2006), 193–294, 196.
7
Hughes cited Art. 12.9 of the DSU but for the proposition that: ‘[f]rom composition of the panel to the issuance
of the final report to the disputing parties it is to take no more than six to nine months’. See the discussion of Art. 12.9
below. She also referred to the practice of non-WTO tribunals.
8
Sources: WTO Secretariat, Update of WTO Dispute Settlement Cases, WT/DS/OV/34, 26 Jan. 2009; 2009
Overview of the State of Play of WTO Disputes, WT/DSB/49/Add.1 and the WT/DS document series. The date of
issuance of the final report to the parties is usually indicated in the procedural aspects or interim review section of the final
report when it is circulated to Members. Where it is not, it has been estimated, usually from the dates of issuance of the
interim report and circulation of the final report.
37
WHY ARE WTO PANELS TAKING LONGER? AND WHAT CAN BE DONE ABOUT IT? 223
composition, before the panel begins work, and translation of the report for circulation to
Members, after the panel has completed its work.
Trends in the time actually taken by panels to complete their work can be analysed on an
annual basis. As annual data sets can be small, averages are sometimes skewed by one or two panels
that adopted extraordinary procedures. The use of medians is more helpful, although medians too
may be influenced by the clustering of extraordinary panel proceedings in certain years. Adjusted
annual averages for panels in more ordinary cases can therefore be calculated as a comparator,
although the choice of which panels to exclude as ‘extraordinary’ is necessarily subjective.
The chart below indicates the median times, and adjusted average times, that panels
have taken to complete their work since the establishment of the WTO. The data comprise
a total of 124 panels that were composed in original disputes each year from 1995 to 2009
and that later circulated a report or that have estimated a period within which they will do
so. The data do not include panels in compliance disputes. The adjusted averages exclude
the big five cases, plus ten other panels that were procedurally unusual in certain ways9 and
the time that elapsed in three disputes prior to replacement of a panellist.10 The data for
2009 are mostly based on estimates from panel Chairs and may increase.
Figure 1. Time Taken to Issue Panel Reports

16
14
12
Months after composition
10
0
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
Year of panel composition

Annual median Adjusted annual average
9
The following panels have been excluded from the averages for the following reasons: Brazil – Aircraft, WT/
DS46; Canada – Aircraft, WT/DS70; Australia – Automotive Leather, WT/DS126; and Canada – Aircraft Credits and
Guarantees, WT/DS222 – accelerated timetables under the SCM Agreement; Mexico – Telecoms, WT/DS204; Mexico –
Anti-dumping Measures on Rice, WT/DS295; Mexico – Soft Drinks, WT/DS308; and Mexico – Olive Oil, WT/DS341 –
translation of panel report prior to issuance to the parties; US – Shrimp (Ecuador), WT/DS335; and US – Carrier Bags,
WT/DS383 – claims not fully contested.
10
EC – DRAMs, WT/DS299; Korea – Commercial Vessels, WT/DS373; US – Continued Zeroing, WT/DS350.
38
The median and average11 times taken for WTO panels to issue a final report to the parties
began at around eight months. The time taken gradually increased by about two months
over ten years but then, in a new departure, sharply increased by double that amount in half
the time.
The cases on EC – Biotech (GMOs), US/Canada – Continued Suspension of Obligations
in the EC – Hormones Dispute, Australia – Apples, Airbus and Boeing took (or are taking)
much longer than indicated on the figure. However, the jump from 2005 onwards cannot
be attributed solely to the delays caused by these disputes as they have been excluded from
the adjusted averages and the trend remains the same.
3. DSU TIME FRAMES FOR PANEL REPORTS
The applicable benchmarks for WTO panels are those in the DSU. The DSU addresses the
issue of procedural speed upfront, locating it among the purposes of the dispute settlement
system. Article 3.3 provides that the ‘prompt settlement’ of disputes is ‘essential’ to the
effective functioning of the WTO and the maintenance of a proper balance between the
rights and obligations of Members. The DSU is also replete with time frames for various
steps in its procedures, which reflects the drafters’ preoccupation with the time taken to
resolve disputes.
‘Time frames’ remain under review in the ongoing negotiations on DSU reform and
are a theme of the 2008 consolidated drafting text.12 Proposals were made in the DSU
negotiations concerning time frames at the panel stage as long ago as 2003. Besides
proposals to reform the panel composition process, there were a few proposals to revise
the time for parties’ submissions in line with current practice;13 or to reduce the overall
time for panel proceedings in particular circumstances.14 Since those proposals were made,
the time taken by panels to complete their work has increased considerably.
The function of panels is set out in Article 11 of the DSU. In concrete terms, it
comprises: (1) making an objective assessment of the facts of the case (findings of fact);
(2) making an objective assessment of the applicability of and conformity with the relevant
covered agreements (findings of law and conclusions); and (3) making such other findings as
will assist the Dispute Settlement Body (‘DSB’) in making the recommendations or in
giving the rulings provided for in the covered agreements (the remedy). Article 12 sets out,
11
The averages for the first three years are not adjusted.
12
See ‘Report by the Chairman of the Special Session of the DSB to the Trade Negotiations Committee (‘‘TNC’’)
for the Purpose of the TNC Stocktaking Exercise’ (22 Mar. 2010), TN/DS/24, para. 2 and fn. 3 referring to Job
document (08)/81, 18 Jul. 2008.
13
Proposals have been made to invert the amount of time for the parties’ respective first written submissions and to
increase the available time for developing country parties’ submissions, as compared to those proposed in Appendix 3 of
the DSU: see TN/DS/W/32 (Japan); TN/DS/W/51/Rev.1 (China).
14
The circumstances include disputes under certain contingent trade remedy agreements: TN/DS/W/49 (Aus-
tralia); TN/DS/W/51/Rev.1 and W/57 (China); where repetition of a violation under a discretionary measure is highly
probable: TN/DS/W/32 (Japan); and where a measure has already been found WTO-inconsistent in a previous dispute:
TN/DS/W/45/Rev.1 (Brazil). See also some ideas from Mexico on diagnosis of the problems affecting the dispute
settlement mechanism, TN/DS/W/90, 13–16.
39
in skeletal form, provisions on panel procedures, including the timetable, parties’ submis-
sions, a written panel report where the case does not settle, and time frames for issuance and
circulation of that report.
The DSU contains two key time frames within which panels must perform their
function.15 First, Article 12.8 provides that the period from the composition of the panel to
the date the panel issues its final report to the parties shall as a general rule not exceed six
months, which can be extended under the first sentence of Article 12.9.16 However, the
second sentence of Article 12.9 creates a second time frame that in no case should the
period from the establishment of the panel to the circulation of the report to the Members
exceed nine months. This is a distinct time frame because it commences and terminates at
different points from the first one. Even if a panel can comply with the six-month time
frame, it will not necessarily comply with the nine-month time frame17 because the
nine-month time frame includes two additional steps, that is, composition of the panel
and translation of the panel report before circulation to Members. Consequently,
Article 12.9 does not grant a three-month extension for the panel to complete its work.
In fact, nowadays, the nine-month time frame implies a reduction in time for the panellists
because the additional steps of panel composition and translation, combined, usually take
more than three months.
Both these time frames are indicative. The time frame in Article 12.8 is ‘a general rule’
and may be extended. The second sentence of Article 12.9 begins: ‘[i]n no case’ but then
uses the auxiliary verb ‘should’, rather than ‘shall’, unlike other time frames in the DSU,18
which appears to indicate that greater efforts should be made to comply with the longer
time frame, without creating an absolute deadline.
DSU practice does not conform to either of these time frames. The six-month time
frame in Article 12.8 has only been respected in five ordinary disputes conducted under the
rules of the DSU.19 The nine-month time frame in Article 12.9 has only been respected in
seven ordinary disputes conducted under the rules of the DSU.20
This is not a recent phenomenon. Even in the first three years of the WTO, only
three out of twenty-one panels composed complied with the six-month time frame and
15
Two other time frames in Arts 20 and 21.4 of the DSU include extensions of time under Art. 12.9. See
below n. 25.
16
The second sentence of Art. 12.8 refers to a three-month time frame in cases of urgency.
17
This occurred in India – Patents (US), WT/DS50; Canada – Pharmaceutical Patents, WT/DS114; and US – Carrier
Bags, WT/DS383.
18
Articles 12.8, 17.5 (1st and 4th sentences), 20, and 21.4 of the DSU (most of which are otherwise qualified).
The special time frames for panels in Arts 4.6 and 7.5 of the SCM Agreement use the word ‘shall’ without qualification.
The special time frame for panels in Art. XXII:6 of the Agreement on Government Procurement uses the words ‘shall
attempt’.
19
US – Wool Shirts, WT/DS33; India – Patents (US), WT/DS50; Canada – Pharmaceutical Patents, above n. 17;
Canada – Patent Term, WT/DS170; and Brazil – Coconut, in which the panel did not reach any conclusion on the
substantive questions because the issue of applicable law was dispositive: WT/DS22/R, s. VII; plus four disputes under the
SCM Agreement, above n. 9, two disputes that were not fully contested: US – Shrimp (Thailand), WT/DS335 (estimated);
and US – Carrier Bags, above n. 17; and possibly Thailand – H-Beams, WT/DS122.
20
US – Underwear, WT/DS24; Canada – Periodicals, WT/DS31; US – Wool Shirts, above n. 19; Argentina –
Footwear, WT/DS56; EC – Poultry, WT/DS69; Canada – Patent Term, above n. 19; and Brazil – Coconut, above n. 19; plus
Brazil – Aircraft, above n. 11; Canada – Aircraft, above n. 11; and US – Shrimp (Thailand), above n. 19.
40
only six out of twenty-three panels established complied with the nine-month time frame.
In the fourth year, 1998, three panels subject to shorter time frames under the Agreement
on Subsidies and Countervailing Measures (SCM Agreement) complied notionally with
one or both of the DSU time frames, but none of the nine ordinary panels complied with
either. Since 2000, the proportion of panels that has complied with either time frame has
been negligible.
The median and average21 times that panels have taken to issue a final report to the
parties have exceeded the six-month time frame in every year by a margin of at least one
third since the inception of the WTO dispute settlement system, as shown in Figure 1.
The rate and margin by which panels have exceeded that six-month time frame show that
it has never been applied as a general rule.
The breadth of the margin by which median and average times have exceeded the six-
month time frame indicates that the firmer nine-month time frame in Article 12.9 was also
unrealistic. This is because the nine month timeframe includes the time from establishment
of the panel before the panel is composed, which requires a minimum of thirty days in the
absence of mutual agreement,22 and it also includes the time after the final report has been
issued to the parties prior to the date of its circulation to Members, for which Appendix 3
of the DSU proposes three weeks. After subtracting these two additional time periods, the
text basically envisages only five additional weeks for the panel to issue its final report
within the nine-month time frame, and less if the panel is not composed promptly. Yet the
median length of panel proceedings without these extra steps has never been less than seven
additional weeks in any year except one, 1996. The longer times taken for panel composi-
tion and translation of reports in later years have only compounded the problem.
The starting point of the nine-month time frame in Article 12.9 of the DSU is
curious, if it is intended to discipline the time that panellists take to complete their work.
The nine months begin to run from the date of establishment of the panel, that is, before
the panellists are appointed. In contrast, the shorter general rule in Article 12.8 and the
shorter special time frames for panels under other WTO agreements all commence from
the point at which the panel is composed and its terms of reference are established or
agreed.23 The time frame in Article 21.5 of the DSU also runs from the time when a
compliance panel can begin work, if the original panellists are all available.24 The timetable
for the panel process is not fixed under Article 12.3 of the DSU until after the panel has
been composed. On the other hand, the time frames in Articles 20 and 21.4 of the DSU
run from the date of establishment of the panel but include later stages as well.25
21
See above n. 11.
22
Twenty days to attempt agreement on the panellists plus ten days for the Director-General to compose the panel
under Art. 8.7 of the DSU.
23
Articles 4.6 and 7.5 of the SCM Agreement; Art. XXII:6 of the Agreement on Government Procurement.
24
Article 21.5 of the DSU also envisages the possibility that the original panellists are no longer available. In the
last five years, five out of fifteen compliance panels required a new composition process taking up to two and a half
months.
25
These time frames include time for adoption of the panel report or notification of appeal, appellate review, and
adoption of reports, if applicable and, in the case of Art. 21.4, determination of the reasonable period of time for
implementation.
41
The time frames in Articles 12.8 and 12.9 of the DSU were based on the GATT
customary practice of over thirty years ago. Both provisions are very similar to provisions of
the 1989 Improvements,26 but the nine-month period can be traced further back to a
description of GATT customary practice in the Annex to the 1979 Understanding.27
GATT panels were established under one agreement only and panel reports in the 1970s
were usually the size of a weekend newspaper supplement.28
As regards compliance panels, the basic time frame in Article 21.5 of the DSU is
ninety days, or one third of the time frame in Article 12.9 of the DSU, but it expressly
allows for extensions. This includes time for re-composition of the panel, if any of the
original panellists are unavailable.29 Compliance panels usually take less time to issue a
report than a panel in an original dispute as they only schedule one meeting with the parties
and similar issues are likely to have arisen in the original proceeding. However, the
compliance panel must still assess the existence or consistency of ‘measures taken to
comply’, including new facts, new arguments, and any new provisions of the covered
agreements that the compliance panel request may invoke.30 The parties still make two
main written submissions and submit responses to questions in writing. Translation of the
report is also included in the Article 21.5 time frame. This all makes the basic ninety-day
time frame unfeasible in most cases.31
Consultation of technical experts and expert review groups is expressly foreseen
during the panel stage, including the compliance panel stage, under Article 13 of the
DSU and certain additional rules in other covered agreements.32 The time frames in
Articles 12.9 and 21.5 of the DSU make no allowance for such procedures, except insofar
as they do not set strict deadlines.
Shorter time frames of ninety and 120 days from panel composition to the circulation
of the panel report apply under special rules and procedures in Parts II and III of the SCM
26
Decision of 12 Apr. 1989 on Improvements to the GATT Dispute Settlement Rules and Procedures, BISD
36S/61, s. F, paras (f)(5) and (6).
27
Understanding on Notification, Consultation, Dispute Settlement and Surveillance of 28 Nov. 1979 (L/4907,
BISD 26S/210), para. 20: ‘The time required by panels will vary with the particular case’. A footnote referred to para.
6(ix) of the Annex setting out Agreed Description of the Customary Practice of the GATT in the Field of Dispute
Settlement (Art. XXIII:2), which stated: ‘Although the CONTRACTING PARTIES have never established precise
deadlines for the different phases of the procedure, probably because the matters submitted to panels differ as to their
complexity and their urgency, in most cases the proceedings of the panels have been completed within a reasonable period of
time, extending from three to nine months’ (emphasis added).
28
See, for example, GATT panel reports in US – DISC (1976) BISD 23S/98 (seventeen pages); EEC – Animal
Feed Proteins (1977) BISD 25S/49 (twenty-two pages); EEC – Minimum Import Prices (1978) BISD 25S/68 (forty-five
pages). The brevity of these reports does not imply any judgment as to their content. Arithmetical comparisons with
recent WTO panel reports must be approached with caution as WTO panel reports can include many annexes submitted
by the parties and responses to questions.
29
In the last five years, five out of fifteen compliance panels have required a new composition process, which took
up to two and a half months.
30
Appellate Body reports in Canada – Aircraft (Article 21.5), WT/DS70/AB/R, paras 40–42; EC – Bed Linen
(Article 21.5), WT/DS141/AB/RW, para. 79.
31
The panels in Australia – Automotive Leather (Article 21.5), WT/DS126, and Canada – Dairy (Article 21.5 – US
and New Zealand) (first recourse), WT/DS103, circulated their respective reports to Members three months after they
were re-composed, but more than three months after the compliance dispute was referred by the DSB.
32
Article 11.2 of the SPS Agreement; Arts 14.2 and 14.3 and Annex 2 of the Agreement on Technical Barriers to
Trade (TBT Agreement); Art. 19.4 of the Customs Valuation Agreement, all listed in Appendix 2 of the DSU.
42
Agreement.33 Four SCM panels accelerated their timetables considerably but found it
unfeasible to circulate a report within ninety days.34 A shorter time frame also applies
under the Agreement on Government Procurement of four months, or seven months in
case of delay, from panel composition to issuance of the final report to the parties.35 One
panel has issued a report under that agreement, outside the longer time frame.36
Comparisons are inevitable between the panel practice of exceeding DSU time frames
and the Appellate Body’s usual respect for its deadline. However, comparisons of these
disparate time frames cannot be taken at face value. The DSU contains two applicable time
frames for appellate proceedings, both in Article 17.5 of the DSU. The first provides that, as
a general rule, the proceedings of the Appellate Body shall not exceed sixty days from the
date that a party formally notifies its decision to appeal to the date the Appellate Body
circulates its report.37 The Appellate Body has complied with this rule only twice in an
original dispute, twice in a compliance dispute, and never since 2000, except after an appeal
was withdrawn.38 Notices of delay from the Appellate Body under Article 17.5 are the
norm. The second time frame provides that: ‘[i]n no case shall’ the proceedings of the
Appellate Body exceed ninety days. This covers the same procedural steps as the sixty-day
rule. The Appellate Body has issued most of its reports within this extended time frame39
although one tenth of Appellate Body reports have been issued later by between two weeks
and fifty days.40
The Appellate Body’s record on time frames, though imperfect, is clearly much
better than that of the panels. The first reason is that the Appellate Body believes that it
has an obligation to respect the ninety-day time frame, as evidenced by the number of
Appellate Body reports issued exactly ninety (or ninety-one) days after the initiation of an
appeal, and the fact that the Appellate Body seeks the consent or agreement of the parties
to exceed the time frame and to ‘deem’ the report to have been circulated under
Article 17.5 of the DSU.41 Article 17.5 of the DSU provides that in no case ‘shall’ an
33
Articles 4.6 and 7.5 of the SCM Agreement.
34
Canada – Aircraft, Brazil – Aircraft, Australia – Automotive Leather, Canada – Aircraft Credits and Guarantees,
above n. 11.
35
Article XXII:6 of the Agreement on Government Procurement.
36
Korea – Government Procurement, WT/DS163.
37
Shorter time frames of thirty days and sixty days apply to certain Appellate Body proceedings under Art 4.9 of
the SCM Agreement.
38
Japan – Alcoholic Beverages, WT/DS8; US – Wool Shirts, WT/DS33; Brazil – Aircraft (Article 21.5), WT/DS46;
Canada – Aircraft (Article 21.5), WT/DS70; as well as India – Motor Vehicles, WT/DS146, in which the appeal was
withdrawn.
39
This includes many reports issued a day later where the time frame expired on a non-WTO working day, in
accordance with Rule 17(2) of the Working Procedures for Appellate Review, WT/AB/WP/6.
40
EC – Hormones, WT/DS26 and WT/DS48 (114 days); Thailand – H-Beams, WT/DS122 (140 days); EC –
Asbestos, WT/DS135 (140 days); US – Lead and Bismuth II, WT/DS138 (104 days); EC – Sugar WT/DS265 (105 days);
US – Upland Cotton, WT/DS267 (136 days); Mexico – Anti-dumping Measures on Rice, WT/DS295 (132 days); US – Upland
Cotton (Article 21.5), WT/DS267 (111 days); US/Canada – Continued Suspension of Obligations in the EC-Harmony Dispute,
WT/DS320 and WT/DS321 (140 days).
41
EC – Asbestos, WT/DS135/AB/R, para. 8; US – Lead and Bismuth II, WT/DS138/AB/R, para. 8; EC – Sugar,
WT/DS265/AB/R, para. 7; US – Upland Cotton, WT/DS267/AB/R, para. 8; Mexico – Anti-dumping Measures on Rice,
WT/DS295/AB/R, para. 7; US – Upland Cotton (Article 21.5), WT/DS267/AB/RW/R, para. 14; US/Canada –
Continued Suspension of Obligations in the EC-Harmony Dispute, WT/DS320/AB/R, para. 29.
43
appellate review proceeding exceed ninety days, unlike Article 12.9 of the DSU, which
uses the word ‘should’.
Second, the time frames in Articles 17.5 and 12.9 do not commence when the
Appellate Body and panels commence work, for opposite reasons. The DSU allows the
Appellate Body to prepare for an appeal before its own proceedings commence, from
the time that the panel report is circulated. Article 16.4 of the DSU allows the parties up to
sixty days to file an appeal, and on average, they take six weeks to do so.42 The Appellate
Body receives the matter digested for it by the panel and can use the time before the appeal
is filed to familiarize itself with the measures at issue and the panel’s findings and to consider
possible grounds of appeal. Upon notification of an appeal, the Working Procedures for
Appellate Review contain no requirement to deliver the panel report with the rest of the
panel record because the Appellate Body already has it.43 In contrast, the time frame for
panels in Article 12.9 of the DSU commences, on average, two months before the panellists
are even appointed. For another month or so, the panellists still have little information
regarding the matter in dispute, only the complainant’s panel request and the parties’
statements in the DSB. The respondent’s statements, for tactical reasons, often reveal
nothing at all of its defences.
Third, the Appellate Body obtains further time to prepare for eventual appeals
through informal procedures. The Appellate Body requests copies of panel reports before
circulation to Members in order to have more time to prepare for an eventual appeal.44
Final panel reports are issued to parties on average about six weeks before they are
circulated to Members, which, if added to the time from circulation of the report to
notification of an appeal, may allow the Appellate Body ninety days to prepare for
an appeal, granting it a total of about six months to comply with a ninety-day
deadline. In addition, certain notices of appeal have been filed more than sixty days after
circulation of the panel report45 while others have been lodged, then withdrawn and then
filed again at a later date to avoid having the ninety-day period end at the core of the
winter vacation.46
The Appellate Body’s time frames under the DSU are much shorter than those for
panels, but this reflects the Appellate Body’s more limited function, which is only to review
issues and interpretations in the panel report; only to review issues of law, not of fact; and
only to review issues raised on appeal.47 The Appellate Body also drafts its reports in the
knowledge that they are not subject to review on a further appeal.
42
Forty-four days on average.
43
Rule 25 of the Working Procedures for Appellate Review, WT/AB/WP/6.
44
The Appellate Body referred to the confidential version of a final panel report issued to the parties in advance of
circulation to Members in Japan – DRAMs (Korea), WT/DS336/AB/R, paras 7 and 279.
45
Chile – Alcoholic Beverages, WT/DS87/AB/R, paras 3 and 4; Korea – Dairy, WT/DS98/AB/R, paras 3 and 4;
Argentina – Footwear, WT/DS121/AB/R, paras 4 and 6; EC – Sugar, WT/DS265/AB/R, paras 4–6; US – Zeroing (EC),
WT/DS294/AB/R, paras 3 and 4; Brazil – Tyres, WT/DS332/AB/R, para. 6.
46
US – FSC, WT/DS108/AB/R, para. 4; US – Line Pipe, WT/DS202/AB/R, para. 13; US – Softwood
Lumber IV, WT/DS257/AB/R, para. 6.
47
Articles 17.6, 17.12, and 17.13 of the DSU.
44
In summary, WTO panel practice does not resemble the time frames for panel
proceedings set out in the DSU and other covered agreements. Delays beyond those time
frames are not necessarily undue. The reasons for these delays are examined further in the
sections below.
4. TIME FOR PARTIES’ SUBMISSIONS
Article 12.2 of the DSU requires panels to strike a balance in their procedures between the
need for ‘high-quality panel reports’ and the need to avoid undue delay in the panel
process. Article 12.3 obliges panels to consult the parties to the dispute before fixing the
timetable for the panel process but not to obtain their consent. However, Article 12.4
expressly states that, in determining the timetable for the panel process, ‘the panel shall
provide sufficient time for the parties to the dispute to prepare their submissions’. When
consulted upon a draft panel timetable, parties are adamant regarding their need for
sufficient time to prepare their own submissions, as guaranteed by Article 12.4.
Many recent panels have cited among the reasons for issuing their reports more than
six months after panel composition the fact that the timetables were agreed after consulta-
tions with the parties or took into account the views of the parties.48 Others have referred
to the need to allow the parties sufficient time to submit evidence and to comment on
evidence submitted by each other,49 requests by the parties to postpone deadlines during
the proceedings,50 requests by the parties for additional time to prepare their rebuttals,51 or
referred to the time frames in Appendix 3 to the DSU.52 Even the two most common
reasons cited for delay, namely ‘scheduling conflicts’ and ‘the scope and complexity of the
dispute’, can partly reflect the time allowed for parties’ submissions.
The DSU contains a proposed timetable in Appendix 3 that indicates the time to be
allowed for each submission and meeting in terms of a range. Parties tend to treat the
maximum time for each step in the proposed timetable in Appendix 3 of the DSU as an
entitlement.53 The point to note is that the maximum times allowed for each step in the
proposed timetable in Appendix 3 of the DSU from composition to issuance of the final
48
US – Oil Country Tubular Goods, WT/DS282/4; US – Zeroing, WT/DS294/9; US – DRAMs, WT/DS296/4;
EC – DRAMs, WT/DS299/5; Mexico – Soft Drinks, WT/DS308/6; Mexico – Steel Plate, WT/DS331/4; Brazil –
Retreaded Tyres, WT/DS332/6; Turkey – Rice, WT/DS334/6; EC – Salmon, WT/DS337/4; Australia – Apples, WT/
DS367/8 and 9.
49
Japan – Laver, WT/DS323/4.
50
US – Poultry, WT/DS392/4.
51
EC – Biotech (GMOs), WT/DS291/25 and 26.
52
Dominican Republic – Cigarettes, WT/DS302/7.
53
There are exceptions, with little or no net time saving. Time frames in item (a) in the proposed timetable for first
written submissions are generally viewed as a mistake as they allow more time for the complainant (three to six weeks)
than for the respondent (two to three weeks) despite the fact that the complainant controls the timing of its request for
establishment of a panel and has time before the panel is established to prepare its submission. Most panels grant less than
the maximum to the complainant and more to the respondent. The time frame in item (i) for an interim review meeting is
also reduced where parties submit only written comments, but more time is granted at item (h) for initial comments at
interim review.
45
report to the parties (items (a) to (j)) add up to thirty-one weeks, which exceed the time
frame of six months (or twenty-six weeks) in Article 12.8 of the DSU.54
Notwithstanding Appendix 3, parties expect additional procedures that have devel-
oped over time and nowadays are standard practice. These comprise one or two weeks for
third parties’ submissions before the first substantive meeting55 plus two weeks to reply to
panel questions after the first substantive meeting, two weeks to reply to panel questions
after the second substantive meeting, and a final opportunity to comment on those
replies.56 Collectively, they add about seven weeks to the timetable. The later steps in
the timetable are not compressed to compensate for the additional steps earlier in the panel
proceeding.
The sum of the maximum time frames in Appendix 3 and the additional procedures
that form part of current WTO practice is about thirty-eight weeks (or nine months) from
composition to issuance of the final report to the parties. These render the general rule of
six months (or twenty-six weeks) in Article 12.8 of the DSU unfeasible. The nine-month
time frame in Article 12.9 of the DSU is also unfeasible because it is counted from
establishment of the panel (including the time before composition of the panel) to the
date of circulation of the report to the Members (including the time for translation after
issuance of the final report to the parties).
The DSU contains provisions that lengthen panels’ timetables beyond nine months.
Article 12.10 of the DSU requires panels to accord sufficient time for developing country
respondents to prepare and present their argumentation. This envisages additional time, or
it would not add anything to Article 12.4.57 Additional time for a developing country
respondent is not compensated by less time for any other party: on the contrary, it creates
additional time for the complainant as well because parties’ second written submissions are
simultaneous, in accordance with Article 12.6.
Respondents at all levels of development mine a rich lode of Appellate Body findings
on due process58 (or procedural fairness) to justify requests for extensions of panel time-
tables. The inequality of arms between multiple complainants and a sole respondent, or
simply the scope of a large matter, are inevitably raised to justify more time under
Article 12.4 of the DSU, and more time than that allowed by the proposed timetable in
Appendix 3. This occurs even where multiple complainants make the same claims, unless
54
Appendix 3 does not state from what date the proposed timetable begins to run. It is sometimes counted from
the date on which the panel fixes its timetable in accordance with Art. 12.3 of the DSU, rather than from composition,
which adds about another week.
55
Article 10.2 of the DSU. Art. 12.4 and Appendix 3, para. 6 of the DSU do not mention the filing of such third
party written submissions.
56
Appendix 3, para. 8 of the DSU refers to panel questions as an optional procedure. Parties expect two weeks to
respond in writing to the panel’s questions after each of the first and second substantive meetings and one week to
comment in writing on each other’s responses after the second meeting.
57
Article 12.10 states that: ‘it does not affect the timeframes in paragraph 1 [sic] of Article 20 and paragraph 4 of
Article 21’.
58
Appellate Body reports in Brazil – Desiccated Coconut, WT/DS22/AB/R, 22, DSR 1997:I, 167, 186; Australia –
Salmon, WT/DS18/AB/R, paras 272 and 278; Chile – Price Band System, WT/DS207/AB/R, para. 176, Mexico – Corn
Syrup (Article 21.5 – US), WT/DS132/AB/RW/R, paras 47 and 107; Thailand – H-Beams, WT/DS122/AB/R, para. 88;
US – Gambling, WT/DS285/AB/R, paras 269–273; Canada – Wheat Exports and Grain Imports, WT/DS276/AB/R,
para. 177.
46
they offer to file joint submissions, because the facts and arguments may be different.
Translation is another example, either of other parties’ submissions into the working
language of the respondent or translation of a party’s own submissions into the working
language of the panel or translation of a panel’s report into another WTO official language
prior to issuance to the parties, all of which increase the time taken from composition to
issuance of the final report to the parties.59
Parties on both sides of a dispute may seek additional time above the maximum time
frames set out in Appendix 3 with good reason. The effectiveness of the WTO dispute
settlement system is related to the quality of the parties’ submissions. Time allowed for a
party’s submission in a panel timetable includes not only actual drafting but also other
internal steps, such as obtaining approvals from different authorities. Where there are
many facts, measures, claims, defences, and provisions in covered agreements in issue,
parties are likely to need more than three weeks to present a comprehensive and coherent
submission.
The increase in the time for parties’ submissions and in the number of filings that
parties make inevitably increases the average time that panels take to draft their findings.
Panels not only have more material to review, but they usually have more issues to address
in their findings. This partly explains why the gradual increase in the length of panel
proceedings, at least to 2004, exceeded the sum of the maximum time frames in Appendix 3
plus the additional procedural steps.
Not all additional time granted to the parties is devoted to improving submissions.
Delegates representing parties to disputes are keenly aware of their national holidays,
summer holidays, and personal and family circumstances when they comment on panels’
proposed timetables and at all times thereafter during a panel proceeding. Although it has
been observed in another connection that the DSU makes no reference to summer
holidays,60 panels usually accommodate requests not to schedule a meeting or submission
on a national holiday of a party to the dispute and may accede to requests for additional
time to prepare submissions around these dates.61 National holidays can last over a week.
Meetings and submissions are not scheduled on WTO official holidays either. These and
other scheduling considerations can collectively add a couple of weeks to a panel
timetable.
59
Mexico – Anti-dumping Measures on Rice, WT/DS295/5; Mexico – Soft Drinks, WT/DS308/7; Mexico – Olive Oil,
WT/DS341/4 and Chile – Price Bands (Article 21.5), WT/DS307/7. This is not requested where the Spanish-speaking
party is complainant, for example, US – Cement (Mexico), WT/DS281/4; US – Oil Country Tubular Goods, WT/DS282/4;
EC – Biotech (Argentina), WT/DS293; US – Oil Country Tubular Goods Sunset Reviews (Article 21.5), WT/DS268/18;
EC – Bananas (Article 21.5 – Ecuador) (II), WT/DS27/85.
60
Award of the Arbitrator on reasonable period of time under Art. 21.3(c) of the DSU in Canada – Pharmaceuticals,
WT/DS114/13, para. 61.
61
The Appellate Body might not be so accommodating. It held its oral hearing in US – Offset Act (Byrd
Amendment) on Thanksgiving Day in 2002: WT/DS217/AB/R, para. 10. In 2004, the Appellate Body proposed to
introduce two annual holiday periods of three weeks each during which its proceedings would be suspended but, after
receiving Members’ comments, decided against it: see proposed amendments, WT/AB/WP/W/8, s. V, and Appellate
Body Annual Report for 2004, WT/AB/3, 8.
47
Only two ordinary original panels established in the last five years took less than nine
months from composition to issue their final report to the parties. Both of them offered less
time for parties’ submissions than contemporaneous panels. The panel in India – Additional
Import Duties (composed in 2007) allowed nineteen weeks up to the end of the second
meeting,62 a total not exceeding the upper time frames in Appendix 363 plus time for third
party submissions and responses to questions after the first meeting. This enabled the panel
to issue its final report to the parties in eight months and three weeks.64 No panel in a fully
contested original dispute has fixed such a briefing and meeting schedule since 2004.65
The panel in Japan – DRAMs (Korea) (composed in 2006) allowed roughly twenty-two
weeks to the second meeting but only three weeks for interim review, less than the five
weeks proposed in Appendix 3, and it was able to issue its final report to the parties in eight
and a half months.66 Two other panels took less time still where the respondent did not
contest the central claims, by dispensing with one round of submissions and one or both
meetings, with the parties’ consent.67
If Members wished to reverse the gradual increase in the length of panel proceedings
in typical cases below the 2004 level, one strategy would be to shorten the time frames for
the parties’ submissions or eliminate certain procedural steps altogether. Parties may find
this unpalatable, in which case a de facto minimum time frame will continue to operate of
over nine months from panel composition to the issuance of the final report to the
parties.
Some complainants may be willing to reduce the time taken overall at the panel stage
in appropriate cases, such as where the measure at issue is of limited duration or there are
few claims. Various procedures are open to reconsideration. For example, an organized
complainant can offer to waive some or all of the time allowed in the timetable to prepare
its first written submission. Panels would probably be more willing to agree to such a
proposal if the DSB confirmed that this time is the complainant’s alone to waive and not
time to which the respondent is also entitled. The second substantive meeting can be
eliminated and, with it, one round of oral statements and one round of questions. Parties
are accustomed to making lengthy filings without a corresponding oral presentation, even
in original disputes, in the form of written responses to questions. A decision to allow both
rounds of submissions before a single meeting would, incidentally, enhance third party
rights as third parties receive all of the submissions made before the first meeting in
accordance with Article 10.3 of the DSU.68 Fewer panel questions (discussed below)
would enable the time for responses to questions to be reduced. Interim review can also
62
Note that parties make four filings after the second meeting, i.e., responses to questions, comments on responses,
requests for interim review, and comments on requests for interim review.
63
Sixteen weeks through item (d), the panel’s second substantive meeting with the parties.
64
Panel Report, WT/DS360/R, paras 1.4–1.8. Another proceeding in a substantially similar matter had already
commenced and the parties had agreed to harmonize the two timetables: see para. 1.9.
65
The panel in Dominican Republic – Cigarettes, WT/DS302, composed in 2004, also allowed nineteen weeks from
composition to the second meeting and issued its report to the parties in eight months and three days.
66
Panel report, WT/DS336/R, paras 1.5, 1.7 and 1.8; Appellate Body report, WT/DS336/AB/R, para. 7.
67
US – Shrimp (Ecuador), WT/DS335; US – Carrier Bags, WT/DS383.
68
Appellate Body report in US – FSC (Article 21.5), WT/DS108/AB/R, para. 245.
48
be accelerated. This stage is required by Article 15.2 of the DSU and useful in practice, but
interim review meetings are not usually held and the review could frequently be conducted
within less than the total of five weeks proposed in Appendix 3. These Modifications of
specific procedures could save up to two months of time in appropriate cases and do not
require amendment of the DSU.
5. UNEXPECTED DELAYS
The gradual increase in panel timetables over the first ten years of the WTO gave way to a
sharp increase among panels composed in 2005 and following years. This steeper increase
has continued to the present time and cannot be attributed solely to the time allowed for
parties’ submissions.
Very long delays began to occur from 2005 in a small number of high-profile disputes
in which the panels each took over two years to issue their final reports to the parties.
The lengths of these delays were unprecedented. During the first ten years of the DSU’s
operation, WTO panels had rarely taken more than a year to issue a report and only three
had ever taken over fifteen months.69 It was fairly exceptional for a panel to issue a second
notice of delay revising its own estimate of the period within which it would complete
its work. However, from 2005, the DSB began receiving cascading estimates of the
further time that certain panels would require to complete their work, which demonstrated
to even a casual observer that there is no upper limit in the DSU on the time that a panel
can take.
The panel in EC – Biotech (GMOs) was composed in 2004 and issued its final report to
the parties two years and two months later, in 2006. This was a dispute under the SPS
Agreement. The panel issued seven notices of delay, citing as reasons the number of claims
and products involved in the dispute, the consultation of scientific experts, the volume of
material on the record, and human resources constraints, and later it cited a lack of facts
regarding the respondent’s measures and a lack of greater coordination among the com-
plainants as well.70 The panel in US/Canada – Continued Suspension of Obligations in the EC
Hormones Dispute was composed in 2005 and issued its report two years and seven months
later, in 2007. This panel also consulted scientific experts on SPS matters and considered
novel issues regarding termination of authorization to suspend concessions under the DSU.
The panel attributed the delay in issuing its report to the complexity of the scientific and
technical issues and difficulties in scheduling the second open hearing with the parties and
the experts, among other factors.
Two panels were also composed in the Airbus/Boeing disputes in 2005. A special
information-gathering process under Annex V of the SCM Agreement continued for the
69
EC – Asbestos, WT/DS135; EC – Tube or Pipe Fittings, WT/DS291 (both over fifteen months); and Mexico –
Telecoms, WT/DS204 (eighteen and one half months including translation prior to issuance).
70
WT/DS291/25, 26, 27, 28, 29, 30, and 31; see also a note from the Chair upon circulation, WT/DS291/32;
and the panel report, WT/DS291/R, paras 7.37–7.46.
49
first four months after the panels were composed, after which the panels set aside their
timetables at the request of the parties for periods of six months or more.71 The panel in
EC and certain Member States – Large Civil Aircraft recommenced its work at the request of
the complainant in September 2006 and issued its report three and a half years later, in
March 2010.72 The panel in US – Large Civil Aircraft (2nd complaint) was composed in
November 2006 and does not expect to complete its work until the first half of 2011, over
four years later.73 Both panels cited, as reasons for the delay, ‘the substantive and procedural
complexities’ of the disputes together with, in the first case, the ‘volume of materials
involved in this dispute’.74
The longer delays have not been confined to these four disputes. The median length
of panel proceedings sharply increased from 2005, as did the average time taken in more
ordinary disputes, as shown in Figure 1. During this period, there was an increase in the
number of unexpected delays, as evidenced by the higher proportion of panels that issued
more than one notice of delay. Initial notices of delay are often circulated around the time
of the panel’s second meeting and can usually take account of the time for parties’
submissions. Second and subsequent notices of delay revise initial estimates and usually
reflect delays that arise when the panel is drafting its findings. Since 2005, half of all panels
in original disputes have issued two, three, or even more such notices.75
All the panels in the longer proceedings in recent years encountered unexpected
delays.76 The actual reasons varied. The panel in US – Zeroing (Japan) was composed in
2005 and issued its report more than fifteen months later. The Chair, Mr David Unter-
halter, who is now a member of the Appellate Body, issued a notice of delay two months
after the panel’s second meeting estimating that the panel would require eleven months
‘inter alia, due to the complexity of the issues and other unavoidable postponements in the
timetable for the work of the Panel’.77 The postponements do not appear to have occurred
during the time for parties’ submissions, at least before the second meeting, as that meeting
was held five months after the panel was composed.78 The Chair later revised his estimate
during interim review due to ‘the complexity of the issues involved’.79 The panel report
shows that the complexity at this later stage was caused by the release of the Appellate Body
report in US – Zeroing (EC), which had a direct bearing on the contents of the panel’s
interim report if not on the panel’s conclusions.80
71
WT/DS316/R, paras 1.7–1.9; WT/DS317/6.
72
WT/DS316/R, paras 1.9 and 1.14.
73
WT/DS353/3 and 7.
74
WT/DS316/9, 10, and 11; WT/DS353/4, 5, 6, and 7.
75
Twenty out of forty-one original panels. Source: Communications from panel Chairs in the WT/DS document
series.
76
All issued revised estimates of delay except EC – DRAMs, in which the Chair resigned, WT/DS299/4; Mexico –
Olive Oil, in which the report was issued to the parties four months later than the Chair’s estimate, WT/DS341/4, WT/
DS341/R, para. 6.3; and Colombia – Ports of Entry, in which the Chair’s estimate was ‘early in 2009’ and the report was
issued to the parties in April, WT/DS366/8, WT/DS366/R, para. 1.8.
77
WT/DS322/10.
78
79
WT/DS322/11.
80
WT/DS322/R, para. 6.2.
50
The panel in Brazil – Retreaded Tyres, composed in 2006, took over thirteen months to
issue its final report to the parties. The Chair scheduled nine months but later announced a
four-month delay without stating any reason.81 Six months elapsed between the panel’s last
meeting with the parties and the issuance of the interim report.82 The panel in China –
Auto Parts, composed in 2007, issued its reports to the parties nineteen months after
composition. The Chair attributed the delays to the complexity of the issues.83 The panel
in Mexico – Olive Oil, also composed in 2007, took over eighteen months to issue its final
report to the parties. The Chair had estimated that the panel would need fourteen months,
citing ‘the complexity of issues presented in this case as well as the need to translate the
parties’ submissions’.84 The translation of parties’ submissions would normally cause delays
earlier in the proceedings, and indeed, the report shows that the panel’s second meeting
with the parties was held over seven months after composition. However, over ten more
months elapsed after that meeting before the panel issued its final report, in English and
Spanish. The Chair did not inform the DSB of the last four months of delay, nor revise her
original estimated date of completion.85
Only three panels were composed in original disputes in 2008 and all of them took
over a year to issue their final reports to the parties. One was the panel in China –
Audiovisual Services, which took fifteen months to issue its final report. The Chair estimated
at the time of the second meeting that the panel would require eleven months ‘mainly due
to scheduling difficulties’. He later revised that estimate twice by a total of four months due
to the need ‘to seek out the services of an outside expert to address a number of translation
issues’, ‘the complexity of the matter and scheduling problems’, ‘the length of the report’,
and sufficient time for the parties at interim review.86 Another was the panel in Colombia –
Ports of Entry, which took fourteen months to issue its final report to the parties. The Chair
cited ‘scheduling difficulties and translation issues’.87 The other panel was in Australia –
Apples, an SPS dispute in which the panel consulted scientific experts and issued its final
report to the parties over two years after it was composed.88
More recently, the panel in EC – IT Products, composed in 2009, issued its final
reports to the parties eighteen months after it was composed. The Chair cited the number
of complicated issues and multiplicity of complainants.89 The panel in Thailand – Cigarettes
(Philippines), also composed in 2009, has not yet circulated its final report, but the Chair has
estimated that the panel will issue its report sixteen months after it was composed, due to
81
WT/DS332/6. The second notice simply stated that ‘it will not be possible for the Panel’ to comply with its
original schedule: see WT/DS332/7.
82
83
WT/DS339/10 and 11.
84
WT/DS341/4.
85
WT/DS341/R, paras 1.4, 1.6, and 6.3. The report does not indicate when the panel issued its interim report.
86
WT/DS363/8 and 9.
87
WT/DS366/8. The panel report also drew attention to the fact that certain relevant provisions in the covered
agreements had not been interpreted before by a panel or the Appellate Body, which the panel felt made its task ‘arduous’:
WT/DS366/R, para. 7.388; see also para. 7.131.
88
WT/DS367/R, para. 1.54, which refers to the interim report issued earlier.
89
WT/DS375/10, 11, and 12.
51
‘the procedural delays caused by the administrative matters involved and the complexity of
the dispute’.90
The delays in issuing these reports do not correlate with an upswing in the number of
panels composed. The number of panels composed in a calendar year, in both original and
compliance disputes, has been less than in 2004 in every subsequent year.91 The number of
panels composed in a calendar year peaked in 1999.
6. SCHEDULING CONFLICTS
One of the most common reasons given for delay in issuance of panel reports is ‘scheduling
conflicts’ or ‘postponements’ in panel timetables. This can cover not only the time for
parties’ submissions but also panellists’ own scheduling constraints, although it is not always
clear in a given case which type of scheduling conflict has been a reason for delay.92
Panellists must dedicate considerable time to a dispute, including blocks of a full week for
each meeting, comprising meetings with parties, internal meetings, and for most panellists,
travel to and from Geneva. A panel report must be approved by all the panellists.
Procedural issues may arise at any time, not just during meetings, and often require
consultation and decisions in very short time frames.
Panellists work on WTO disputes on a part-time basis. They usually have other
full-time jobs and other commitments from which government officials must often take
leave to serve on a panel. Panellists are appointed on the basis of the criteria set out in
Article 8 of the DSU, which include independence, diversity, wide experience, and
nationality and, to the extent possible, the criteria stated by the parties to the dispute.
There is also a negotiating proposal that expects more specific expertise in panellists.93
However, availability of time for sustained examination of submissions and draft findings is
not mentioned.
Nominations for panellists are accompanied by a copy of their respective curricula
vitae, which are intended to indicate satisfaction of the DSU criteria, but they do not
necessarily indicate nominees’ other commitments from time to time. It has been observed
in the DSU negotiations that parties may not have any information about a nominee besides
what appears on that curriculum vitae.94 The Secretariat may not always either. Given that
the DSU recognizes that prompt settlement of disputes is essential to the effective function-
ing of the WTO95 and that many panel reports are delayed due to scheduling conflicts,
sufficient availability to work on a dispute should be a criterion in selecting panellists.
The projected time frame of the panel proceeding should be raised with all panellists before
they accept appointment. Panellists’ other commitments create scheduling conflicts that may
90
WT/DS371/6 and 5.
91
Sources: above n. 8.
92
Notices of delay sometimes expressly cite panellists’ other constraints, e.g., WT/DS114/8; WT/DS291/28.
93
TN/DS/W/89 (Chile and the United States).
94
TN/DS/W/61 (Indonesia and Thailand).
95
DSU, Art. 3.3.
52
be accommodated, with difficulty, in fixing an original timetable and in dealing with

procedural issues that arise during a proceeding. However, they risk adding weeks or more
to a timetable if any date in the timetable should change for any other reason and new dates
on which all three panellists and the parties are available must be found.
As an illustration, one can consider the longest panel proceeding in the WTO to date,
US – Large Civil Aircraft (2nd complaint) (in progress over four years and still counting).96
The panel’s terms of reference include twenty-eight alleged subsidy programmes granted by
at least three US States, NASA, three US government departments, and US federal tax
legislation and involve allegations of many different forms of adverse effects in three
separate aircraft product markets.97 This is a very large matter. When the panel was
composed by Deputy Director-General Alejandro Jara in November 2006,98 the person
that he appointed to chair the panel was already chairing the Doha Round agriculture
negotiations.99 It is difficult to imagine how one person could discharge both these
responsibilities, as well as those of an Ambassador of a Member to the WTO, expeditiously
and simultaneously.100 Successive revisions of draft modalities in the agriculture negotia-
tions were issued in July 2006, August 2007, and February, May, July, and December
2008.101 The panel held meetings with the parties in September 2007 and January 2008.102
The Chair of the panel ceased to chair the agriculture negotiations in early 2009103 and he
estimates that the panel will complete its work in the first half of 2011.104
A strategy for minimizing scheduling conflicts would be for the parties and the
Secretariat to give greater weight to proposed nominees’ actual availability. Parties to a
dispute who are anxious to avoid undue delays should consider nominees’ current and
projected responsibilities and can cite it as a compelling reason to object to a nominee.
During panel composition, parties to a dispute can ask the Secretariat to provide details of a
nominee’s current workload and seek confirmation that he or she is not, for example,
serving on another panel, representing a Member in another dispute before a panel or the
Appellate Body, chairing a body in the WTO or another international organization, or
engaged in detailed bilateral negotiations.
The ultimate scheduling conflict arises when a panellist needs to be replaced, which
happens from time to time. Ms Luz Elena de la Torre resigned as Chair of the panel in
EC – DRAMs.105 Ms Lilia Bautista joined the Appellate Body during US – Continued
96
Above n. 75.
97
WT/DS317/5, WT/DS353/2, and first written submission of the European Communities, <http://trade.
ec.europa.eu/doclib/docs/2007/september/tradoc_136101.pdf>, 16 Sep. 2010.
98
WT/DS353/3.
99
The then-Chair of the Committee on Agriculture in Special Session: see TN/AG/R/21, para. 3.
100
This does not imply that there may not have been other reasons for delay in this proceeding.
101
TN/AG/W/3, W/4, W/4/Rev.1, Rev.2, Rev.3, Rev.4, W/5, W/6, and W/7.
102
European Commission External Trade Press Information Service, ‘Update September 2007’, <http://trade.
ec.europa.eu/doclib/docs/2007/september/tradoc_136049.pdf> and <http://trade.ec.europa.eu/doclib/docs/2008/
january/tradoc_137417.pdf>, 16 Sep. 2010.
103
Minutes of Meeting of the Special Session of the Committee on Agriculture of 22 Apr. 2009, TN/AG/R/22,
para. 3.
104
WT/DS353/7.
105
WT/DS299/4.
53
Zeroing.106 Mr Harsha Singh accepted appointment as a WTO Deputy Director-General

during US – Softwood Lumber V (Article 21.5).107 Panellists have also passed away during
panel proceedings.108
Secretariat staff members who assist panellists are mostly employed on a full-time basis,
but they, too, have other responsibilities besides dispute settlement. The panel in EC –
Biotech (GMOs) expressly cited human resources limitations as one of the reasons for its
delay,109 but ‘scheduling conflicts’ in other cases might indicate that staff have insufficient
time. A complementary strategy to avoid scheduling conflicts is to ensure that sufficient
Secretariat staff members are assigned to assist a panel and that they are not assigned to
other urgent tasks, including other disputes, during key periods. As an illustration, one
can consider EC and certain Member States – Large Civil Aircraft, the vast scope of which
is described below. Up until the time shortly before its first meeting with the parties,
that panel was assisted by only two legal officers and a panel secretary, all of whom
were assigned to another panel or WTO committee or other WTO responsibility at the
same time.
The allocation of more Secretariat staff to dispute settlement is simply logical in light
of the fact that the prompt settlement of disputes is essential to the effective functioning of
the WTO and all the more appropriate in light of the disproportionately large contribution
that the dispute settlement mechanism actually makes to the success of the WTO.
7. SCOPE AND COMPLEXITY OF DISPUTES
The other common group of reasons given for delay in issuance of panel reports is the
‘scope’ or ‘complexity’ of the issues before the panel, or both. The scope of a matter
depends on the number of parties, the number of measures at issue, the volume of
evidence, the number of claims, the number of defences, and the number of issues that
they raise, both procedural and substantive. Complexity depends, among other things, on
the number of issues of first impression and the degree to which the parties’ respective cases
are finely balanced.
There is a wide variety in the scope of the matters that come before the DSB.
A dispute can be as small as Canada – Patent Term, which involved only one measure and
one primary claim. The panel reported in 2000, complying with both the six-month time
frame in Article 12.8 of the DSU and the nine-month time frame in Article 12.9.110 It was
the last panel to do so in an ordinary dispute, due to the small scope of the matter.111
106
WT/DS350/10.
107
WT/DS264/24; WT/DS264/RW, para. 1.5.
108
Korea – Commercial Vessels, WT/DS273/7; and US – Tuna II (Mexico), WT/DS381/7.
109
WT/DS291/28, 30, and 32.
110
WT/DS170/R.
111
The panel in Canada – Aircraft Credits and Guarantees issued its report to the parties in 2002 less than six months
after its composition but was subject to a shorter time frame in the SCM Agreement. The panels in US – Shrimp (Ecuador)
and US – Carrier Bags issued their reports to the parties in less than six months, but the respondent did not fully contest the
claims. See above n. 9.
54
At the other end of the spectrum lies EC and certain Member States – Large Civil Aircraft.
The panel’s terms of reference included launch aid provided by four countries and the EC
to several Airbus-related companies for up to eight models of aircraft, which were
challenged both as actionable subsidies and, in some cases, prohibited subsidies. However,
the dispute also concerns the provision of infrastructure and grants for the construction of
infrastructure by various cities and regions, as well as bank loans, capital restructurings
including debt forgiveness, and equity transfers, and several research and development
funding programmes. With respect to each of these measures, the panel was called upon
to conduct a separate examination to determine whether there was a financial contribution
by government, whether it conferred a benefit, and in some cases whether it was specific.
Having thus identified the subsidies, the panel then analysed whether they caused serious
prejudice to Boeing through the displacement of imports into the EU market, exports into
third country markets and lost sales in the same market, or through price undercutting,
price suppression, or price depression. The panel was also faced with preliminary issues
regarding the applicable agreement between the parties regarding civil aircraft, the temporal
scope of the SCM Agreement, and whether certain measures were within the panel’s terms
of reference. The panel also dealt with a significant number of requests from the parties
concerning the treatment of two categories of business confidential information in the
submissions and exhibits.112 Examination of this matter was a very large exercise indeed,
quite beyond the scope of the usual DSU time frames.
At the outset, the scope and complexity of a dispute are in the hands of the parties.
Complainants may deliberately expand the scope through the inclusion of weak, conse-
quential, and speculative claims in their panel requests. Some alternative claims are included
to anticipate potential defences to the primary claims for sound tactical reasons; others are
apparently included to create more work for the respondent or to impress domestic
stakeholders. Additional claims expand the scope of the matter before the panel and may
create delay in issuance of an interim report. Respondents, for their part, may gratefully
note the number of unsuccessful claims at the time of adoption of the panel report or in
remarks to the press. They can also expand the scope of matters by running unfounded but
time-consuming defences, raising many procedural issues, and engaging in circumvention
tactics.113
Complainants are entitled to decide how many measures and claims they raise in a
dispute. The alleged violation may only arise from the combined effect of a large number of
measures and the desired remedy may only be available where several claims are upheld.
However, in large matters there is a trade-off between the scope of the complaint and the
time within which a panel will deliver its report and, hence, make a remedy possible under
the DSU. Complainants may wish to consider the radical step of pursuing fewer claims in
112
WT/DS316/R.
113
Panels have not cited the participation of private lawyers as a reason for delay and the participation of private
lawyers is not usually disclosed in panel reports. This article therefore does not analyse the impact of that factor, if any, on
the scope of matters before panels.
55
exchange for quicker panel reports. Industry and other domestic stakeholders can question
the impact of the inclusion of many additional and minor claims on the timing of the panel
report. Complainants might be more willing to reduce the scope of a dispute if they were
confident that panels would not propose a standard-length timetable regardless of the
number of claims.
Panels themselves contribute to the scope of the matter as it develops through their
written questions. Astute panellists can analyse which are the key issues sufficient to dispose
of a case at an early stage and pose questions accordingly. This requires proper considera-
tion of parties’ submissions before the end of the meeting schedule and is most useful when
the claims or defences rely on multiple premises. It would be an error to pose too few
questions only to be unclear as to how to proceed when it comes time to draft certain
findings. However, panels pose an increasing number of questions and some fall into the
trap of covering every issue in equal detail, posing too many questions on too many issues,
generating large numbers of responses and also larger rebuttal submissions and comments
on responses, as the parties take their cue from the panel’s questions to some extent in the
issues that they address later.
The panel report in EC – Biotech (GMOs) illustrates the boomerang effect of panel
questions. This was a large dispute from the outset, but the panel also posed 201 questions
to the parties and 114 questions to six scientific experts.114 Many of these questions
generated multiple responses as well as arguments in rebuttal. The panel does not appear
to have foreseen the quantity of material that its large number of questions would inevitably
produce. Even after the panel had posed most of its questions to the experts the Chair still
estimated that the panel could complete the case within a further eight months or fifteen
months from the date of composition.115 He revised that estimate after the panel had
received responses, met the experts, and held its second meeting with the parties.116 When
the report was eventually issued twenty-six months from the date of its composition, the
panel partly blamed the lengthy delay on the volume of material that its own questions had
generated.117
A strategy, if not a solution, for a panel to deal with a large matter is to balance the
need for a high-quality report against the risk of undue delay before it poses questions. Panels
should consider the likely volume of material that they will have to review in the responses,
particularly in view of the common practice of posing many questions to more than one
party, and the signal that the questions may unwittingly send about what the panel
considers important. Panels can address judicial economy overtly early on to narrow down
the volume of material and not just relegate it to the interim review stage. Panels should not
only eliminate redundant or duplicative questions but also consider the net value of each
question to the resolution of the dispute and the likely impact of the total number of
114
EC – Biotech (GMOs), WT/DS291/R, paras 7.14 and 7.39.
115
Notice dated 2 Nov. 2004, WT/DS291/27; cf. WT/DS291/R, para. 7.29.
116
Notice dated 13 Jun. 2005, WT/DS291/28; cf. WT/DS291/R, para. 7.29.
117
WT/DS291/R, para. 7.39.
56
responses on the panel’s timetable. A complementary strategy is to ensure effective docu-

ment management, as a panel may need to review thousands of pages of submissions and
hundreds of exhibits and retrieve them at short notice during the drafting of questions and
findings.
8. LENGTH OF REPORTS
The scope and complexity of the issues that delay panel proceedings often mean, in more
concrete terms, the length of the panel report, in particular the findings section. Panel
reports with longer findings sections generally take longer to draft. The length of the
findings is partly a product of the amount of explanation that a panel feels it needs to give.
Article 12.7 of the DSU provides that ‘the report of a panel shall set out the findings of fact,
the applicability of relevant provisions and the basic rationale behind any findings and
recommendations that it makes’ (emphasis added). The Appellate Body in Mexico – Corn
Syrup (Article 21.5) interpreted the ‘basic rationale’ to mean explanations and reasons
sufficient to disclose the essential, or fundamental, justification for those findings and
recommendations. It clarified that this must be judged on a case-by-case basis and also
explained that this ‘does not, however, necessarily imply that Article 12.7 requires panels to
expound at length on the reasons for their findings and recommendations’.118
Few achieve brevity. Panels do sometimes dismiss a highly contested key issue in short
order where they are confident that their reasoning is solid enough to commend itself over
the alternatives. For example, in US – Upland Cotton, a threshold issue concerned the
panel’s ability to make any findings at all with respect to subsidies whose legislative basis had
expired but the effects of which continued. The panel informed the parties of its ruling at a
preliminary stage and addressed the issue in three pages of the final report, including parties’
arguments. The Appellate Body acknowledged that the panel’s findings on this issue ‘may
be brief’ but dismissed an appeal under Article 12.7 of the DSU.119 Similarly, in EC –
Geographical Indications, one of the two main bones of contention was whether Article 24.5
of the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS
Agreement) authorized the coexistence of geographical indications and prior trademarks.
The panel disposed of this issue in five pages, including parties’ arguments, without appeal
or challenge on interim review.120 However, these examples are not typical, as shown by
the length of the other findings in the same panel reports.
The most notorious example of lengthy findings may be those in EC – Biotech
(GMOs). The first section of the panel’s findings included a detailed discussion of the
relevance of other rules of international law in which the panel offered its interpretation of
Article 31.3(c) of the Vienna Convention of the Law of Treaties, surveyed other
118
Appellate Body report in Mexico – Corn Syrup (Article 21.5), WT/DS132/AB/R, paras 106, 108, and 109.
119
Panel report in US – Upland Cotton, WT/DS267/R, paras 7.104–7.122; Appellate Body report, WT/DS267/
AB/R, para. 277; see also para. 458 regarding the panel’s serious prejudice analysis.
120
EC – Geographical Indications (US), WT/DS174/R; EC – Geographical Indications (Australia), WT/DS290/R,
paras 7.604–7.625.
57
conventions, and reviewed decisions of other international tribunals, only to add that the
respondent’s interpretation was unreasonable in any case and to reach non-conclusions that
the panel did not need to take a position or that it was not necessary to rely on other treaty
provisions.121 The panel report included a commendably brief factual description of the
measures at issue122 but almost 300 paragraphs (seventy pages) on the question whether
three EC approval procedures were SPS measures, a point on which the parties actually
agreed in part.123 The panel then devoted 850 paragraphs (189 pages) to its examination of
whether the first measure at issue (a moratorium on product approvals) had ever existed,124
before finding in a repetitive analysis that most claims regarding that measure failed because
it was not an SPS measure.125 The foregoing comprised almost half of the panel’s findings.
The panel eventually reached the product-specific measures at paragraph 1628 (out of a
total of 3,430) of the findings section. One of the panel’s notices of delay referred to the
complexity ‘of the Panel reports’ (rather than ‘of the dispute’) without apparent irony.126
Yet the complexity of the reports was partly due to choices made as to how to conduct the
panel’s assessment.
An obvious strategy to reduce the time spent drafting findings is to write shorter
reports. Unfortunately, parsimonious findings may not always be adequate to satisfy the
DSU as it is applied. The Appellate Body has explained that the purposes of the require-
ment to provide the ‘basic rationale’ are, first, that a respondent Member is entitled to
know the reasons why it has been found to have acted inconsistently with its obligations
under the covered agreements as a matter of due process and, second, to assist such a
respondent in understanding the nature of its obligations and to make informed decisions
about: (i) what must be done in order to implement the eventual rulings and recommenda-
tions made by the DSB; and (ii) whether and what to appeal.127 Each of these purposes has
separate implications beyond what is strictly the panel’s ‘basic rationale’.
First, a respondent is entitled to know the reasons why its measures have been found
inconsistent with the WTO agreements. This means, in practice, that respondents expect
all of their arguments to be addressed where the panel reaches an adverse conclusion.
Likewise, complainants expect panels to address all of their significant arguments with
respect to any claim on which they are unsuccessful. It is settled law that panels are not
obliged to address every argument raised by a party,128 but panels are loath to ignore any
significant argument that, if accepted, would alter the result. The Appellate Body is not. As
an illustration, one can consider US – Zeroing (EC), in which the Appellate Body found
121
WT/DS291/R, paras 7.49–7.96, note paras 7.71, 7.89, 7.95, 290–292.
122
Ibid., paras 7.106–7.146, 345–354.
123
Ibid., paras 7.147–7.437, 354–427.
124
Ibid., paras 7.438–7.1285, 427–616.
125
Ibid., paras 7.1326–7.1465, 625–657, paras 7.1605–7.1617, 687–689. Certain claims under Art. 8 and Annex C
of the SPS Agreement were successful: paras 7.1466–7.1570.
126
WT/DS291/31.
127
Appellate Body report in Mexico – Corn Syrup (Article 21.5), WT/DS132/AB/R, para. 107.
128
See, for example, the Appellate Body reports in US – Lead and Bismuth II, para. 71; Canada – Autos, paras
112–114.
58
against the respondent.129 The respondent, the United States, criticized the Appellate
Body for failing to address arguments that it had raised regarding the use of ‘zeroing’ in
the post-investigation phases of an anti-dumping proceeding, as well as arguments
regarding the mandatory/discretionary distinction. Even though the Appellate Body is
not, strictly speaking, subject to Article 12.7 of the DSU, the United States was left
without a full explanation of why the Appellate Body found its measures inconsistent
with the Anti-dumping Agreement.130 The United States raised the same objection at the
time of adoption of the Appellate Body report in US – Zeroing (Japan).131 At the panel
stage, unsuccessful parties take a similar approach to the treatment of their significant
arguments.
Second, the findings assist the respondent in implementation. Overly narrow reason-
ing or careless drafting may speed up issuance of a report only to doom the parties to a
compliance dispute. This can also occur on appeal. In US – Gambling, the United States
understood that the Appellate Body report suggested that it could comply with the rulings
by more fully arguing its original defence, an approach that the compliance panel
rejected.132 With the benefit of additional drafting time, the findings in the Appellate
Body report might have disabused the respondent of that notion.
Third, complainants and respondents both refer to a panel’s reasoning (or lack of it) in
deciding whether and what to appeal.133 Appellate review did not exist when the practice
of setting out the ‘basic rationale’ behind findings and conclusions developed.134 Panels are
often at pains to render rulings that are genuinely persuasive, if not in the subjective views
of the parties then at least upon objective review by the Appellate Body. Persuasive
reasoning often requires considerable exposition of the reasons that support a conclusion,
in addition to the dismissal seriatim of the arguments of the unsuccessful party on a particular
point.
These considerations all create an upward pressure on the length of findings in a panel
report, which implies additional time for drafting prior to issuance to the parties. Judicial
economy is an important technique to reduce the scope of a large matter that threatens to
swamp a panel.135 Naturally, it is false economy to save time by shortening a report without
fulfilling the panel’s mandate of contributing to a positive resolution of the dispute.136
Panels from time to time also make additional findings of fact, where these do not unduly
129
WT/DS294/AB/R.
130
See minutes of the meeting of the DSB, WT/DSB/M/211, paras 37–40, and a communication from the
United States dated 9 May 2006, WT/DS294/16, paras 16–17 and 27–28.
131
See minutes of the meeting of the DSB, WT/DSB/M/225, para. 74.
132
See, for example, the Appellate Body report in US – Gambling, para. 371. See also the panel report in US –
Gambling (Article 21.5), WT/DS285/RW, para. 6.40.
133
Appellate Body report in Mexico – Corn Syrup (Article 21.5), WT/DS132/AB/R, para. 107.
134
Article 12.7 of the DSU repeats almost verbatim the terms of the Annex on Customary Practice in the 1979
Understanding, above n. 27, para. 6(v).
135
The panel in US – Upland Cotton chose not to resolve certain complex issues that would have considerably
delayed its report. Both parties appealed this exercise of discretion, unsuccessfully. See the Appellate Body report,
WT/DS267/AB/R, paras 717–718 and 731–732.
136
Article 3.7 of the DSU; Appellate Body report in Australia – Salmon, WT/DS18/AB/R, para. 223.
59
delay the report, lest their legal reasoning be changed on appeal and the Appellate Body
chooses to complete the analysis.137
A solution to lengthier reports is for some brave panels to make shorter findings on
complex issues, setting out their essential reasons without labouring the point in the
interests of avoiding undue delay, and to exercise judicial economy where appropriate.
The fate of such findings on appeal is not assured. For example, the Appellate Body in
China – Audiovisual Services disapproved the panel’s decision to avoid a complex issue that
was unnecessary to reach a result, despite the facts that neither party asked the panel to
address the issue and that the report was already very lengthy.138
Additional time alone confers no advantage where the panel makes a fundamental error
at the outset or its reasoning and conclusions are otherwise very badly flawed. For example,
the panel in US/Canada – Continued Suspension of Obligations in the EC-Hormones Dispute
began part of its analysis by finding that two Members withholding concessions pursuant to
DSB authorization were actively seeking redress of a violation without recourse to Dispute
settlement on the basis of statements that they had made to the DSB.139 Errors such as these
cannot be addressed through timetables and working procedures.
9. PROCEDURAL ISSUES
Procedural issues expand the scope and complexity of the issues that panels must address as
well as lengthen reports. Procedural requests are not new. However, parties are now
engaging some panels in correspondence on an array of issues that concern the conduct
of the proceedings, such as the timing of requests for preliminary rulings, enhanced third
party rights, the bracketing of business confidential information, objections to the untimely
submission of factual evidence, unsolicited submissions by non-parties, alterations to the
panel timetable, consultation of experts and open hearings,140 as well as procedural issues
that affect the substance of the matter in dispute, typically the panel’s terms of reference.
The correspondence not only concerns requests to adopt or apply these procedures but also
the relevant modalities, in particular, which parties and third parties may comment on what
request and when. These matters distract panels from the substantive issues.
The recent panel report in Australia – Apples illustrates the phenomenon. That panel
faced a volley of procedural issues throughout the dispute, beginning with a request for a
preliminary ruling before the first written submissions, revisions of the timetable, a second
request for a preliminary ruling before the first substantive meeting, an unsolicited letter
137
The panel in Argentina – Peaches made additional findings for this express reason, see WT/DS238/R, paras
7.135–7.139. The Appellate Body regretted that the panel in US – Softwood Lumber IV had not done so, WT/DS257/AB/
R, para. 118. The Appellate Body expressly approved the discretion of a panel to make such additional findings in US –
Gambling, WT/DS285/AB/R, para. 344.
138
Appellate Body report in China – Audiovisual Services, WT/DS363/AB/R, paras 213–215.
139
Panel report in US – Continued Suspension, WT/DS320/R, paras 7.205–7.209 and 7.222–7.226.
140
The panel in US/Canada – Continued Suspension of Obligations in the EC-Harmony Dispute cited the difficulties in
scheduling the second open hearing of the panel with the parties and the experts consulted by the panel as one of the reasons
for delay: WT/DS320/10 and WT/DS321/10. As regards the modalities of open hearings, see Lothar Ehring, ‘Public Access
to Dispute Settlement Hearings in the World Trade Organization’, J. Int. Economic Law 11, no. 4 (2008): 1021–1034.
60
from a private party, the opening of the panel’s first substantive meeting with the parties
and third parties to public viewing, a settlement of certain claims, the selection of experts
and preparation of questions, alleged due process concerns, additional revisions of the panel
timetable due to the consultation of experts, and the opening of the panel’s second
substantive meeting with the parties and its meeting with the experts to public viewing.141
The panel report in Australia – Apples also illustrates the volume of correspondence
and material that a single procedural issue can generate. For example, the first request for a
preliminary ruling involved consideration not only of the panel’s terms of reference but also
of the procedures that applied to the making of such a ruling. The respondent requested the
preliminary ruling before the organizational meeting, both parties then wrote to the panel
(at its request) with their preliminary reactions, including on the issue of timing, both
parties made written submissions on the terms of reference, the complainant submitted an
unsolicited letter, a third party then requested an opportunity to comment, the panel
proposed a procedure for third party comments, the panel consulted the parties on this
third party procedure, two third parties submitted comments, both principal parties
commented on the third party comments, the panel issued the preliminary ruling, and
the panel also consulted the parties on whether to circulate the preliminary ruling.142 This
process took just under three months. More time dedicated to procedural issues entails less
time to address the substantive issues.
One strategy to manage the time devoted to procedural issues might be a set of
procedural rules.143 There are certain issues that recur, such as whether or not third parties
should have the right to comment on additional working procedures144 and, if they should,
how much time they should have for comment. It is not clear that Members would be able
to agree on a set of rules, but to the extent that they could devise or approve even a few
standard procedural rules, this could create a presumption in certain situations, which
would allow panels to take less time to deliberate. The difficulty is that there can be few
hard and fast rules so that any procedures must make allowance for extraordinary situations.
Members would still have the right to submit their views, which may do little to stem the
tide of correspondence between parties and panels. The Secretariat can collate the practice
from previous disputes, in which circumstances may have been different, and inform
panellists accordingly. However, when panels deal with procedural issues there is no
substitute for a nimble Chair, especially one who has arbitration or litigation experience.
There is already a procedural rule concerning the timeliness of submission of evidence.
Standard panel working procedures require each party to submit all factual evidence no
later than the first substantive meeting. This is subject to a major exception for evidence
necessary for purposes of rebuttals or answers to questions and a discretionary exception for
141
Panel report in Australia – Apples, WT/DS367/R, paras 1.7–1.52.
142
Ibid., paras 1.7–1.10, Annex A-2 and WT/DS367/7.
143
Jacques Bourgeois, ‘The Umpire Needs Better Rules of the Game’, in The WTO at Ten: The Contribution of the
Dispute Settlement System, ed. G. Sacerdoti et al. above n.6, 235–245.
144
See the European Communities and its Member States’ contribution to the DSU negotiations, 20 Jan. 2003,
TN/DS/W/38.
61
good cause. Parties sometimes submit lengthy, complex factual evidence after the first
substantive meeting that the other party alleges is not ‘necessary’ for the purpose of rebuttal.
Panels tend to admit the new factual evidence and grant the other party a period of time for
comment.145 The other party, in its comments, might then submit new factual evidence of
its own in a competition to have the last word. The procedural rule is not always effective
and generates issues of its own.
Evidentiary issues also arise. A complainant in a contingent trade remedy case may
object to evidence that includes information that was not before the investigating authority
at the time it made its determination. A panel may be required to examine in detail what
information was available at what point in time to determine the admissibility of this
evidence under specific provisions in the covered agreements.146 In other situations, panels
have handled questions akin to admissibility on the basis of general considerations of
relevance or reliability in the course of weighing the evidence as part of their objective
assessment under Article 11 of the DSU.147 The WTO dispute settlement system manages
without comprehensive rules of evidence such as exist in common law jurisdictions. It is
doubtful that such a set of formal rules would improve on the current situation or lessen
demands on panels’ time, especially if the rules were riddled with exceptions. If there is a
need to improve clarity and predictability as regards standards for evidence, it might be
more productive for Members in the DSB to discuss directly the specific issues that recur.
Preliminary rulings on a fundamental question, such as the panel’s jurisdiction, can be
issued before the parties file full submissions. This can save time if the rulings identify major
flaws in the terms of reference148 or where they confirm that alleged flaws do not exist so
that a respondent knows it must address certain measures and claims.149 Requests for the
rulings to be issued at an early stage have increased.150 However, it is likely that a round of
initial submissions on a preliminary ruling will create delay regarding issues that can be dealt
with in the panel report, at the end. The Secretariat can also collate practice regarding
preliminary ruling procedures in previous disputes.
A multi-stage procedure was adopted by the panel in US – Upland Cotton to reduce
the scope of the issues under consideration at any given time to manageable proportions.
That panel observed that most of the claims were subject to a raft of preliminary conditions,
namely non-compliance with the different aspects of the ‘peace clause’ in Article 13 of the
Agreement on Agriculture (that has now expired). The panel addressed the peace clause at
an early stage that enabled it to manage the conduct of its proceedings in an orderly
145
See, for example, Argentina – Textiles and Apparel, WT/DS56/R, para. 6.55, upheld on appeal in WT/DS56/
AB/R, paras 77–81.
146
See, for example, the panel report in EC – Salmon from Norway, WT/DS337/R, paras 7.385–7.860, regarding
Art. 17.5(ii) of the Anti-dumping Agreement.
147
See, for example, EC – Geographical Indications (Australia), WT/DS290/R, paras 7.83–7.84; China – Intellectual
Property Rights, WT/DS362/R, paras 7.627 and 7.629.
148
Preliminary ruling by the Panel in Canada – Wheat Exports and Grain Imports, 25 Jun. 2003, WT/DS276/12.
149
Panel report in US – Upland Cotton, WT/DS267/R, paras 7.55–7.69, fn. 134, paras 7.70–7.103, fn. 134, paras
7.104–7.128, fn. 180.
150
For example, preliminary rulings in Australia – Apples, 6 Jun. 2008, WT/DS367/7; China – Raw Materials,
7 May 2010, WT/DS394/9.
62
manner. If the ruling on the peace clause had been negative (which it was not), it would
also have rendered much of the rest of the case moot.151 The panel was able to issue its final
report in that very large and complex dispute in just under thirteen months from the date of
its composition. In other disputes, the claims may not be amenable to this type of sequential
analysis. Complainants may also be reluctant to support multi-stage procedures as additional
steps in a timetable necessarily entail some delay, but the alternative in the largest cases has
been a series of unexpected delays after panels become bogged down in too many issues at
once, which does not serve the parties’ interests better.
10. SCIENTIFIC AND TECHNICAL EXPERTS
The need to consult scientific experts has been cited as a reason for delay by several panels,
notably those examining disputes under the SPS Agreement, but others as well.152 Consulta-
tion of experts is never given as the only reason for delay, as the need for expert evidence is
related to the scope and complexity of the relevant disputes. Expert consultation increases the
volume of material that the panel must examine, both directly through the experts’ responses
to questions and indirectly through parties’ reactions to the expert evidence, and it also relates
to the complexity of the dispute, where the expert evidence addresses scientific issues.
The procedures for consultation with technical experts also create their own delays.153
Additional time is required to draft and consult the parties on the working procedures for
expert evidence, to identify the areas on which expert advice is required, to identify and
then select experts, to invite or propose questions to the experts and finalize them, to give
the experts time to respond, and then to hold a meeting with the experts in the presence
of the parties. Each of these steps can be highly contested because expert evidence goes to a
central issue in the dispute. The parties may appeal not only the panel’s findings on the
substantive issues but also the panel’s handling of the procedural aspects of expert consulta-
tion. In US/Canada – Continued Suspension of Obligations in the EC-Harmony Dispute, the
Appellate Body upheld an appeal that the Panel failed to comply with its duties under
Article 11 of the DSU to make an objective assessment of the matter because of the identity
of two experts that it had selected.154
Up until 2005, all but one of the panels that consulted scientific experts managed
to issue their final reports to the parties within twelve months of composition.155
151
Panel report in US – Upland Cotton, WT/DS267/R/Add.3, Annex L-4, paras 18–21.
152
Australia – Salmon, WT/DS18/4; Australia – Salmon (Article 21.5), WT/DS18/17; EC – Hormones, WT/DS26/8
(but not WT/DS48); US – Shrimp, WT/DS58/10; Japan – Agricultural Products II, WT/DS76/4; EC – Asbestos, WT/
DS135/5; Japan – Apples, WT/DS245/4; Japan – Apples (Article 21.5), WT/DS245/17; EC – Biotech (GMOs), WT/
DS291/26, 27, and 28; Australia – Apples, WT/DS367/8 and 9; Korea – Bovine Meat (Canada), WT/DS391/5. The panel
in the US/Canada – Continued Suspension of Obligations in the EC-Harmony Dispute cases referred to the scientific and
technical issues in general: WT/DS320/10 and 11; WT/DS321/10 and 11.
153
See working procedures for expert consultation at WT/DS245/R, para. 6.2; WT/DS245/RW, para. 6.2;
WT/DS320/R, Annex A-5; and WT/DS321/R, Annex A-5.
154
WT/DS320/AB/R, paras 433–484, 736(b).
155
Australia – Salmon, Australia – Salmon (Article 21.5), EC – Hormones, US – Shrimp, Japan – Varietals, Japan –
Apples, and Japan – Apples (Article 21.5).
63
The exception was EC – Asbestos (which was not an SPS dispute), in which the panel
issued its final report just under sixteen months from the date it was composed. Since 2006,
the three panels that have consulted scientific and technical experts in SPS disputes have
each taken more than two years to issue their respective reports to the parties.156 This
appears partly to be linked to the increase in complexity of the procedures themselves for
consultation of experts, in particular for the selection of the experts, which is multiplied
when the dispute concerns more than one pest, disease, or food additive.
The panel in EC – Biotech (GMOs) deferred a decision on whether to seek advice
from scientific and technical experts until after it had received the second written submis-
sions of the parties. There was a ten-week expert selection process, involving nominations
by six international organizations and by the parties of 100 candidates, of which forty-eight
expressed interest, with a second round that took an additional month after it turned out
that the first four experts selected were not likely to provide advice on two issues. The
experts answered questions posed by the panel within eleven weeks or less. The panel also
sought advice from seven international organizations regarding the meaning of certain
terms and concepts, many of which appear in the WTO agreements (such as ‘pest’).157
This expert review process lasted six months.
The panels in US/Canada – Continued Suspension of Obligations in the EC-Harmony
Dispute decided to seek advice from scientific and technical experts after the first substantive
meeting with the parties.158 Five weeks later, the panel adopted special working proce-
dures.159 There was then a four-month two-round expert nomination process, involving
seventy-one nominations by international organizations and the parties, of which thirty-
five candidates expressed availability and provided a curriculum vitae. The panel selected
six experts, to which there was a series of objections (one of which was ultimately upheld
on appeal) plus two withdrawals and replacements. It then took seven weeks to finalize the
draft questions after which the experts replied in eight weeks, followed by two rounds of
comments by the parties and a meeting open to the public (which created an additional
scheduling constraint).160 The panel posed sixty-two questions to the experts in seven fields
as well as twenty-six questions to three international entities more focused on definitional
issues, as well as institutional and procedural issues. This expert consultation process lasted
eleven months.161
The panel in Australia – Apples cited its decision to seek scientific and technical expert
advice as a reason for delay and also spelt out why this created delay, listing the time needed
to identify and select experts, to prepare the questions for the experts in consultation with
156
Another panel is currently at work in an SPS dispute. This panel, in Korea – Bovine Meat (Canada), has already
cited the expert consultation procedure as a reason for delay. The Chair has estimated that the panel will issue its final
report to the parties seventeen months after composition: WT/DS391/5.
157
WT/DS291/R, paras 7.12–7.32 and Annex H.
158
WT/DS320/R, paras 7.55–7.64.
159
Ibid., paras 7.60 and 7.71.
160
WT/DS320/10, WT/DS321/10.
161
WT/DS320/R, paras 7.76–7.99, Annexes D and E.
64
the parties, the time required by experts to prepare their responses, and the time provided
to Parties to comment on these responses.162
A strategy to deal with the delays caused by expert review would include clarification
of some of the recurring issues, such as what constitutes a conflict of interest in the selection
of experts, in light of the fact that only a limited number of reliable scientists may have
conducted research on a particular health risk and those that have done so are likely to have
expressed a view. The respective roles of the panel and the experts in the consultation of
experts could also be elucidated further. Members are best placed to discuss these issues in
regular sessions of the DSB or the DSU negotiations; otherwise, they are decided by the
Appellate Body by default.
11. TRANSLATION
Translation has been cited as a reason for delay in issuing final report to the parties but in
distinct ways. Submissions from one or both parties may have to be translated.163 This can
cause delays earlier in the proceeding. If both parties make their submissions in Spanish, the
solution is to appoint three panellists that can work in that language, as in Mexico – Stainless
Steel.164 If the parties make their submissions in different languages, and the panel report
must be translated prior to issuance to the parties, some of that time can usually be recouped
later before the report is circulated to Members in three languages.165
The meaning of key words in the measures at issue may give rise to substantive
disagreements where they are only authentic in a non-WTO language. For example, the
panel in China – Audiovisual Services cited translation of measures (from the original Chinese
version) as a reason for delay.166 However, the panel did not ask the parties to explain the
rationale for their respective translations until one month after its second and final meeting
and could not locate an available translator until three months after that meeting (in Kenya).
After the translations of a handful of terms were received, the panel offered the parties two
opportunities to comment, a process that may not have concluded until over four or five
months after the second meeting.167 The solution to translation disagreements is to adopt
working procedures to identify differences in translation, and resolve them, at the earliest
possible stage. The contemporaneous panel in China – Intellectual Property Rights made
provision for translation disagreements in its working procedures, which led to the resolu-
tion of disagreements regarding the meaning of the measures by the time of the second
substantive meeting.168
162
WT/DS367/8, 9, and 10.
163
Mexico – Olive Oil, WT/DS341/4 (complainant: European Communities); Colombia – Ports of Entry,
WT/DS366/8 (complainant: Panama).
164
The complainant was Guatemala, WT/DS331.
165
Above n. 59.
166
WT/DS363/8.
167
Panel report on China – Audiovisual Services, WT/DS363/R, paras 2.4–2.9, 7.928–7.931 and Annex A-1:
‘Translation Differences in the Report – Summary Table’.
168
China – Intellectual Property Rights, WT/DS362, paras 2.5 and 2.6. See also panel report in Japan – Film, WT/
DS44/R, paras 1.9–1.11 and Annex on Translation Problems.
65
12. CIRCULATION OF THE FINAL REPORT
Circulation of the panel’s final report to Members is a precondition to its adoption by the
DSB. Circulation occurs simultaneously in the three WTO official languages; hence,
translation is required before the panel report can have any legal effect. Translation takes,
on average, about two months after the issuance of the final report to the parties.
The proposed timetable in Appendix 3 of the DSU only foresees three weeks for circula-
tion in item (k). Translation times increased as many panels adopted a practice of annexing
parties’ submissions, or executive summaries of them, to the panel report. Matters have
improved since panels ceased to annex parties’ responses to questions, but the time taken
for translation of submissions, as well as findings, remains considerable. Translation times
lengthen when there are many panel proceedings being conducted concurrently and also
where the priority given to the translation of panel reports is downgraded, as may occur
when Appellate Body reports are being translated or during the preparations for a Minis-
terial Conference.
The volume of material to be translated in a panel report includes the descriptive part,
comprising factual and argument sections, as well as the findings and conclusions sections.
Article 15.1 of the DSU requires panels to issue the descriptive part to the parties to the
dispute for comment before they issue the findings section. This requirement has its origins
in GATT practice, according to which panels issued the descriptive part and the conclu-
sions or an outline of them to the parties before circulation in order to encourage
development of a mutually satisfactory solution.169 WTO practice has developed to the
point where the parties prepare executive summaries of their own submissions, so that
many panels now go through the motions of issuing a descriptive part that consists of a
short procedural summary of the dispute together with the executive summaries received
from the parties themselves. Unless a key fact is in dispute, this serves no real purpose prior
to the circulation of the final report. Panels recite the key arguments in the findings to
provide the essential context, where they are read after circulation of the final report.
A more rational approach would be to amend Article 15.1 of the DSU to allow panels
to dispense with a separate descriptive part and maintain the current practice of reciting the
key arguments in the findings. Parties could upload their submissions or non-confidential
summaries onto the WTO website if they wished in order to provide a comprehensive
record. The Appellate Body receives the panel record in full when a notice of appeal is filed
in any case.170 These submissions would be in the original language only. Translators could
then devote their energies to the findings and conclusions sections of panel reports and
reduce the time taken to circulate those reports to Members.
169
1979 Understanding Annex on Agreed Description of the Customary Practice of the GATT in the Field of
Dispute Settlement (Art. XXIII:2), above n. 27, para. 6(vii).
170
Rule 25 of the Working Procedures for Appellate Review, WT/AB/WP/6.
66
13. CONCLUSION
The prompt settlement of disputes is essential to the effective functioning of the WTO,
especially in light of the fact that the WTO dispute settlement system lacks interlocutory
and retrospective relief. Delays occur at different stages of the procedures, but the delays at
the panel stage are increasing, and increasingly frequent. The steady, gradual trend towards
longer panel proceedings throughout the first ten years of the WTO has now given way to
steeper increases. There is an underlying trend towards longer panel proceedings in more
ordinary cases that do not involve extraordinary procedures. If this upward trend goes
unchecked, it may diminish the effectiveness of the WTO dispute settlement system.
The DSU time frames for panel proceedings are not respected because they were
unrealistic from the start. WTO procedures have developed, at the parties’ request, to offer
more time and opportunities to address the issues. These created a de facto minimum time
frame of over nine months from panel composition to issuance of final reports to the
parties, to which other considerations often add more time. Comparisons with the time
frames for appellate review cannot be taken at face value. The DSU contains a binding,
extended time frame for appeals that begins to run long after the Appellate Body has
actually commenced work, while for panels the position is just the opposite. A realistic time
frame also creates a downward pressure on the length of reports, something that panels lack.
Panel proceedings have sharply increased in length since 2004 to 2005 when two long
SPS disputes and the Airbus and Boeing cases demonstrated that there is no upper limit to
the length of WTO panel proceedings. Complainants’ and respondents’ submissions
increase the scope of matters in dispute, as do the growing number of procedural issues,
but many of the longer delays in more ordinary cases have occurred after the parties have
completed their submissions. Panellists’ other commitments can cause delay and should be
given more consideration during panel composition. Panels sometimes pose too many
written questions and panel reports are sometimes excessively detailed. Sufficient Secretariat
staff should always be made available to assist panels.
Parties to disputes and the collective membership can re-examine standard panel
procedures to which procedures they are willing to accelerate or eliminate without
compromising the quality of panel reports. Proposals to modify times for specific submis-
sions and meetings could save up to two months of time in appropriate cases and could be
more effective than broad proposals to compress overall time frames. These steps do not
require amendment of the DSU, but panels would probably be more willing to agree to
some of them if the DSB confirmed that they were not per se unfair to either party.
If the time spent at the panel stage continues to increase at its present rate, it will offset
the time savings, and add to the time costs, delivered by proposals to reform other stages of
the dispute settlement procedures. The DSU negotiators could also consider the possibility
of accelerating the circulation of reports by amending Article 15.1 of the DSU to abolish
the requirement for a separate descriptive part of the panel report. Panels could continue
67
the current practice of reciting key facts and arguments in the findings and parties could
upload their submissions or summaries of them on the WTO website.
The sharp increase in the time taken to issue panel reports is neither inevitable nor
irreversible. Panels can certainly strive to fix tighter timetables and ensure that their findings
are drafted sooner, but Members also have their part to play in anticipating undue delays,
not only in the DSU negotiations but also in their capacity as parties to disputes.
68
The Question of Remand Authority for the Appellate Body
THE QUESTION OF REMAND AUTHORITY

FOR THE APPELLATE BODY*
APPELLATE
FERNANDO PIÉROLA†
In this chapter, the author addresses the possibility of

introducing remand authority for the Appellate Body of the
World Trade Organization. In the absence of such authority,
uncertainty and unpredictability can arise in particular
cases due to the current limits on the scope of appellate
review. The willingness of the Appellate Body to complete
the panel’s analysis in some circumstances does not fully
resolve this problem, especially where the panel’s factual
findings are inadequate. Moreover, the legal basis of this
Appellate Body technique may be questioned. The
introduction of remand authority would improve due
process and fairness in WTO dispute settlement, and the
practical implications in terms of additional time and
workload needed for dispute settlement proceedings would
not be overwhelming. On balance, remand authority is an
attractive solution to the existing problems and should be
preferred to the alternatives for reform in this area.
*
A Moza, por todo el amor que nos diste en casa.
†
Counsel, Advisory Centre on WTO Law. The views expressed in this chapter are my own.
I would like to give special thanks to Andrew D Mitchell for his valuable comments and to
William Davey for his comments on a prior draft of this chapter.
193
69
I INTRODUCTION
Under GATT 1947, panels’ recommendations that losing parties bring their
illegal measures into conformity were adopted by the so-called ‘positive
consensus rule’. According to this rule, any GATT contracting party,
including the losing party to a dispute, could block the adoption of a Panel
Report.1 On the one hand, this practice was seen as a legitimate means to
involve GATT contracting parties in the resolution of trade disputes, given
the failure to establish GATT as an intergovernmental organisation and the
consequent lack of treaty-based jurisdiction by panels to examine trade
disputes. On the other hand, this practice sometimes prevented satisfactory
resolution of trade disputes when the matters at issue were politically
contentious. In order to overcome this situation, WTO Members agreed to
change the rule from positive consensus to negative consensus. In other
words, under the DSU, Members should adopt Panel Reports automatically
unless there is a consensus to the contrary.2
The creation of a standing Appellate Body for the judicial settlement of
trade disputes is one of the major innovations of the WTO dispute
settlement system.3 The concept of an Appellate Body can be seen as a
counterpart to the introduction of the negative consensus rule, to prevent
the possibility of panels’ legal errors being adopted ‘untouched’.4
However, the Appellate Body’s powers are limited. In particular, the scope
of appellate review is restricted, and the Appellate Body lacks remand
authority – that is, the authority to return a matter to the original panel to
be decided in accordance with the Appellate Body’s ruling. In some
1
Julio Lacarte and Fernando Piérola,‘Comparing the WTO and GATT Dispute Settlement
Mechanisms: What was Accomplished in the Uruguay Round?’ in Julio Lacarte and Jaime
Granados (ed), Inter-Governmental Trade Dispute Settlement: Multilateral and Regional
Approaches (2004) 33, 51.
2
DSU, art 16.4. The negative consensus rule also applies to the adoption of Appellate Body
reports, the establishment of panels, and the authorisation to suspend concessions and other
obligations.
3
Ernst-Ulrich Petersmann,‘How to Promote the International Rule of Law? Contributions
by the World Trade Organization Appellate Review System’ (1998) 1 Journal of International
Economic Law 25, 38; see also Robert Howse and Michael Trebilcock, The Regulation
International Trade (2nd ed, 1999) 78; David Palmeter and Petros Mavroidis, Dispute
Settlement in the World Trade Organization: Practice and Procedure (1999) 147.
4
Robert Hudec, ‘The New WTO Dispute Settlement Procedure: An Overview of the First
Three Years’ (1999) 8 Minnesota Journal of Global Trade1, 27; John H Jackson,The World
Trading System: Law and Policy of International Economic Relations (2nd ed, 1999) 125;
John H Jackson, William Davey and Alan Sykes,Legal Problems of International Economic
Relations, Cases Materials and Text on the National and International Regulation of
Transnational Economic Relations (3rd ed, 1995) 343; Giorgio Sacerdoti, ‘Appeal and Judicial
Review in International Arbitration and Adjudication: The Case of the WTO Appellate
Review’ in Ernst-Ulrich Petersmann (ed), International Trade Law and the GATT/WTO Dispute
Settlement System, Studies in Transnational Economic Law, vol 11 (1997) 245, 247.
195
70
Challenges and Prospects for the WTO
circumstances, given the negative consensus rule, this means that Panel
Reports may be adopted as modified by an Appellate Body Report, leaving
certain matters unresolved. This may raise doubts as to whether the
relevant recommendations or rulings have ‘achiev[ed] a satisfactory
settlement of the matter’5 and contributed to ‘security and predictability
[in] the multilateral trading system’6 in accordance with the DSU.
In the current negotiations on the DSU, some Members have raised the
possibility of providing the Appellate Body with remand authority.7 This
chapter evaluates this possibility, based on the experience of the WTO
dispute settlement system since 1995. In the following section, I examine
the limits on the powers of the Appellate Body under the DSU, the
particular circumstances in which the absence of remand authority
becomes relevant and problematic, and the existing alternatives of the
Appellate Body when such circumstances arise. Next, I consider the
advantages and disadvantages of instituting remand authority in light of
the alternatives. Finally, the chapter draws a conclusion on the feasibility
of introducing remand authority and its likely impact in WTO disputes.
II PROBLEMS ARISING FROM THE CURRENT SYSTEM
A Powers of the Appellate Body

The DSU confers the Appellate Body two significant types of powers:
those relating to the scope of appellate review, and those relating to the
actions that the Appellate Body may take.
Article 17.1 of the DSU provides for the establishment of the Appellate Body
to ‘hear appeals from panel cases’. This provision implies that the Appellate
Body cannot decline jurisdiction over challenges to panel decisions by way
of appeals. Furthermore, it implies that the Appellate Body’s main function
is to review issues already dealt with at the panel stage, but it does not
expressly exclude de novo reviews under specified circumstances.
The scope of the Appellate Body’s activity is limited under Article 17.6 ‘to
issues of law covered in the panel report and legal interpretations developed
by the panel’. This provision excludes the Appellate Body’s examination of
facts except in two specific circumstances: (i) examination of the facts of the
case in order to examine whether the panel made an objective assessment of
5
DSU, art 3.4.
6
DSU, art 3.2.
7
Communication from Jordan, Jordan’s Further Contribution Towards the Improvement
and Clarification of the Dispute Settlement Understanding, TN/DS/W/56 (19 May 2003);
Communication from the European Communities, Contribution of the European
Communities and Its Member States to the Improvement and Clarification of the WTO
Dispute Settlement Understanding, TN/DS/W/38 (23 January 2003).
196
71
them; and (ii) examination of the facts of the proceedings in order to determine
whether the panel’s procedural conduct was in accordance with due process
requirements. Even though they are based on facts, the Appellate Body has
characterised these two kinds of matters as issues of law.
Article 17.13 provides that ‘[t]he Appellate Body may uphold, modify or
reverse the legal findings and conclusions of the panel’. Thus, after
examining the issues of law raised on appeal, the Appellate Body may
agree with the panel and uphold the panel’s findings, interpretations and
conclusions. Alternatively, it may disagree with the panel. At this stage, the
Appellate Body has to decide how to proceed: it either modifies the panel’s
findings by stating its own legal reasoning or reverses the panel’s findings
and replaces them with its own, or it just modifies or reverses the panel’s
findings without providing subsequent reasoning or findings. Under the
DSU, where the Appellate Body modifies or reverses panel findings, it is
not expressly entitled to remand the matter to the panel to be decided anew.8
Furthermore, Article 19.2. points out that ‘in their findings and
recommendations, the panel and Appellate Body cannot add or diminish
the rights and obligations in the covered agreements’. This provision
imposes a horizontal control on the Appellate Body’s findings and
conclusions – they must be limited to the matters discussed in the course
of the dispute and the parties involved therein.
B Panels’ Use of Judicial Economy

In many instances, panels leave certain factual or legal issues resolved as
a result of the principle of ‘judicial economy’.9 By virtue of this principle,
‘[a] panel need only address those claims which must be addressed in
order to resolve the matter in issue in the dispute’.10 The discretion of
panels to limit their findings to the minimum required to resolve the
dispute, applying the principle of judicial economy, is widely accepted in
most legal systems.11 Acknowledging the value of judicial economy in
the GATT/WTO context, the Appellate Body has stated:
8
Nonetheless, a prestigious expert has submitted that the Appellate Body has implied remand
powers as ‘the negative aspects of the current solution outweigh the interpretation of the
DSU: that the Appellate Body has the prerogative to send the case back to the DSB so that
the proceeding can be completed by a panel in compliance with the system set up by the
DSU’: Jacques Bourgeois, ‘Some Reflections on the WTO Dispute Settlement System from a
Practitioner’s Perspective’ (2001) 4 Journal of International Economic Law 145, 152.
9
William Davey, ‘Has the WTO Dispute Settlement System Exceeded its Authority? A
Consideration of Deference Shown by the System to Member Government Decisions and Its
Use of Issue-Avoidance Techniques’ (2001) 4 Journal of International Economic Law 79, 108.
10
Appellate Body Report, US – Wool Shirts and Blouses, 19.
11
Jeffrey Waincymer, WTO Litigation, Procedural Aspects of Formal Dispute Settlement
(2002) 371.
197
72
Nothing … in previous GATT practice requires a panel

to examine all legal claims made by the complaining
party. Previous GATT 1947 and WTO panels have
frequently addressed only those issues that such panels
considered necessary for the resolution of the matter
between the parties, and have declined to decide other
issues. Thus, if a panel found that a measure was
inconsistent with a particular provision of the GATT
1947, it generally did not go on to examine whether the
measure was also inconsistent with other GATT
provisions that a complaining party may have argued
were violated. In recent WTO practice, panels likewise
have refrained from examining each and every claim
made by the complaining party and have made findings
only on those claims that such panels concluded were
necessary to resolve the particular matter.12
Judicial economy appears to be a valid way of dealing with controversial
points that need not be addressed in order to settle a dispute.13 In addition,
Palmeter and Mavroidis point to the fact that judicial economy is useful
to save judicial resources and a ‘less expansive approach which seems
appropriate to bodies reviewing the programmes and actions of sovereign
governments’.14 Even though judicial economy is frequently exercised in
respect of claims raised by complainants, we consider that this principle
may also be applied to respondents’ defences.
The use of judicial economy by panels has been upheld by the Appellate
Body in several disputes, such as US – Wool Shirts and Blouses,15 Canada –
Autos,16 India – Patents (US),17 Australia – Salmon,18 US – Wheat Gluten,19
and US – Lead and Bismuth II.20 The Appellate Body has also used judicial
economy in its own proceedings.21
Judicial economy appears to be a useful instrument for the settlement of
disputes at the panel stage. It saves judicial costs for the panel and the
WTO Secretariat, minimises the political impact of rulings for the parties
and other WTO Members, and reduces the length of the litigation process.
12
13
Waincymer, above n 11, 368.
14
Palmeter and Mavroidis, above n 3, 149.
15
16
Appellate Body Report, Canada – Autos, para 110.
17
Appellate Body Report, India – Patents (US), para 87.
18
Appellate Body Report, Australia – Salmon, para 223.
19
Appellate Body Report, US – Wheat Gluten, paras 177–186.
20
Appellate Body Report, US – Lead and Bismuth II, para 71.
21
See, eg, Appellate Body Report, Argentina – Footwear (EC), para 98.
198
73
Nonetheless, it also imposes certain shortcomings when the Appellate

Body reverses panel findings.
C Circumstances in which Remand Authority Could be Used

Given the limitation on the Appellate Body’s power to examine issues of
fact and the absence of remand authority, the Appellate Body may in
certain circumstances find itself in a problematic situation when it modifies
or reverses findings, interpretations or conclusions of the panel.
In some instances, modification or reversal of a panel’s legal findings,
interpretations or conclusions may not be an obstacle to the resolution
of a dispute. This may happen, for example, where the Appellate Body
modifies or reverses a legal finding that is the ultimate step in the panel’s
legal analysis under a specified provision, and the facts underlying that
legal issue are either undisputed or sufficiently established by the panel.
For instance, in EC – Bed Linen, the panel interpreted Article 2.2.2(ii) of
the Anti-Dumping Agreement as allowing a Member investigating
dumping to construct an exporter’s administrative, selling and general
costs (SG&A) based on the actual SG&A of one other exporter subject to
investigation, even though the provision refers to the weighted average
from other exporters or producers subject to investigation. Accordingly, the
panel found that the European Communities’ methodology was WTO-
consistent.22 The Appellate Body disagreed with the panel’s interpretation
of Article 2.2.2(ii). From the wording ‘weighted average’, ‘exporters’ and
‘producers’ in that provision, the Appellate Body reasoned that the basis
for the methodology in question could not be information from a single
exporter. The Appellate Body therefore reversed the panel’s interpretation.
As the European Communities did not dispute the fact that it had taken
into account information from only a single exporter (although this factor
is not expressly mentioned in the Appellate Body Report), the Appellate
Body concluded that the European Communities had violated Article
2.2.2(ii) of the Anti-Dumping Agreement.23 As this issue arose only at the
final step of the legal analysis, the absence of remand authority did not
prevent this issue from being resolved.
However, in other cases, reversals or modifications pose difficulties for
the ultimate resolution of disputes. A reversal or modification of certain
legal findings or conclusions may entail:
(a) the resurgence of issues within the legal analysis under a specified
provision; or
22
Panel Report, EC – Bed Linen, para 6.75.
23
Appellate Body Report, EC – Bed Linen, para 85.
199
74
(b) the revival of subsidiary claims or defences before the panel.

Two frequently cited cases24 illustrate these situations. With respect to the
situation described in (a), the Appellate Body in US – Gasoline reversed
the panel’s finding that the measure at issue prescribing quality standards
for gasoline was not a measure ‘relating to’ an exhaustible resource within
the meaning of Article XX(g) of GATT 1994.25 Given this finding, the panel
did not proceed with the following issues: (i) whether the measure was
‘made effective in conjunction with restrictions on domestic production
or consumption’ under Article XX(g); and (ii) whether the measure was
consistent with the requirements of the chapeau of Article XX. After the
Appellate Body’s reversal, these issues became relevant without the benefit
of the panel’s analysis.26 With respect to the situation described in (b)
above, in Canada – Periodicals, the Appellate Body reversed the panel’s
finding that the measure at issue was inconsistent with the first sentence
of Article III:2 of GATT 1994.27 In addition to this claim, Canada had raised
a claim under GATT Article III:2, second sentence. However, the panel
expressly declined to decide this claim as it had already found that the
measure at issue was inconsistent with the first sentence.28
Thus, Appellate Body reversal or modification of panel findings may pose
problems where:
• whether by virtue of the principle of judicial economy’ or otherwise,
the panel did not address the relevant legal issue, claim or defence; or
• the panel has not even made factual findings on the relevant issues,
claims or defences.
D Existing Alternatives for the Appellate Body in the Absence of

Remand Authority
In domestic courts, after reversing the findings of a lower court, a higher
court would typically:
(a) examine the issues that were not addressed by the lower court and
‘complete’ the analysis on these matters;
(b) remand the case to the lower court in order to let it correct any
errors and complete the analysis; or
24
See, Palmeter and Mavroidis, above n 3, 148 and Terence Stewart and Amy Ann Karpel,
‘Review of the Dispute Settlement Understanding: Operation of Panels’ (2000) 31 Law and
Policy in International Business 642.
25
Appellate Body Report, US – Gasoline, 19.
26
Ibid.
27
Appellate Body Report, Canada – Periodicals, 23.
28
Panel Report, Canada – Periodicals, para 5.30.
200
75
(c) leave the issues undecided (although this alternative is negated

under most domestic legal systems as it has the effects equivalent
to non liquet).
Given the absence of remand authority under the DSU, the only options
available to the Appellate Body in these circumstances are (a) or (c). I
now consider these options in turn.
1 Completing the Analysis
(a) The Practice and its Features
(i) Duty or Discretion?

The Appellate Body has repeatedly completed the analysis of panels since
its first appellate review.29 This is demonstrated in the two appeals just
mentioned as raising issues that the panel had not addressed. In US –
Gasoline, after reversing the panel’s findings and conclusions on whether
the measure at issue was a measure ‘relating to’ an exhaustible resource
within the meaning of Article XX(g) of GATT 1994, the Appellate Body
completed the legal analysis in order to examine whether the measure at
issue was ‘necessary’ under Article XX(g) and whether the measure was
consistent with the requirements of the chapeau of GATT Article XX. The
Appellate Body found that the panel had erred in law in failing to decide
the other aspects of the Article XX analysis.30 Completion of the analysis
in this case was within the examination of the same provision at issue,
namely Article XX.
In Canada – Periodicals , the Appellate Body completed the analysis again
but under different circumstances. After reversing the panel’s finding
under Article III:2, first sentence, the Appellate Body considered it
necessary to analyse the complainant’s claim under Article III:2, second
sentence. Unlike in US – Gasoline, the Appellate Body stated that it had a
duty to do so:
As the legal obligations in the first and second
sentences are two closely-linked steps in determining
the consistency of an internal tax measure with the
national treatment obligations of Article III:2, the
Appellate Body would be remiss in not completing
the analysis of Article III:2. In the case at hand, the
Panel made legal findings and conclusions concerning
the first sentence of Article III:2, and because we
29
See the cases cited in Appellate Body Report, EC – Asbestos, para 78.
30
201
76
reverse one of those findings, we need to develop our

analysis based on the Panel Report in order to issue
legal conclusions with respect to Article III:2, second
sentence, of the GATT 1994.31
Similarly, in US – Shrimp,32 the Appellate Body again faced the need to
complete the analysis and stated that this is, in some circumstances, its
responsibility:
[W]e believe that it is our duty and our responsibility
to complete the legal analysis in this case in order to
determine whether Section 609 qualifies for
justification under Article XX. In doing this, we are
fully aware of our jurisdiction and mandate under
Article 17 of the DSU. … We do this, in part, recognizing
that Article 3.7 of the DSU emphasizes that: ‘The aim
of the dispute settlement mechanism is to secure a
positive solution to a dispute.’33
In carrying out this practice under a sense of ‘duty’, which the DSU does
not explicitly impose, it may be argued that the Appellate Body has
developed a customary rule for its own proceedings.
(ii) The Need for a Factual Basis and a Logical Continuum

As previously suggested, the Appellate Body can complete the panel’s
analysis only where the facts allow it. In US – Shrimp, after asserting its
duty to complete the analysis, the Appellate Body stated:
Fortunately, in the present case, as in the mentioned
previous cases, we believe that the facts on the record
of the panel proceedings permit us to undertake the
completion of the analysis required to resolve this
dispute. 34
In Australia – Salmon, the Appellate Body again indicated that the need
for facts to complete the analysis:
[W]e believe that — to the extent possible on the basis
of the factual findings of the Panel and/or of
undisputed facts in the Panel record — we should
complete the legal analysis and determine whether the
31
32
Appellate Body Report, US – Shrimp, para 123.
33
Ibid paras 123–124.
34
Ibid.
202
77
actual SPS measure at issue, ie, Australia’s import

prohibition on fresh, chilled or frozen ocean-caught
Pacific salmon, is based on a risk assessment.35
Conversely, the Appellate Body has refused to complete the analysis on
many occasions in which completion of the analysis was impossible
because of deficiencies in the factual findings or the panel record.36 For
instance, in EC – Hormones , after reversing the panel’s conclusion under
Article 5.5 of the SPS Agreement, the Appellate Body considered that ‘it
[could not] be assumed that all the findings of fact necessary to proceed
to a determination of consistency or inconsistency of the EC measures
with the requirements of Article 5.6 [had] been made by the Panel…’.37
The same occurred in Korea – Dairy, where the Appellate Body found that:
In the absence of any factual findings by the Panel or
undisputed facts in the Panel record relating to whether
the alleged increase in imports was, indeed, ‘a result of
unforeseen developments’ and of the effect of the
obligations incurred by a Member under this agreement,
… we are not in a position, within the scope of our
mandate set forth in Article 17 of the DSU, to complete
the analysis and make a determination as to whether
Korea acted inconsistently with its obligations under
Article XIX: 1(a). Accordingly, we are unable to come to
a conclusion on whether or not Korea violated its
obligation under Article XIX:1(a) of the GATT 1994.38
The criterion of complete factual record or factual findings by the panel
appears to be supplemented by a more substantive element when the
completion of the analysis concerns provisions other than those examined
by the panel. The Appellate Body has stated on certain occasions that, in
order to complete the analysis, there must be a ‘logical continuum’ between
the provisions at issue:
The need for sufficient facts is not the only limit on
our ability to complete the legal analysis in any given
case. In Canada – Periodicals, we reversed the panel’s
conclusion that the measure at issue was inconsistent
with Article III:2, first sentence, of the GATT 1994, and
we then proceeded to examine the United States’
claims under Article III:2, second sentence, which the
35
Appellate Body Report, Australia – Salmon, para 118.
36
See the cases cited in ibid para 79.
37
Appellate Body Report, EC – Hormones, 229–230.
38
Appellate Body Report, Korea – Dairy, para 92.
203
78
panel had not examined at all. However, in embarking

there on an analysis of a provision that the panel had
not considered, we emphasized that ‘the first and
second sentences of Article III:2 are closely related’ and
that those two sentences are ‘part of a logical
continuum.’39
(b) Assessing the Practice
(i) Factual Basis

Even though the Appellate Body has indicated that it will complete the
analysis only when presented with sufficient factual findings by the panel
or undisputed facts in the panel record, it is important to note that the facts
on the panel record do not necessarily constitute factual findings of the
panel.40 In addition, not all undisputed facts contained in the record are
truly ‘undisputed’ in the sense that the parties would agree on them. Rather,
the parties might not have disputed the facts purely because they did not
consider them or did not regard them as relevant in resolving the dispute.
Another concern arising from the Appellate Body’s rulings in this area is
that in some controversial cases it has completed the analysis in
circumstances that required it to resolve issues of fact. As noted by
Appleton,41 in US – Shrimp for example, although the Appellate Body
stated that ‘the record of the panel proceedings permit[s] us to undertake
the completion of the analysis required to resolve this dispute’,42 it made
some factual findings with regard to the actual application of the measure
at issue.43 Furthermore, in US – Wheat Gluten , the Appellate Body also
engaged in a factual analysis in relation to whether the US authorities
had considered the protein content of wheat as a relevant factor for the
assessment of domestic consumption, and also the price of wheat gluten
for the determination of injury.44
(ii) Legal Basis and Uncertainty

The basic argument in favour of completing the analysis is that it enables
the Appellate Body to provide a positive response to complaints when
panels exercise judicial economy or otherwise leave issues unanswered.
39
Appellate Body Report, EC - Asbestos, para 79 (emphasis added, footnotes omitted).
40
Arthur E Appleton, ‘Shrimp/Turtle: Untangling the Nets’ (1999) 2 Journal of International
Economic Law 477, 479.
41
Ibid.
42
Appellate Body Report, US – Shrimp, paras 123–124.
43
Ibid paras 163–164, 179.
44
Appellate Body Report, US – Wheat Gluten, para 59.
204
79
However, the legal basis for the Appellate Body to complete the analysis
is unclear. Even though several provisions of the DSU may implicitly
accommodate this practice,45 Articles 17.6 and 17.13 do not expressly
authorise the Appellate Body to complete a panel’s legal analysis.
In Canada –Periodicals, Canada argued that the Appellate Body lacked
jurisdiction to examine a claim under Article III:2, second sentence, of
GATT 1994, given that no party had appealed panel findings relating to
that provision. Canada also referred to the requirement in Article 17.6 of
the DSU that an appeal be limited to issues of law covered in the Panel
Report and legal interpretations developed by the panel.46 The Appellate
Body responded:
We believe the Appellate Body can, and should,
complete the analysis of Article III:2 of the GATT 1994
in this case by examining the measure with reference
to its consistency with the second sentence of Article
III:2, provided that there is a sufficient basis in the
Panel Report to allow us to do so …47
The Appellate Body provided no explanation as to the legal basis for
completing the analysis in the light of Article 17.6 of the DSU.
In Australia – Salmon, before completing the analysis of a claim that the
panel had not addressed, stated:
We are cognizant of the provisions of Article 17 of the
DSU that state our jurisdiction and our mandate. Article
17.6 of the DSU provides: ‘An appeal shall be limited
to issues of law covered in the panel report and legal
interpretations developed by the panel.’ Article 17.13
of the DSU states: ‘The Appellate Body may uphold,
modify or reverse the legal findings and conclusions of
the panel.’ In certain appeals, when we reverse a panel’s
finding on a legal issue, we may examine and decide
an issue that was not specifically addressed by the panel,
in order to complete the legal analysis and resolve the
dispute between the parties …48
It is important to note that, in addressing claims not addressed by the
panel, the Appellate Body has implicitly enlarged its powers to look not
45
For example, Article 3.3 of the DSU emphasises the importance of ‘prompt settlement’ of
disputes, which would be hindered if issues were left undecided.
46
47
Ibid 24.
48
Ibid para 117.
205
80
only into issues not covered by the Panel Report, but also to make findings
beyond those of the panel. A strict reading of Articles 17.6, 17.12 and 17.13
of the DSU to prevent this kind of review by the Appellate Body could
lead to unsatisfactory results for complainants. This was the case in EC –
Poultry, where the Appellate Body held:
[The panel conclusion] leaves unanswered two issues
that were raised by Brazil before the Panel … With
respect to these two issues, we are mindful of our
mandate under Article 17.6 … and … Article 17.13 of
the DSU… With these constraints in mind, we note that
there is no finding nor any ‘legal interpretation
developed by the panel’ that may be the subject of an
appeal of which the Appellate Body may take
cognizance. 49
Thus, the Appellate Body took the approach of not addressing the issues
raised by Brazil that were contained in the Panel Report. However, in
respect of a different claim, it ruled the contrary:
We are aware of the provisions of Article 17 of the DSU
that state our jurisdiction and our mandate. … In certain
appeals, however, the reversal of a panel’s finding on a
legal issue may require us to make a finding on a legal
issue which was not addressed by the panel … And, in
this appeal, as we have reversed the Panel’s finding on
Article 5.1(b), we believe we should complete our
analysis of the c.i.f. import price by making a finding
with respect to the consistency of the EC regulation with
Article 5.5, which was not addressed by the Panel for
reasons of judicial economy.50
Lichtenbaum queries how to reconcile these two conclusions in the same
document: ‘if the lack of a panel finding precludes Appellate Body
jurisdiction, this rule should also apply in the situation where the lack of
a panel finding was due to the application of the judicial economy
approach.’51 As a result, the circumstances in which the Appellate Body
may or will complete the panel’s analysis are uncertain. It may be argued
that this procedural uncertainty diminishes the parties’ rights to defend
their interests in accordance with due process in the WTO dispute
settlement system, as discussed further below.
49
Appellate Body Report, EC – Poultry, para 107.
50
Ibid para 156
51
Peter Lichtenbaum, ‘Procedural Issues in WTO Dispute Resolution’ (1998) 19 Michigan
Journal of International Law 1195, 1270.
206
81
A strict reading of the competences of the Appellate Body may lead to

even greater difficulties for respondents than complainants. If for any
reason the Appellate Body cannot complete the analysis of a defence that
may justify the WTO-inconsistency of the respondent’s measure, the
respondent would find itself in the curious situation in which it would
have to implement recommendations and rulings regarding a measure
for which it considers it has a defence. The absence of a ruling on such a
defence would be due to a mixture of judicial economy and the distribution
of powers between panels and the Appellate Body — matters that are
completely unrelated to the validity of the defence.
(iii) Absence of Appellate Review

Another problem with the Appellate Body completing the panel’s analysis
relates to the lack of further review of the analysis that is completed.
Palmeter considers that the Appellate Body’s analysis in these
circumstances is the equivalent of de novo review, and that the resulting
decisions are ‘unreviewed’ and ‘unreviewable’ due to the absence of
appeal.52 According to this author, ‘[d]e novo decisions of the Appellate
Body lack the primary benefit of appellate review, which is a second,
more focused examination of a contentious issue, by individuals other
than those who made the initial decision.’53
That decisions do not benefit from a potential review is not a sufficient
argument to dismiss the power of the Appellate Body to complete the
analysis. The functions and tasks of panels and the Appellate Body are
different and, given that the Appellate Body has the power to modify a
panel’s reasoning, it may always come up with new considerations.54 Even
in judicial systems where there is remand authority, it is more efficient in
some instances for the higher court to modify the issues undecided (and
‘complete the legal analysis’) rather than to send the case back to the lower
court to be corrected.
2 Leaving Issues Undecided

This is the only feasible alternative for the Appellate Body when
modification or reversal of a panel finding raises issues for which there
52
David Palmeter, ‘The WTO Appellate Body Needs Remand Authority’ (1998) 32 Journal
of World Trade 41, 43.
53
Ibid.
54
The Appellate Body noted this distinction in US – Wheat Gluten , where it said: in view of
the distinction between the respective roles of the Appellate Body and panels, we have
taken care to emphasize that a panel’s appreciation of the evidence falls, in principle, ‘within
the scope of the panel’s discretion as the trier of facts’. Appellate Body Report, US – Wheat
Gluten, para 151 (footnote omitted).
207
82
are no factual findings or factual basis in the panel record. In Canada –

Periodicals, given the absence of necessary factual elements, the Appellate
Body could not proceed to determine whether the investigated products
were like products. In EC – Computer Equipment,55 the Appellate Body
reversed the panel’s legal conclusion that ‘legitimate expectations’ of an
exporting Member are relevant for interpreting the terms of a tariff
Schedule and therefore for assessing the violation of GATT Article II:1.56
The Appellate Body did not look further into the panel record to determine
whether it could uphold the panel’s final conclusion for different reasons.
The Appellate Body left the substantive issue undecided, with an
unsatisfactory result for the complainant.
This alternative seems to lead to a denial of justice for parties to a dispute,
meaning that complainants would ‘need to bring the entire dispute back
to the panel level to have this issue settled, starting again at the beginning
of the dispute settlement process’,57 and that respondents might face a
recommendation to bring their measure into conformity even though a
defence may have been validly raised. These results do not seem consistent
with the objective of achieving positive and meaningful solutions to trade
disputes.
III EVALUATING THE POSSIBILITY OF REMAND AUTHORITY FOR THE

APPELLATE BODY
Arguments in favour of remand authority for the Appellate Body typically
focus on considerations of two kinds: (a) due process and fairness; and
(b) the proper limits to appellate review. Arguments against remand
authority are typically focus on: (a) delays in the settlement of disputes;
and (b) the institutional implications for the WTO – in particular, resource
implications. I consider these various arguments below.
A Due Process and Fairness

The Appellate Body has repeatedly invoked due process as a principle
informing the DSU.58 In India – Patents (US), the Appellate Body pointed
out that claims and facts should be raised by the parties from the very
beginning, as a matter of due process, implicit in the DSU.59 Likewise, in
that report the Appellate Body related due process to the timely provision
55
Appellate Body Report, EC – Computer Equipment.
56
Ibid para 97.
57
Andrew Shoyer and Eric Solovy, ‘The Process and Procedure of Litigating at the World
Trade Organization: A Review of the work of the Appellate Body’ (2000) 31 Law and Policy
in International Business 677, 690.
58
See eg Appellate Body Report, Mexico – Corn Syrup (Article 21.5 – US), para 107.
59
Appellate Body Report, India – Patents (US), para 38.
208
83
of accurate information and the need for parties to have an opportunity

to defend themselves. The Appellate Body pointed out that: ‘the dictates
of due process could better be served if panels had standard working
procedures that provided for appropriate factual discovery at an early
stage in panel proceedings.’60
This underlying connection between due process, timely information, and
the opportunity to present a defence, has been reaffirmed in many instances.
Thus in EC – Computer Equipment, the Appellate Body stated that: ‘[a]s the
ability of the European Communities to defend itself was not prejudiced
by a lack of knowing the measures at issue, we do not believe that the
fundamental rule of due process was violated by the panel’.61 In Australia –
Salmon, the Appellate Body said that due process ‘entails providing the
parties adequate opportunity to respond to the evidence submitted’62 and
made the further point that a ‘fundamental tenet of due process is that a
party be provided with an opportunity to respond to claims made against
it’.63 In Mexico – Corn Syrup (21.5.), the Appellate Body referred to the concept
of due process in describing the obligation of panels ‘to address issues that
are put before them by the parties to a dispute’.64
Based on the foregoing, it appears that the concept of due process requires
parties to have an opportunity to be heard and present a defence and
requires panels and the Appellate Body to address the issues raised by
the parties. It follows that parties to disputes ought to have an adequate
opportunity to defend their positions when the Appellate Body has to
complete the analysis. At present, this may not necessarily be the case in
all instances.
Remand authority would provide address these concerns of due process
and fairness. If an issue not addressed at the panel stage had to be resolved
to settle the dispute, the Appellate Body could refrain from completing
the analysis and instead send the case back to the panel in order to give
the parties the right to defend themselves. The Appellate Body would
not feel compelled by duty to complete the analysis.65 Thus, remand
authority would: (a) permit the exercise of judicial economy by panels,
where appropriate; and (b) minimise procedural unpredictability for the
parties with respect to the issues on appeal. From a procedural standpoint,
remand authority would address the current problems of the Appellate
Body practice of completing the analysis.
60
Ibid, para 95.
61
Appellate Body Report, EC – Computer Equipment, para 70.
62
Appellate Body Report, Australian – Salmon, para 272.
63
Appellate Body Report, Australian – Salmon para 278.
64
Appellate Body Report, Mexico – Corn Syrup (Article 21.5 – US), para 36.
65
See above nn 31 and 34.
209
84
B Scope of Appellate Review

As previously indicated, a problem with the Appellate Body completing
the panel’s analysis is that the decisions made by the Appellate Body on
new issues are not subject to review.66 If the Appellate Body had remand
authority, it would not need to undertake de novo reviews and thus make
non-reviewable decisions on new issues. Remand authority would also
minimise the Appellate Body’s involvement in issues of fact, exceeding
its express authority. It could thus reinforce the Appellate Body’s function
in dealing with issues of law, preserving the panel as the trier of facts, as
emphasised by the Appellate Body in US – Wheat Gluten.67
C Delays in Dispute Settlement
1 Implications for the Parties

Perhaps the most important issue in the discussions for allowing remands
in the context of the DSU is the likely delays that remand would impose
on all Members.68 Palmeter recalls that one of reasons for the absence of
an Appellate Body remand authority is ‘the perceived need during the
Uruguay Round to fit the entire dispute settlement process into the time-
frame of Section 301 of the United States Trade Act of 1974’.69
For complaining Members, and particularly for their constituents affected
by the measures at issue, delays caused by remands could increase the
injury caused by the challenged measure. Consequences of delays may
be aggravated by the lack of any interim relief in the WTO dispute
settlement system, the impossibility of taking retaliatory actions until
authorisation from the Dispute Settlement Body (DSB) is obtained (after
adoption of the panel report and expiry of the relevant period for
implementation), and the fact that the system does not provide
retrospective compensation for damage already caused.70
For defendant countries, remands could be used as a litigation technique
to enable WTO-inconsistent measures to be maintained for longer. The
possibility of remand could even induce defendants to prompt procedural
failures in order to force remand. If remands were used as litigation
techniques rather than to achieve the prompt settlement of disputes, they
could be contrary to the object and purpose of the DSU: ‘[t]he procedural
66
Palmeter, above n 52, 43.
67
See above n 59.
68
Michael Hathaway, ‘Commentary on “The Appellate Body”‘ (2000) 31 Law and Policy in
International Business 697, 699.
69
Palmeter, above n 52, 43.
70
Article 3.7 of the DSU; Petersmann, above n 3, 42.
210
85
rules of WTO dispute settlement are designed to promote, not the

development of litigation techniques, but simply the fair, prompt and
effective resolution of trade disputes’.71
At first glance, remand authority would undermine the objective of
prompt settlement of disputes. However, this argument is not totally
persuasive, as remand would avoid the reinitiation of dispute settlement
procedures and, therefore, could abbreviate the total length of a dispute.
In other words, at present, if the Appellate Body cannot address some of
the claims validly raised by the parties before the panel, it ought to leave
the issue undecided, obliging the complainant, if it still interested in
asserting its rights, to restart the whole dispute settlement proceedings.
For a complainant, the delay in this case could well be greater than if
remands were possible.72 As a result, although remand authority may
represent delays from a short-term perspective, it may also be seen as a
useful tool for saving time from a longer-term view-point.
2 Attenuation of Delays with other DSU Reforms

Delays in proceedings that remands could cause may be minimised by
other kinds of DSU reforms. We refer to the implementation of provisional
measures and the modification of the effect of the dispute settlement
decisions from an ex nunc (prospective) to an ex tunc (restrospective)
perspective.
The possibility of remands, if implemented together with the possibility
of provisional measures, would minimise the nullification or impairment
of benefits that might be brought about any procedural delays. The
relevant issue would be to determine the stage at which such measures
should be implemented. They could be implemented after the circulation
of the Panel Report to Members, or when the panel is convinced that
there is sufficient evidence to presume nullification or impairment.
However, this last proposal would impose a duty on panels or the
Appellate Body to clearly define uniform rules for making this assessment.
As recalled by Petersmann: ‘[i]n GATT dispute settlement proceedings
under Article XXIII, the claimant states have regularly requested only
withdrawal of the illegal act in an ex nunc manner, without demanding
71
Appellate Body Report, US – FSC, para 166. It is likely that the possibility of remands will
encourage WTO Members to rely on more private lawyers. In fact, this could lead to greater
clarification and evolution of WTO law
72
Of course, were the Appellate Body to leave an issue undecided, some of the proceedings
would not last as long as they did the first time. For instance, consultations or the
establishment of the panel and the term of reference could be abbreviated. The closer the
error is located to the beginning of the WTO dispute settlement procedure, the longer the
remanded case would last: Palmeter and Mavroidis, above n 3, 31.
211
86
reestablishment of the status quo ante or of the situation that would have
existed in the absence of the illegal act.’73 According to this author, the
only exception to this rule was five antidumping and countervailing duties
cases where the panels recommended not only the removal of the five
GATT inconsistent measures but also the reimbursement of the duties to
importers. This author concludes that ‘GATT law seems to recognise only
a limited legal basis for an international right to reimbursement of certain
illegal duties.’74
This practice appears to prevail in the WTO context, on the basis of Article
19.1 of the DSU, under which panels and the Appellate Body typically
recommend that Members ‘bring into conformity’ with the relevant WTO
agreement any measure found to be inconsistent. Under the SCM
Agreement, however, the practice appears to provide an exception to this
general rule. According to Article 4.7 of the SCM Agreement, prohibited
subsidies must be withdrawn without delay. In Australia – Automotive
Leather II (21.5.) a subsidy was granted by the Australian Government to
domestic producers and exporters of automotive leather. The subsidy was
granted only once, prior to the initiation of dispute settlement proceedings.
Applying the general rule in Article 19.1 of the DSU and the ex-nunc effect
from the GATT and WTO practices would have provided no significance
or effect utile to Article 4.7 of the SCM Agreement. Accordingly, the panel
found that the obligation to withdraw a subsidy under Article 4.7 is not
limited to prospective action;75 rather, Article 4.7 of the SCM Agreement,
in this case, required repayment of the prohibited subsidy.76 This Panel
report was not appealed. Given the absence of subsequent jurisprudence
on prohibited subsidies that no longer exist at the time of the final rulings
and recommendations in a WTO dispute, the question of ex tunc effects in
relation to such subsidies remains uncertain.
In Guatemala – Cement II, the panel denied Mexico’s request concerning
the reimbursement of antidumping duties because it considered that the
issue had not been fully explored in the dispute. Unlike the previous case,
the request of Mexico was based on Article 19.1 of the DSU, which is
recognised as establishing the ex nunc effect:
[W]e note that Guatemala has now maintained a WTO-
inconsistent anti-dumping measure in place for a period
of three and a half years … Mexico’s request raises
73
Ernst-Ulrich Petersmann, ‘International Trade Law and the GATT/WTO Dispute Settlement
System 1948-1966: An Introduction’ in Ernst-Ulrich Petersmann (ed), International Trade
Law and the GATT/WTO Dispute Settlement System (1997) 5, 42.
74
Ibid 43.
75
Panel Report, Australia – Automotive Leather II, para 6.39.
76
Ibid para 6.45.
212
87
important systemic issues regarding the nature of the

actions necessary to implement a recommendation
under Article 19.1 of the DSU, issues which have not
been fully explored in this dispute. Thus, we decline
Mexico’s request to suggest that Guatemala refund the
anti-dumping duties collected.77
DSU reform to enable decisions of WTO panels and the Appellate Body
to have ex tunc effect would oblige Members to withdraw WTO-
inconsistent measures or to offer compensation for the time they have
been in place. Losing parties would therefore have to restore the situation
to that existing before the nullification or impairment. This kind of reform
would create incentives for overcoming the fears of delays in dispute
settlement proceedings, which would be more acute if remand were
introduced.
D Resource Implications for the WTO Adjudicating Bodies

For panels and the Appellate Body, remands could require major
additional resources in terms of time and personnel. However, an overall
assessment would also indicate significant benefits. For the Appellate
Body, with remands the workload could decrease. The Appellate Body
could avoid completing the analysis if it could refer the issue back to the
panel. In addition, the Appellate Body would not need to delve into factual
issues in order to advance a prompt settlement of a dispute. Remands
could thus promote expeditious appellate review.
For panels, the introduction of remands could raise important challenges.
The DSB would probably receive the remanded case from the Appellate
Body and then reconvene the original panel to the extent possible. The
first issue this raises is whether the original panelists would be willing to
re-examine the remanded file, taking account of the fact that the current
panel system is ad hoc. Panelists are appointed on a case-by-case basis
and are not remunerated as permanent WTO staff members. Under this
system, one might expect that it would be difficult to find specialists who
would agree to serve as panelists without a clear limit on the time required.
Article 21.5 of the DSU provides some guidance as to the difficulties that
may arise with recalling original panelists. This provision may be regarded
as an exception to the current ad hoc character of the panel system. It
provides:
Where there is disagreement as to the existence or
consistency with a covered agreement of measures
77
Panel Report, Guatemala – Cement II, para 9.7.
213
88
taken to comply with the recommendations and

rulings such dispute shall be decided through recourse
to these dispute settlement procedures, including
wherever possible resort to the original panel…78
Although the words ‘wherever possible’ indicate that the original panelists
are not necessarily bound to serve in a Article 21.5 panel, of the more
than twenty Article 21.5 panels established to date, in only four have any
of the original panelists had to be replaced.79 Moreover, in two other cases,
the same panelists have served on two successive Article 21.5 panels in
relation to the same dispute.80
This evidence suggests that panelists under the current ad hoc system
would often be willing to get involved again if the Appellate Body
remanded their original case. Accordingly, it would seem unnecessary to
change the current ad hoc panel system to a system of permanent panelists
purely to accommodate a new remand procedure.
Although remand authority is sometimes conceived as a procedural tool
to permit panels to exercise judicial economy,81 it may contrarily create
incentives for panelists to address all claims raised by the parties in order
to avoid the possibility of remands. This could be seen as a negative,
over-cautious and time-consuming approach. In fact, seen in a more
positive light, remand authority could induce ad hoc panelists to be more
careful with their findings and the use of judicial economy in order to
avoid remands. They could refrain from the use of judicial economy in
instances when it might be risky, in order to provide the Appellate Body
with all the necessary elements to resolve the dispute.
E Possible Alternatives to Remand Authority

There are good arguments for considering remand authority as an
alternative to deal with the problems arising from modifications and
reversals of panel’s findings and conclusions. Other alternatives for reform
could include enlarging the task of panels and reducing their use of judicial
economy. However, this would increase the burden of panel work. Another
alternative would be to expand the Appellate Body’s powers of review
so as to allow it to delve into the facts of a case and make factual findings
on its own. However, this would obviously increase the workload of the
Appellate Body.
78
Emphasis added.
79
These cases were: Mexico – Corn Syrup (Article 21.5 – US); Canada – Dairy (Article 21.5 –
New Zealand and US); EC – Bed Linen (Article 21.5); Canada – Aircraft (Article 21.5).
80
Brazil – Aircraft and EC – Bananas.
81
Palmeter and Mavroidis, above n 3, 149–150.
214
89
In the negotiations regarding reform of the DSU, it may be considered

that the Appellate Body should advise the parties at the outset of its
intentions in relation to matters on which it may be required to complete
the analysis. Thus, one way for the Appellate Body to ensure that parties
have an opportunity to present their views on new issues it intends to
address would be to notify its intention to analyse legal issues not
developed by the panel. This would provide the parties with an
opportunity to submit their positions and would improve predictability
in appellate review. However, the Appellate Body might be reluctant to
disclose in advance its intention concerning the legal issues raised on
appeal. In addition, this could be burdensome for the parties and increase
the workload of the Appellate Body.
IV CONCLUSION
Two Members have proposed remand authority in the current DSU reform
process:82 the European Communities and Jordan suggest amending the
powers of the Appellate Body under Article 17.12 of the DSU so as to
allow remands. Both proposals highlight the need for remand when the
factual basis at the panel stage prevents the Appellate Body from
performing its task and emphasise the guidance that the Appellate Body
could provide to the panel for the remand. However, Jordan’s proposal
contemplates automatic remands by the Appellate Body to the DSB,
whereas the European Communities’ proposal contemplates remands only
upon request by a party when the relevant Appellate Body Report is before
the DSB for adoption.
Apart from the constraints imposed by the current DSU rules on the
powers of the Appellate Body, remand authority appears to be a feasible
alternative to preserve the distribution of competence between panels
and the Appellate Body, to preserve the practice of judicial economy, and
to provide higher due process guarantees to the parties to WTO disputes.
The experience with Article 21.5 procedures suggests that the current ad
hoc panel system does not impose an insurmountable obstacle to the
introduction of remand proceedings.
82
Communication from Jordan, Jordan’s Further Contribution Towards the Improvement
and Clarification of the Dispute Settlement Understanding, TN/DS/W/56 (19 May 2003);
Communication from the European Communities, Contribution of the European
Communities and Its Member States to the Improvement and Clarification of the WTO
Dispute Settlement Understanding, TN/DS/W/38 (23 January 2003).
215
90
Journal of World Trade 38(5): 757±772, 2004.
# 2004 Kluwer Law International. Printed in The Netherlands.
Third Party Rights and the Concept of Legal

Interest in World Trade Organization Dispute
Settlement: Extending Participatory Rights to
Enforcement Rights
Ngangjoh H. YENKONG*
I. INTRODUCTION
On 28 May 2004, the Chairman of the World Trade Organization (WTO)

Dispute Settlement Body (DSB) Special Session issued a report to the Trade
Negotiations Committee (TNC) confirming the failure of Members to reach a
consensus on the Doha Mandate on the negotiations on improvements and
clarifications of the Dispute Settlement Understanding (DSU) by the end of May
2004, as was formally scheduled. The Chairman's report to the TNC also
unambiguously confirms that the negotiations on the clarifications and
improvements of the DSU as provided in paragraph 30 of the Doha Ministerial
Declaration would again be postponed for a future date.
One of the most contentious issues leading to the deadlock in the negotiations
(although not exclusively), among the delegations is on the form the DSU remedies
should take. Some Members are of the opinion that retaliation, which is the remedy of
last resort,1 should be retroactive and be tradable,2 while others are of the opinion that
retaliation should take the form of United Nations Chapter VII Mandate, where all
other Members shall be allowed to retaliate so as to maintain a level playing field within
the system and to preserve its legitimacy.3 Some Members are sceptical about these two
views.4 Some of the proposals also touched on the enhancement of third party rights.
With regard to the latter proposals, it is not very clear as to what extent such third party
rights should be enhanced, although the issue of third party participation in dispute
settlement not only goes far beyond mere participation but also has great implications as
regards the legitimacy of the system. By taking inspiration from the concept of
reparations developed by the Permanent Court of International Justice (PCIJ) in the
* Visiting Scholar at the World Trade Institute-Bern and Research Fellow at the Faculty of Law, University
of Helsinki. I am grateful to Professor Thomas Cottier, Dr Edwini Kessie, Dr Matthias Oesch, and Fernando
Pierola for sharing with me their views on this topic. The views expressed in this article are strictly personal and do
not represent those of the organizations to which the author is associated. E-mail: <hodu.ngangjoh@helsinki.fi>.
1 DSU Article 22.1.
2 Dispute Settlement Body (DSB) Special Session, 13±15 November 2002, Proposal by Mexico, TN/DS/M/
6, 3 (31 March 2003).
3 Dispute Settlement Body (DSB) Special Session, Negotiations on the DSU, Proposal by the Least
Developed Country (LDC) Group, TN/DS/W/17 4 (9 October 2002).
4 See for instance comments on the proposals by Mexico to the DSB Special Session, TN/DS/M/6, 6±27.
91
Chorzow Factories dispute, this article examines the importance of third party and co-
complainant participation in the WTO dispute settlement enforcement regime. To
highlight the development of third parties' rights in international disputes settlement,
the second part of this article examines the role played by third parties in the
enforcement of the judgments of the International Court of Justice (ICJ) and other
tribunals. Section III focuses on the role of third parties within the WTO enforcement
regime and section IV critically examines the concept of legal interest before Panels and
or the Appellate Body (AB), and sets forth suggestions for improving the practical
enforcement of the WTO rules. Section V contains some concluding remarks. In some
of the explanations, analogies have also been made to the 2001 International Law
Commission (ILC) Articles on State Responsibility.
II. PRACTICES IN OTHER TRIBUNALS
It is a well-established practice under the ICJ as well as other international tribunals

that individual States that have not suffered a ``direct injury'' in a particular dispute may
ground a claim in a broad concept of legal interest and be bound by the decision of the
court.5 Apart from being bound by the decision of the Court, such States may impliedly
be able to seek reparation. The participation of third States in disputes is provided under
the Statute of the ICJ as well as was that of the Permanent Court of International Justice
(PCIJ).6 If the case concerns the construction of a convention to which the intervening
State is a party, then its right of participation cannot be disputed.7 This issue was
contested by the Union of South Africa in the South West Africa case.8 Here, the Union of
South Africa argued that Ethiopia and Liberia had no locus standi in the case, as the dispute
did not affect any material interests of the applicant States or nationals of both Ethiopia
and Liberia. In rejecting the arguments of the Union of South Africa, the court held that:
``For the manifest scope and purport of the provisions of the Article indicate that the Members
of the League were understood to have a legal right or interest in the observance by the
Mandatory of its obligations both toward the inhabitant of the Mandatory Territory and towards
the League of Nations and its Members.''9
5 See Ian Brownlie, Principles of Public International Law (5th edn, Oxford: Clarendon Press, 1998), pp. 469±
476. See also Article 63.2 of the Statute of the ICJ.
6 When we try to establish a case law argument for third states remedies in international law, we would often
rely upon the landmark Wimbledon case in which the Permanent Court of International Justice (PCIJ) allowed third
parties members of a multilateral treaty to intervene (SS Wimbledon, UK, France, Italy, Japan v. Germany, PCIJ 1923
(Ser. A), No. 1, 21). The PCIJ in this case accepted Poland to intervene, though it had no interest in the case.
Poland's basis for intervention rested on the fact that it was a party to the Treaty of Versailles, which had been
violated by Germany. The case provides some support for the application of collective countermeasures in
international law. For further discussion, see Brownlie, as note 5 above, at 471.
7 See ICJ Article 63. Article 62 also provides for a situation where a non-party to a particular treaty or
convention may participate in a dispute concerning such convention or treaty.
8 South West Africa cases (Ethiopia v. South Africa, Liberia v. South Africa), Preliminary Objections, Judgment of
21 December 1962, ICJ Reports (1962).
9 ICJ Reports, 343 (1962). In his separate opinion, Judge Jessup also pointed out that `Ìnternational Law has
long recognised that States may have legal interests in matters which do not affect their financial, economic or other
material or say, physical or tangible interests''. See ICJ Reports, 424±433 (1962).
92
THIRD PARTY RIGHTS IN WTO DISPUTE SETTLEMENT 759
Having rejected the arguments of South Africa, by a slight majority, the court
found that it had proper jurisdiction to rule on the merits of the case. Interestingly,
neither of the applicant states had argued that it had suffered material injury. Rather, the
applicants successfully argued that as former members of the League of Nations, they
had right under the mandate to sue for performance. But this view was rejected during
the second phase of the dispute as the composition of the court changed.
However, the issue of intervention by states not directly injured by a particular
treaty violation has not been easy to settle. The ICJ has rather construed `èvery state . . .
has the right to intervene in the proceedings'', as provided for in Article 63.2 of the
Statute of the ICJ narrowly.10 The court has been reluctant to grant locus standi to third
parties in several cases before it.11 In the South West Africa case, the court found that the
issue of legal interest was a question to be decided on merits as opposed to admissibility
upheld in the preliminary Objections.12 The majority of the members of the court in
1966 found that the issue of legal interest was not to be interpreted in a teleological
manner. As a matter of interpretation, individual States only had a legal interest in
respect of certain provisions of the mandate.13 It is worth pointing out here that this
narrow approach to the concept of legal interest has been widely criticized, as the court
observed a more broad approach in later cases.14
In the context of a multilateral standard-setting treaty such as the World Trade
Organization, it seems obvious that being a Member to the Marrakech Agreement would
suffice to have locus standi in any case concerning the invocation of the Agreements or any
of its related covered agreements without showing that it has suffered any material
injury.15 At the regional level, the European Commission and the European Court of
Human Rights have pointed out that public rights of Europe have consequences in the
procedural sphere. Thus, by implication, locus standi cannot be refused in a particular case
just because the intervening party has not suffered a material injury.16
10 See, for instance, C.M. Chinkin, Third Party Intervention Before the International Court of Justice, 80 A.J.I.L.
(July 1986), pp. 495±531; Jonathan I. Charney, Third State Remedies in International Law, 10 Michigan J. Int'l L.
(Winter 1989), p. 57; Ian Brownlie, The Rule of Law in International Affairs: International Law at the Fiftieth Anniversary
of the United Nations (The Hague: Martinus Nijhoff, 1998), pp. 74±78.
11 Malta was refused right to intervene in the Continental Shelf (Tunisia v. Libya Arab Jamahiriya), ICJ, 18, 91,
94 (1982). A similar result was witnessed when Italy requested intervention in the Continental Shelf (Libya Arab
Jamahiriya v. Malta), ICJ 3 (1984)ÐApplication by Italy for Permission to Intervene.
12 ICJ Reports, 36±38 (1962).
13 Relevant to this view is ICJ Reports in the cases of Nicaragua v. United States, ICJ Reports 6 (1959); the
Northern Cameroons Case, ICJ Reports, 33±34 (1963). See also Brownlie, as note 5 above, at 74±78.
14 In addressing the issue of locus standi in the Barcelona Traction case, the ICJ affirmed a rather different
category of obligation. In the view of the court, ``. . . an essential distinction should be made between the
obligations a State towards the international community as a whole, and those arising vis-aÁ-vis another State in the
field of diplomatic protection. By their very nature, the former are the concern of all States. In view of the
importance of the rights involved, all States can be held to have a legal interest in their protection; they are
obligations erga omnes.'' See Barcelona Traction, Light and Power Company, Limited (Second Phase), 32 (Judgment of
5 February 1970). See also the case concerning East Timor (Portugal v. Australia) (30 June 1995).
15 See Charney, as note 10 above, at 72±75; and Brownlie, as note 5 above, at 75±76.
16 In the case of Loizidou v. Turkey, the European Court rejected an argument on behalf of Turkey that the
regime of reservations to acceptances of jurisdiction should be the same as that of the International Court. See
European Court of Human Rights, Series A, Judgments and Decisions, Vol. 310, 29; see also Brownlie, as note 5
above, at 76.
93
It appears that, incrementally, there is a progression towards the notion of actio

popularis. In the wake of this development, there may be no limitation of material losses
to third State remedies under multilateral agreements such as the WTO. It seems the
most significant support of this view is found in the work of the International Law
Commission on the Law of State Responsibility. Any State other than the directly
injured State is allowed to invoke the responsibility of another State and request for
cessation and ``full'' reparation in this respect.17 However, the ILC has varied its
approach on the nature of third State remedies. Just before the Commission completed
its work, James Crawford made several proposals in 2000 on third State enforcement,
which nevertheless was received with much scepticism by other members of the
Commission.18 As mentioned elsewhere, the ILC Articles were adopted by the
Commission in 2001 and apply to norms established by multilateral treaties or by
customary international law. The Articles apply to the extent where there are no special
rules of international law governing international wrongful conduct of States and there
is no particular specification that a State that has been injured by a violation of a norm of
customary law or a particular convention has a tangible injury.19 By the same token,
does it not flow from this that any WTO Member can invoke the responsibility of
another Member for continuously failing to respect its obligation under a particular
WTO covered agreement?
III. THIRD PARTY RIGHTS IN THE CONTEXT OF THE DSU
Article 10 of the DSU provides the possibilities and conditions under which other
WTO Members may participate in disputes related to a dispute formally brought
before the Panel. Such intervenors (third parties) shall address their requests to the
DSB. Third parties have the right to make both written and oral submissions and
would also be allowed to receive the first written submission of the complainants and
the respondents. Pursuant to the provisions of DSU Article 4.11 and GATT 1994
Article XXII, third parties may also make a request to the consulting Members and the
DSB to joint in the consultation process.20 Article 4.11 requires that the requesting
third party should have a ``substantial trade interest'' in the alleged measure. Also,
Article 10.2 relating to third parties' participation at the Panel stage requires instead a
17 See Article 48 of the Draft Articles on Responsibility of State for Internationally Wrongful Acts, adopted
by the International Law Commission at its 53rd session (2001).
18 See the comments of James Crawford on the final Article of the ILC, The ILC's Articles on Responsibility of
State for International Wrongful Acts: A Retrospect, 96 A.J.I.L. (2002), p. 874. See also Martti Koskenniemi, Solidarity
Measures: State Responsibility as a New International Order?, British Year Book of International Law (Oxford: Oxford
University Press, 2002).
19 See Article 55 of the ILC Articles.
20 With regard to consultation, third parties interested in joining the consultation may notify the Consulting
Members and the DSB of their desire to be joined in the consultation within 10 days after the circulation of the
request for consultation. If the parties to the consultations find that the requesting Member's substantial interest is
well founded it will accept the request and notify the DSB of such acceptance. If third party request for consultation
is rejected, such third party may be free to request for consultation under the provisions of GATS Article XXIII
para. 1, or corresponding provisions in the covered agreement concerned.
94
``substantial interest''.21 Pursuant to the provisions of DSU Article 17.4, third

participants may also appear in the appeal stage subject to prior notification to the DSB
and provided that they also indicate clearly their ``substantial interest'' in the case.22
The Working Procedures of the Appellate Body23 contains further rules for the participation
of third parties.
With this right, third parties seek to influence the proceedings and their
outcomes.24 The participation of third parties is important in the course of litigation
because it could serve to highlight a broad range of issues relevant for the security and
predictability of the system, which might have been overlooked by the complainant(s)
or the respondent(s) as the case may be. Thus, third parties help to ``multilateralize'' a
case, highlighting concerns that might otherwise not be brought to the fore by the
complainant(s).25 The encouraged participation of third parties may also deter potential
violators because of the pressure that they may exert by highlighting other broader
implications of a particular nullification or impairment that the complainant itself might
not have noticed.26 And the participation of a third party is also important for
transparency.27 It is, however, important to mention here that third parties are limited
to WTO Members as opposed to non-governmental organizations.28
However, the practice has been so far different. From past experiences, it appears
that the Panel or AB would not be ready to go beyond issues addressed in its terms of
21 It is, however, important to note that Article 10.2 is about ``substantial interest'' and not ``substantial trade
interest''. From this perspective, substantial interest can also be tied to the interest in maintaining the security and
predictability of the WTO, which is the major objective of the DSU. See DSU Article 3.2.
22 Such third party must also have participated in the dispute at the Panel stage.
23 Working Procedures of the Appellate Body, adopted 15 February 1996, WT/AB/WP/1. It has been revised a
number of times, the latest one being in document WT/AB/WP/7 (1 May 2003). See also document WT/AB/WP/
8, which is still under consideration.
24 In a recent case concerning CanadaÐMeasures Relating to Exports of Wheat and Treatment of Imported Grain,
Panel Report, WT/DS276/R (6 April 2004), the Panel decided as follows ``. . . in the specific circumstances of this
case, we consider it appropriate to invite the third parties to address the Panel both in writing and orally, during a
session of the preliminary hearing of the Panel process set aside for that purpose. In this regard, we note that, in the
interest of an expeditious disposition of the preliminary issues raised by Canada, it was necessary to set a tight
deadline for the third parties to provide written submissions on the preliminary issues. As a consequence, the Panel
found it appropriate . . . to give an additional opportunity to the third parties to develop and refine their arguments
during a session of a preliminary hearing of the Panel set aside for that purpose''. See para. 6.6.
25 Lanye Zhu, The Effect of the WTO Dispute Settlement Panel and Appellate Body Reports: Is the Dispute Settlement
Body Resolving Specific Disputes Only or Making Precedent at the Same Time?, 17 Temple Int'l & Comparative L.J. 1
(2003), p. 235.
26 Of course, this is only to the extent that third parties participate only on the side of the complainants.
27 This is because in most cases details about negotiated settlements might only be known by the complainant
and the respondent. Such undisclosed settlements might tend not to be WTO-inconsistent.
28 However, in accordance with Article 13 of the DSU, Panels may accept briefings from non-governmental
organizations. In order to maintain transparency and prompt settlement of disputes, the acceptance of amicus curiae
briefs should be discouraged, though there are some views to the contrary. See, for instance, Michael J. Trebilcock
and Robert Howse, Regulation of International Trade (2nd edn, London: Routledge 2001), pp. 65±66, who argue for
the continuous acceptance of such briefings. See also Nick Covelli and Rajeev Sharma, Proposals for Reform of the
WTO Dispute Settlement Understanding in Respect of Third Parties, 9 Int'l Trade L. & Regulation 1 (January 2003). In
fact, there are fears that accepting briefs from unwarranted sources might not be in the interest of the majority of the
WTO membership since most of those briefs are from sources that represent very different interests. Furthermore,
the criticisms of WTO Members relating to the decision of the AB to adopt a special procedure for receiving amicus
curiae briefs in the European CommunitiesÐMeasures Affecting Asbestos and Asbestos-Containing Products (ECÐAsbestos),
WT/DS135/AB/R, adopted 5 April 2001, clearly demonstrates the reservations of a majority of WTO Members
against the acceptance of such briefs. See comments of Members in the General Council meeting on amicus curiae
briefs, 22 November 2000, WT/GC/M/60 (23 January 2001).
95
reference, even if there are other relevant issues relating to the dispute that have been
addressed by third parties.29 This was the situation in the United StatesÐStandards for
Reformulated and Conventional Gasoline.30 The European Communities acting as one of
the third parties in the disputes argued that the US defence of the Gasoline Rule, which
indirectly allowed another foreign importer, Canada, to define its own baseline,
provided that it imported into the United States 75 percent of the gasoline produced at
that refinery in 1990, was a possible violation of GATT Article I most-favoured-nation
(MFN) principle. The EC requested the Panel to rule on the so-called ``75 percent
Rule'', even if it could no longer be invoked. In requesting the Panel to make a finding
on the 75 percent Rule, the interest of the EC was seemingly to prevent potential ``law
breakers'' as it pointed out that `à determination on this point would help avoid the
inclusion of this kind of measure in future WTO members' legislation''.31
By taking this standpoint, the EC sought to enhance the security and predictability
of the multilateral trading system, which is the main objective of the DSU.32 However,
the Panel in the Gasoline case had not been asked by the disputing parties in the terms of
reference to review this matter, nor was it ready to extend the scope of the terms of
reference. The Panel in a compromising note declined to examine the matters
addressed by the EC for the reason that at the time the terms of reference were drafted,
the 75 percent Rule was no longer in existence and there was no prospect of the
respondent re-introducing it. Does this mean that the Panel may under some
circumstances consider issues raised by third parties that were terminated after the terms
of reference had been agreed? Or whether the Panel would be willing to review similar
issues raised by third parties, if there was the likelihood of it being re-introduced? The
DSU is not very revealing in this respect. Thus, Panels and the AB seem to have a wide
discretion, first, as to whether to accept or reject third parties' participation and, second,
as to whether to review any issue requested by such third parties.
In USÐAnti-Dumping Act of 1916, Japan requested that it be granted enhanced
third party rights to participate in the dispute.33 In particular, Japan requested to receive
all the necessary documents, including submissions and written versions of statements
of the parties and to attend all the sessions of the substantive meeting of the Panel. At
the request of the Panel, both the complainant and respondentÐthe United States and
the ECÐcommented on this request. The EC accepted Japan's request on the
condition that the EC's request of a similar nature in a case initiated by Japan in
document WT/DS162, would also be accepted. On its part, the United States rejected
Japan's request, arguing that the particular circumstances of the case did not warrant an
29 DSU Article 7.2 states that, ``Panels shall address the relevant provisions in any covered agreement or
agreements cited by the parties to the disputes''.
30 United StatesÐStandard for Reformulated and Conventional Gasoline, Report of the Panel, WT/DS2/R (29
January 1996).
31 WT/DS2/R, para. 4.5.
32 DSU Article 3.2.
33 United StatesÐAnti-Dumping Act of 1916, Complaint by the European Communities, Panel Report, WT/
DS136/R, para. 3.9 (31 March 2000).
96
enhanced participation by third parties as opposed to ECÐHormones, where it was

deemed necessary by the Panel to enable it to conduct concurrent examinations of the
claims by the United States and Canada.34 In rejecting Japan's request, the Panel noted:
``. . . while the DSU does not provide for enhanced third party rights, neither Article 10 of the
DSU nor any other provision of the DSU prohibits Panels from granting third party rights
beyond those expressly mentioned in Article 10 (footnote omitted). The Appellate Body in the
ECÐHormones case confirmed that granting enhanced third party rights was part of the
discretion of the Panels under Article 12.1 of the DSU. (footnote omitted) . . . however, the
DSU differentiates in terms of rights between main parties and that this principle should be
respected in order to keep with the spirit of the DSU in that respect. (emphasis added) We
therefore find . . . no reason to grant enhanced third party rights to Japan in these
proceedings.''35
Notwithstanding the approach of the Panel in USÐAnti-Dumping Act of 1916, it is

difficult to refute the assertion that the DSU recognizes in several ways the relevance of
third parties' participation in the WTO disputes. Third parties have the right to
participate in the dispute from the consultation stage to the appeal stage. At the Panel
stage, a third party has the right to be heard by the Panel and to make written
submissions.36 A third party is also entitled to hear the submissions of the disputing
parties to the first Panel meeting and may comment on their submissions and can also
address questions directed to it by the Panel during such sessions.37 These rights may
also be available at the appeal stage, although it is clearly stated that only the parties to
the dispute (complainants or defendants) may appeal.38
Evidence that the Panels and the AB attach importance to the participation of third
parties can easily be ascertained. In CanadaÐWheat, for example, the Panel noted:
``. . . we think it would be incongruous to give the third parties access to the parties preliminary
written submissions without also giving them an opportunity to address the Panel on the issues
discussed therein . . . Since this Panel intends to issue a preliminary ruling before the due date of
the parties' first written submissions, . . . the Panel necessarily needs to receive the third parties'
arguments with respect to those issues before it issues its preliminary ruling.''39
Again, in ChileÐAlcoholic Beverages, in rejecting one of the arguments presented by

Chile,40 the Panel pointed out that, ``. . . the EC as well as the third parties objected that
such a notion was absurd''.41 The question however is whether these third parties by
being heard can also be able to actually obtain a particular remedy say compensation or
34 Para. 6.3 of the Panel Report.

35 WT/DS136/R, 116, paras 6.32±6.36.
36 DSU Article 10.2.
37 See DSU Appendix 3.6 and 3.8.
38 DSU Article 17.4, Rules 27 and 28.1 of the Working Procedure of the Appellate Review, WT/AB/WP/6 (10
April 2003).
39 CanadaÐMeasures Relating to Exports of Wheat and Treatment of Imported Grain, Panel Report, WT/DS276/
R, 113, para. 6.6 (6 April 2004).
40 Chile argued that under its legislation, it is still possible to sell diluted alcohol contrary to the view of the
EC.
41 ChileÐTaxes on Alcoholic Beverages, Panel Report, WT/DS87/R, WT/DS110/R, 188, para. 7136 (15 June
1999). See also Panel Reports in United StatesÐSections 301±310 of the Trade Act of 1974, WT/DS152/R, para. 7.129
(adopted 27 January 2000).
97
retaliation, when a Panel establishes nullification or impairment and makes its

recommendations to efface such inconsistent measures. Or whether by actively
participating in the dispute settlement process, a third party would be able to influence
the implementation of the rulings and recommendations of the DSB?
The DSU leaves it open for any WTO Member, and not only the disputing
parties, to raise questions relating to unimplemented DSB reports in the DSB meeting
following the adoption of such report.42 Despite this, Article 22.2 on a possible
negotiation of a particular remedial action does not seem to accord any place for a third
party. This means that even if the implementing Member fails to comply with the
rulings and recommendations of the DSB, DSU Article 21.6 does not guarantee in any
case an event where a third party may take steps to enforce the rulings and
recommendations of the DSB, even if such third parties' arguments were to have any
influence on the rulings and recommendations of the Panel or AB.
However, in the travaux preÂparatoires of the DSU, some Member States seemed to
have been very concerned about undermining third parties' remedial rights. Elsewhere,
such concern was also a major topic of discussion throughout the work of the
International Law Commission on State Responsibility. In his Preliminary Report on
State Responsibility, for instance, Willem Riphagen suggested that the breach of a
provision `èssential to the accomplishment of the object or purpose of the treaty''
would almost inevitably directly affect all parties to the treaty.43 Presumably, based on
Riphagen's line of thinking, the GATT contracting parties during the negotiating
history of the DSU, thought that expanding third parties' rights in the DSU would be
essential for building confidence in the multilateral trading system.44 Thus, according to
these delegates, original parties to the dispute should not have absolute control over
which provisions of the General Agreement were implicated in a dispute.45 Also, in
recent negotiations on the clarifications and improvements of the DSU, some Members
have raised similar concerns, starting at least with the question of the requirement to be
a third party in a dispute.46
For security and predictability of the system, a logical starting point for the current
negotiations should not only be to allow third parties to attend all substantive Panel
meetings and receive copies of the parties' submissions to the Panel prior to the issuance

43 Willem Riphagen, Preliminary Report on State Responsibility, II Year Book of the International Law
Commission Part 1 (1980), 115, paras 107±109. Riphagen in his preliminary Report on State Responsibility
pointed out here that a multilateral treaty has a character that a breach by one party may entitle another state party
whose material interests are not directly affected to claim from the guilty State a re-establishment ex nunc and
perhaps even a guarantee ex ante.
44 Terence P. Stewart, The GATT Uruguay Round: A Negotiating History (1986±1992), Vol. II (Deventer:
Kluwer Law International, 1993), pp. 2750±2751, 2774±2775. On their parts, Hong Kong, Hungary, Singapore,
and Switzerland in 1990 proposed to the negotiating group that to enable the possibility of meaningful
presentations and transparency in the dispute settlement process, third parties should be granted an unconditional
access to written submissions of the parties during the Panel process.
45 Ibid., at 2751.
46 See Edwini Kessie, ``The Èarly Harvest Negotiations' in 2003'', in F. Ortino and E.-U. Petersmann (eds),
The WTO Dispute Settlement System 1995±2003 (The Hague: Kluwer Law International, 2004), p. 127.
98
of the interim reports as proposed by some Members,47 but also issues addressed by
third parties in their submissions should be reflected in the rulings and
recommendations of the Panel and/or AB.48 It is true from DSU Article 10.4 and
the decision in ECÐBananas, as discussed later, that instead of participating in a dispute
as a third party, which may only entail limited rights in the process, any WTO Member
may become a co-complainant. But for Members with very limited resources,
especially least-developed countries (LDC) and developing countries, it may only be
viable for them to participate as third parties. Thus, expanding third parties' rights to the
extent of influencing the implementation process of DSB rulings and recommendations
would be relevant from a systemic viewpoint. This may further lead to a sort of
threatening posture for potential WTO lawbreakers to rethink possible WTO
inconsistent measures they may be intending to adopt. A winning Member with
relatively weaker economic power would also be able to benefit from the strength of
third parties in the event where it has to resort to retaliation against an economically
stronger implementing Member. Thus, in this context, it could be suggested that third
parties to a particular dispute may also be allowed to request DSU Article 22.6
arbitration to calculate rather retroactively the level of damages it has suffered, and
further DSU Article 22.2 request for authorization to retaliate from the DSB.
It is, however, argued that the essence of the dispute settlement is to settle dispute
between WTO Members as promptly as possible. With this in mind, some Members
are concerned that expanding third parties' rights might tend to delay the process. This
argument might not stand if one calculates the current length of time taken to complete
a particular dispute to the adoption stage and the length of time that a weak
enforcement mechanism as it presently stands might take to induce the implementing
Member to comply with an adopted DSB Report. To highlight the risk of a weak
enforcement mechanism, consider this scenario. Imagine five countries A, B, C, D and
E that are involved in a particular trade dispute. Country A as the complainant against
country B; while countries C, D and E as third parties with substantial interests,
participate on the side of country A. Country A wins the case and country B refuses to
comply because the congress in country B does not want to amend the WTO-
inconsistent legislation. Countries C, D, and E have also suffered damages due to
country B's inconsistent legislation. Countries C, D and E are willing and strong
47 Chairman's Text as of 28 May, Dispute Settlement Body Special Session, TN/DS/9, 5±7 (dated 6 June
2003). In this respect Members have also made proposals relating to an expanded third party rights in the appeal
stage. There seem to be no proposals up to May 2004, relating to the possibility of a third party assisting the
winning Member in forcing the losing Member to implement the recommendations and rulings of the DSB.
48 It is also important from this perspective to take into account the proposals by some Member states to
allow third parties to participate in the appeal stage even if such third parties did not participate at the Panel stage.
The important argument supporting this proposal is that some Members might only realize the implications of a
particular dispute after the issuance of the Panel reports. Thus, allowing them to participate at the appeal stage to
address certain interpretative issues developed by the Panel might be good for the security and predictability of the
system. While it is still very uncertain about the role of precedents in the WTO dispute settlement practice, it is
certainly clear by virtue of Article XVI.1, of the Marrakech Agreement establishing the WTO (the WTO Charter)
that Panels and the AB would hardly making reference to past practices in order to give credence to their particular
line of interpretation.
99
enough to effect retaliation, but are not allowed to do so. Country A requests the DSB
under Article 22.2 to retaliate and is granted such right. Due to the size of country B,
country A's retaliation does not induce country B to comply with the DSB rulings and
recommendations. Would it not also be logical to allow countries C, D and E to
request for Article 22.6 arbitration so that they also can be able to retaliate? Even if
country B complies in response to country A's retaliation, there would still be systemic
problems because it has not complied in relation to countries C, D and E. Although
satisfactory from the perspective of countries A and B, this result would be unfortunate
from the standpoint of those Members who see the WTO rules as a collective good.
IV. THE CONCEPT OF LEGAL INTEREST
The DSU does not say explicitly if a WTO Member should have a legal interest
before bringing a particular case. Article 10.2 talks about a ``substantial interest'' for
interested third parties, without actually defining exactly what could amount to a
``substantial interest''. Should systemic interest be interpreted to mean ``substantial
interest''? Article 10.1 of the DSU states, ``The interests of the parties to the dispute and
those of other Members under a covered agreement at issue in the dispute shall be fully
taken into account during the Panel process.''
Article 21.6 of the DSU further states that, ``. . . The issue of implementation of
the recommendations or rulings may be raised at the DSB by any Member at any time
following their adoption.''49
These two provisions could arguably give a leeway to third parties to challenge
WTO-inconsistent measures as an act beneficial to the collective membership of the
organization. It also demonstrates that a case of a particular violation can be contested
by any WTO Member without necessarily showing its legal interest in the dispute.
Truly, the international law principle of pacta sunt servanda, which is now a part of the
WTO acquis, requires WTO Members to respect and fulfil their obligations in good
faith.50 Thus, all parties to a treaty, as is the case with the WTO treaty, have an interest
in its observance.51 This approach has already been followed by the AB and needs only
further strengthening. The AB in ECÐBananas, explicitly agreed with the view of the
Panel that:
``. . . with the increased interdependence of the global economy, . . . Members have a greater
stake in enforcing WTO rules than in the past since any deviation from the negotiated balance of
rights and obligations is more likely than ever to affect them, directly or indirectly.''52
49 In ECÐBananas, the AB pointed out that WTO Members have broad discretion in deciding whether to
bring a case against another Member under the DSU. WT/DS27/AB/R, paras 88±89.
50 See Ngangjoh H. Yenkong, Pacta Sunt Servanda and Complaints in the WTO Dispute Settlement,
forthcoming Manchester J. Int'l Econ. L. (2004).
51 AB Report, ECÐBananas, 8, para. 16.
52 Ibid., 62, para. 136. See also a previous GATT Panel Report in United StatesÐTaxes on Petroleum and
Certain Imported Substances (USÐSuperfund), L/6175±34S/136 (adopted 17 June 1987), where in citing the decision
in Japanese Measures on Imports of Leather, it was upheld that ``the existence of a quantitative restriction should be
presumed to cause nullification or impairment not only because of any effect it had had on the volume of trade but
100
In ECÐBananas, the AB upheld that the United States could bring a case under
the GATT 1994 notwithstanding the fact that the United States hardly produced
bananas and was not an exporter of bananas. The European Communities argued that
the WTO Agreement must contain a rejection of the requirement of a legal interest or
an acceptance of the notion of actio popularis in order to conclude that the WTO dispute
settlement has set aside the requirement of a legal interest. According to the EC, the
fact that the WTO and its covered agreements are silent about the requirement of legal
interest is not sufficient for the Panel to accord a legal standing to the United States to
participate in the dispute therein.53 Thus in the view of the EC, ``. . . the United States
has no actual or potential trade interest justifying its claim, since its banana production
is minimal, it has never exported bananas, and this situation is unlikely to change due to
climatic and economic conditions in the United States''.54
The AB here observed that GATT 1994 Article XXIII:1 and DSU Article 3.7,
give WTO Members a broad mandate to consider whether to bring a case or not.55 If
the conduct of another Member could frustrate any anticipated gains accruing to it
directly or indirectly as a result of its membership to the organization; or that the
attainment of any particular objective of the WTO Agreement is being impeded, any
WTO Member would have a locus standi before the Panel or AB to contest such a
measure.56 Thus, the Panel or the AB may not enter into any review as to whether a
particular Member contesting a WTO-inconsistent measure has a legal interest to do
so.57 This approach seems to have been endorsed by WTO Members in several cases,
the most recent being USÐSteel Safeguards.58 Here, for instance, Norway and
Switzerland had a trading interest only in one of the steel products each (welded pipe in
cont.
also for other reasons, e.g. it would lead to increased transaction costs and would create uncertainties which could
affect investment plans'' (BISD 31S/113).
53 Ibid., 8±9, paras 16±18. In the view of the EC, the absence of an explicit requirement of legal interest in
the DSU or other WTO covered agreements indicates that general international law must be applied.
54 Ibid., 8, para. 17.
55 This may be compared with the situation that arose in the South West Africa case, where in the 1962
findings, the ICJ concluded that Ethiopia and Liberia, as former members of the League of Nations, had
jurisdiction to bring a suit against South Africa with respect to the latter's failure to respect its obligations under the
mandate. Here the rationale for the court decision was based on the fact that mandate itself permitted any former
member of the League to bring a suit against the mandate in order to enforce the obligations of the mandate. See
South West Africa Cases, Preliminary Objections.
56 ECÐBananas, para. 134.
57 In reconsidering the question of legal interest in MexicoÐAnti-Dumping Investigation of High Fructose Corn
Syrup (HFCS) from the United States (MexicoÐCorn Syrup, Recourse to Article 21.5 by the United States), the AB
concurred, ``Given the `largely self-regulating' nature of the requirement in the first sentence of Article 3.7, Panels
and the Appellate Body must presume, whenever a Member submits a request for establishment of a Panel, that
such Member does so in good faith, having duly exercised its judgment as to whether recourse to that Panel would
be `fruitful'. Article 3.7 neither requires nor authorizes a Panel to look behind the Member's decision and to
question its exercise of judgment.'' WT/DS132/AB/RW, 23±24, para. 74 (22 October 2001).
58 United StatesÐDefinitive Safeguards Measures on Imports of Certain Steel Products, WT/DS248/R, WT/DS249/
R, WT/DS251/R, WT/DS252/R, WT/DS253/R, WT/DS254/R, WT/DS258/R, WT/DS259/R (May 2 2003).
See also the GATT Panel decision in USÐSuperfund, where in recognizing the general interest of the Contracting
Parties in relation to GATT commitments, the Panel defined the objectives of the GATT Articles III and XI as, ``to
protect expectations of the contracting parties as to the competitive relationship between their products and those
of the other contracting parties. Both articles are not only to protect current trade but also to create the
predictability needed to plan future trade.'' Para. 5.2.2.
101
the case of Switzerland) included in the US safeguards measure, but nevertheless in

their submissions pointed out that in its recommendations, the DSB should also take
into account the other measures complained of by the EC and the other co-
complainants, namely Brazil, China, Japan, Korea and New Zealand.59
However, the decisions of the AB in these cases, especially in ECÐBananas, have
significant implications with regard to co-complainants' enforcement rights. First, a WTO
Member may win a case all through the dispute settlement process without certainly
demonstrating a substantial legal interest in the contested measure. So long as a Member
ascertains that a particular conduct may put the attainment of the global welfare objective
of the General Agreement in jeopardy, it would relevantly have a claim before the Panel or
the AB.60 This approach seems to be sanctioned by the fact that the Panels and the AB in
these cases consider the WTO obligations as a sort of ``global common''.61 The fact that
the WTO principal objective seems to be that of trade liberalization in the sense that it
increases the overall global welfare as a result of a better allocation of the world's resources
including that of the violating state, it should be irrelevant to require WTO Members to
prove their substantial legal interest before participating in a case. This view is, however,
highly reflected in the concept of MFN, which is the core of the WTO Agreement.62 As a
consequence, the MFN makes WTO bilateral engagements collective.
The nature of the rulings of the AB in ECÐBananas tells us that the rights and
obligations of WTO Members are interlocked in the negotiated tariff concessions
where all involved have duties to perform and expectations arising. After all,
compliance with WTO obligations leads to the achievement of worldwide welfare.
This seems also to be in line with Article 48 of the 2001 International Law Commission
(ILC) Articles on the Responsibility of States for International Wrongful Acts.63
The second implication of the AB rulings in ECÐBananas is that a party could
succeed in a WTO dispute but would be unable to claim compensation or retaliate as
retaliation is linked to the concept of nullification or impairment.64 The United
States realized this limitation in the DSU when it pointed out that Article 3.8
presupposes a finding of infringement prior to a finding of nullification or
impairment even if no compensation was due.65 With this approach, it may be
59 See Panel Reports, paras 4.7±4.11. Information on the submissions obtained from Matthias Oesch, who
was counsel for Switzerland in the Steel case.
60 The preamble of the Marrakech Agreement Establishing the WTO outlines the basic goals of the WTO
Agreement.
61 See DSU Article 3.8, which considers a presumption of nullification or impairment once a violation has
been found.
62 The MFN provision is well defined under GATT Article I. According to this provision, any trade
advantage a country accords to one country must be multilateralized.
63 According to Article 48 of the ILC Articles, states that are not directly `ìnjured states'' may invoke the
responsibility of another state for committing an international wrongful act.
64 Following the rulings of the AB in ECÐBananas, retaliation or countermeasures could have been possible
for states not ``directly injured'' by a particular violation, were DSU Article 22.4 eliminated. According to this
provision, ``The level of suspension of concessions or other obligations authorized by the DSB shall be equivalent
to the level of the nullification and impairment''.
65 WT/DS27/R/USA, para. 7.48.
102
vividly ascertained that the DSU enforcement practice seems to be deviating from
the provisions of Article 60.2(a) of the Vienna Convention on the Law of Treaties
(VCLT), which provides a leeway to other members to a multilateral treaty to
suspend the operation in part or in whole of the treaty to another member in breach
of that treaty.66 Article 60.2(a) of the VCLT accords a collective enforcement rights
to other state parties to a multilateral treaty even if they did not invoke the dispute in
question.
The DSU, by contrast, provides the possibility to sue by any Member of the
WTO when there is a violation of any of the provisions of the covered agreements
even if such Member has not suffered a ``direct injury''. The DSU does not extend
such rights to the level of enforcement of the rulings and recommendations of the
DSB. However, by providing for such collective enforcement rights in the VCLT,
the route was mapped out clearly for further considerations in the context of state
responsibility.67 The ILC codified the system of collective enforcement in a case of a
violation of multilateral obligations. The Commission's Articles canvassed different
scenarios where the international community should collectively respond to a breach
of a convention. The obligations breached should be owed to the international
community at large and/or an obligation set up for the collective interest.68 If those
two tests can be met, then all parties to the convention are `ìnjured states''. If the
fundamental character of the WTO treaty system can be construed as a ``global
common'', does it not follow from this that a particular breach would require a
collective action to pursue the lawbreaking State?
A contrario, the DSU does not require any trade effect or injury suffered to bring a
complaint before a Panel or AB,69 but rather requires that retaliation should be subject to
an `èquivalent'' amount of nullification or impairment suffered.70 Precedents in the
WTO jurisprudence do not tell very much about the true meaning of `èquivalent''. But,
if we go by the Concise Oxford Dictionary meaning of the term `èquivalent'', retaliation
must be equal in value or have the same effect as that of the nullification or impairment
that has been suffered.71 Up to February 2004, DSU Article 22.6 arbitration on the
calculation of the level of nullification had been resorted to seven times,72 all of them
66 The Vienna Convention was adopted on 22 May 1969 and opened for signature on 23 May 1969 by the
United Nations Conference on the Law of Treaties. In accordance with Article 84.1, it entered into force on
27 January 1980. See United Nations, Treaty Series, Vol. 1155, 331.
67 For further discussion on collective enforcement in the context of the ILC work on State Responsibility,
see Koskenniemi as note 18 above.
68 Article 48 of the ILC Articles.
69 The exception being in the context of non-violation and situation complaints (DSU Article 26 and GATT
1994 Article XXIII(b) and (c)). See, further, United StatesÐMeasures Affecting Alcoholic and Malt Beverages (USÐMalt
Beverages) (adopted 19 June 1992), BISD 39S/206, where legislation was found to constitute a GATT violation
even though it was not being enforced.
71 Judy Pearsall (ed.), The Concise Oxford Dictionary (10th revised edn, London: Oxford University Press,
2001).
72 USÐ1916 Act (WT/DS136/ARB), CanadaÐAircraft II (WT/DS222/ARB), USÐFSC (WT/DS108/
ARB), BrazilÐAircraft (WT/DS46/ARB), ECÐBananas (Ecuador) (WT/DS27/ARB), ECÐHormones (United
States, Canada), (WT/DS26/ARB, WT/DS48/ARB), ECÐBananas (United States), (WT/DS27/ARB).
103
pronouncing the level in monetary terms except in one case.73 The arbitrator in ECÐ
Bananas, observed:
``We agree with the United States that . . . it is the purpose of countermeasures to induce
compliance. But this purpose does not mean that the DSB should grant authorization to suspend
concessions beyond what is equivalent to the level of nullification or impairment. In our view,
there is nothing in Article 22.1 of the DSU, let alone in paragraphs 4 and 7 of Article 22, that
could be read as a justification of countermeasures of a punitive nature.''74
Thus, the DSU rules out any form of countermeasure that would be punitive in
nature. By the same token, it does not even allow Members which have not suffered a
direct nullification or impairment to effect countermeasures, even if they were co-
complainants in the dispute and the ``directly injured'' State is incapable of effecting an
`èquivalent'' countermeasure.
It is an essential principle which seems to be embedded in the framework of
international practices that reparation must, as far as possible, wipe out the
consequences of an illegal act and re-establishes the situation which would, in all
probability, have existed if that act had not been committed.75 With this in mind, the
calculation of the level of suspension of concessions should be done in a manner that
the retaliation would be able to wipe out the effects of a particular nullification or
impairment and then, compensate the winning member for the frustrated legitimate
expectations due under the negotiated tariff commitments.76 Countermeasures in the
WTO only try to fulfil the latter condition, but unsuccessfully. Contrary to the
guiding principle for awarding reparation, established by the PCIJ in the Chorzow
Factory case, the DSU countermeasure only ex nunc. It is worth pointing out that
countermeasures in the WTO are essentially trade restrictive, in the sense that the
implementing member opens up a sector of its market to the winning Member by,
say, reducing the tariffs below that stated in the losing Member's schedule of
commitments.
Having said this, the arbitrators in ECÐBananas rightly pointed out that
countermeasures or suspension of concessions are not in the economic interest of
either party.77 But the rationale for countermeasures is to induce compliance by a
violating Member State.78 This sounds very relevant for the DSU since the primary
goal is compliance.79 But, it is controversial at the same time because it is difficult to see
how a lawbreaking State can be induced by the imposition of a particular level of
retaliation, which must be `èquivalent'' to the level of nullification or impairment.
73 In United StatesÐAnti-Dumping Act of 1916, the arbitrators did not specify any monetary figure on which
the EC was to base its countermeasures. The arbitrators only established certain parameters to be used by the EC in
determining the level of nullification or impairment, namely, final court judgment and settlement award under the
1916 Act, WT/DS136/ARB, paras 5.59, 5.62 (24 February 2004).
74 Para. 6.3.
75 Chorzow Factory case, PCIJ, Ser. A, No. 17, 47 (1928).
76 This is far-fetched in the WTO practice.
77 ECÐBananas arbitration report, WTO document WT/DS27/ARB, para. 2.13.
78 Arbitration report, ECÐBananas, para. 6.3.
104
Thus, to the extent that retaliations that are strong enough to be deterring, this must be
a coincidence that the violating Member perceives the cost of maintaining the
inconsistent measure to be too high to be worthwhile keeping.80 Retaliation in this case
can only be effective when the winning party has a comparable stronger economic
power than the losing party.
It is well documented from the WTO jurisprudence81 and also from empirical
evidence drawn from different economic analogies, that retaliation does not do any
welfare good to the retaliating State or the implementing Member State.82 To this effect,
to discourage retaliation within the context of the WTO would only be by raising
higher the cost of not complying with the WTO obligations. Retaliation as a last resort
remedy in the WTO must thus be strong enough to be deterrent by itself, as the purpose
stands.83 It is also true that Members of the WTO are greatly different in their economic
size and even politically. Thus, combining these differences, a possible solution to ensure
a sort of power balance in terms of retaliation would be to allow co-complainants who
participated in the dispute for systemic reasons to retaliate. For this to happen, it would
be relevant to reconsider the concept of equivalence and who can be allowed to retaliate.
Although there might be some practical difficulties, retaliation could be tied to violation
instead of the present status quo, where it is tied to the level of nullification or
impairment. In this context, such collective enforcement possibility would not only
increase the cost of not implementing the rulings and recommendations of the DSB, but
would also reduce the cost of bringing a case since Members would be motivated to
bring and to consolidate their claims before the Panel and/or the AB. This would also
allow the developing and least-developed country Members that presently find it
financially difficult to pursue WTO-violating Members to do so. By increasing the
effectiveness of retaliatory regime, the level of compliance would also be increased and
thereby reducing the resort to retaliation in itself.
80 The effectiveness of retaliation would depend on the economic and political importance to the country
adopting it. Another important remark here is the fact that for retaliation to be effective, the adopting Member's
market should also be attractive so that the implementing Member would find it difficult to be replaced by
diverting its exports to alternative markets. For a study on economics rationale for retaliation, see Monika Bu Ètler
and Heinz Hauser, The WTO Dispute Settlement System: First Assessment from an Economic Perspective, 16 J.L.E.O.
2 (2000), pp. 503±533. See also the in-depth analysis of Countermeasures provided in Petros C. Mavroidis,
Remedies in the WTO Legal System: Between a Rock and a Hard Place, 11 E.J.I.L. (2000), pp. 763±813.
81 The ECÐBananas case was the most outstanding example, where Ecuador thought it irrational to retaliate
even though it was given the green light to effect cross-retaliation. Also, in BrazilÐExport Financing Programme for
Aircraft, Canada was accorded a retaliatory right of Can$ 344.2 million a year and could not fully retaliate because of
its impact on the Canadian consumers. See Press Release, ``Canada±Brazil WTO Regional Aircraft Dispute: An
Up-date on Compliance Negotiations'', Department of Foreign Affairs and International Trade Canada, at
<www.dfait-maeci.gc.ca> (updated 18 September 2000). See also USÐFSC, where the EC had to delay its
retaliation against the United States until 1 March 2004.
82 See, for instance, Jim Powell, Why Trade Retaliation Closes Markets and Impoverishes People, Cato Policy
Analysis 143 (20 November 1990). Also obtainable at <www.cato.org/pubs/pas/pa-143.html>, last visited 7 June
2004. See also Bu Ètler and Hauser, as note 80 above.
83 See AB Reports, ECÐBananas, para. 6.3.
105
V. CONCLUSION
Although there are disagreements among WTO Members as to what form the
improvements and clarifications to the DSU should take, there is a general consensus
that some of the proposed changes are desirable. It is true that the political reality in a
multilateral setting such as the WTO indicates that accepting most of the suggestions in
this article to strengthen the WTO enforcement system and the remedies it provides,
for instance retroactivity on the calculation of the level of breaches of the WTO rules
and collective retaliation, is still very much uncertain. This is not to say that WTO
Members should be inflexible in making changes to the DSU provisions on remedies.
There is a strong majority and, of course, legitimacy interest in reaching a consensus on
these changes. The WTO rules are gradually evolving as international law obligations
binding on any Member once its accession to the multilateral trading system is finalized.
Thus, the impact of a particular violation is not only felt by governments as political
actors, but also by individuals and companies within the global market space. The only
effective due process guaranteed once this view is accepted by the membership of the
WTO would be to introduce a system of retaliation that would be effected by parties
who participate in the dispute solely for systemic reasons. Such a retaliatory regime
could be necessary for the multilateral spirit required for the survival of the World
Trade Organization.
106
Journal of World Trade 35(5): 825–845, 2001.
© 2001 Kluwer Law International. Printed in The Netherlands.
The Time to Comply with an Adverse WTO Ruling

Promptness within Reason
Pierre MONNIER*
The time period for compliance is the eye in the cyclone of the WTO Dispute
Settlement Mechanism. It starts with the adoption of the panel’s and/or Appellate
Body’s report, when the Dispute Settlement Body (DSB) enjoins the defending Mem-
ber to bring its inconsistent measure into conformity, and opens the door to a second
round of reports (compliance panel and/or arbitration about retaliation).
For example, the right for the complainant to retaliate is clearly linked to the
expiry of the reasonable period of time.1 Moreover, even if the compliance panel of
Article 21.5 of the DSU does not refer explicitly to the reasonable period of time, the
21:5 panel report Australia—Salmon chose the expiry of the reasonable period of time
as a benchmark to assess whether “measures taken to comply existed or will exist in the
sense of Article 21.5”.2
If the concept of compliance period relies mainly on the “Reasonable Period of
Time” set forth in Article 21.3 of the DSU, the concept of “without delay” provided
for prohibited subsidy cases should not be forgotten.3 In both cases, the basic principle
is immediate compliance. For example, Article 21.1 of the DSU prescribes that
“Prompt compliance with recommendations or rulings of the DSB is essential in order
to ensure effective resolution of disputes to the benefit of all Members”, and Article 4.7
of the Agreement on Subsidies and Countervailing Measures (ASCM) enjoins that “If
the measure in question is found to be a prohibited subsidy, the panel shall recommend
that the subsidizing Member withdraw the subsidy without delay”. But this “zero
delay” standard looked probably unrealistic to the negotiators of the WTO agreements
who installed safety-valves, such as “Reasonable Period of Time” in the DSU or
“time-period within which the measure must be withdrawn”, in Part II of the ASCM.
1 * Juge Administratif, France. For publishing reasons this article is accurate as of 31 August 2001. Many
thanks to Gabrielle Marceau and Jean-François Brakeland for their kind comments.
1 Article 22.2 of the Understanding on rules and procedures governing the settlement of disputes (DSU) says
“… within 20 days after the date of expiry of the reasonable period of time, any party having invoked the dispute
settlement procedures may request authorisation from the DSB to suspend concessions or other obligations …”
2 See report Australia—Measures Affecting Importation of Salmon—Recourse to Article 21.5 by Canada, WT/DS18/
RW, para. 7.30, adopted by the DSB on 20 March 2000. But the exiry of RPT seems to function more like a
minimum than an absolute benchmark for compliance panels. For example, the Shrimp compliance panel
considered that, “while mindful of the obligation of the United States to bring its legislation into conformity by
the end of the reasonable period of time, it is consistent with the spirit of Article 3.3 of the DSU to take
into account any relevant facts until the date on which the matter was referred to the Panel.” Also see report
WT/DS58/RW, para. 5.13, dated 15 June 2001, currently under appeal.
3 For the sake of completeness, the six-month period provided by Article 7.9 of the Agreement on subsidies
for actionable subsidy cases should be mentioned, but such a period has never been applied yet.
107
How to be prompt without being unreasonable? A solution to this dilemma could

be found through four fundamental questions:
1) How are compliance periods determined?
2) How long are they?
3) What are the relevant criteria set forth in the case law?
4) What is the relationship, if any, between the “reasonable period of time” and
the “without delay” provided for prohibited subsidy cases?
I. THE DETERMINATION OF THE TIME PERIOD FOR COMPLIANCE

A. THE REASONABLE PERIOD OF TIME (RPT) OF ARTICLE 21.3 DSU
Article 21.3 DSU states:
“The reasonable period of time shall be:
(a) the period of time proposed by the Member concerned, provided that such period is
approved by the DSB; or, in the absence of such approval,
(b) a period of time mutually agreed by the parties to the dispute within 45 days after the
date of adoption of the recommendations and rulings; or, in the absence of such
agreement,
(c) a period of time determined through binding arbitration within 90 days after the date
of adoption of the recommendations and rulings.”
These provisions seem to imply a preferential order between the three paragraphs.
Strangely enough, paragraph (a) was never applied,4 whereas paragraphs (b) and (c)
have been used, respectively, 18 and 13 times (see Annex to this article). However,
during its meeting on 24 July 2001, the DSB approved the US proposals to extend two
RPTs previously settled by arbitrators with regard to the disputes Section 110(5) of the
US Copyright Act5 and Anti-Dumping Act of 1916.6 These DSB agreements could be
seen as an implicit application of paragraph (a).7
The first question arising from the provisions of Article 21.3 is whether there is
any other possibility than those set forth in the three paragraphs to determine the RPT.
The 301 panel reports replied negatively to this question, pointing out that “It is for the
WTO or both of the disputing parties, through the procedures set forth in Article 21—
not for an individual WTO Member—to determine the reasonable period of time for
the Member concerned to implement DSB recommendations and rulings.”8
Nevertheless, this does not seem to preclude a panel, under its power of
suggestions in implementation set forth in Article 19.1 DSU second sentence, to make
4 In Section 110(5) of the US Copyright Act, the 15-month period proposed by the US was not approved
by the DSB (see Document WT/DS160/9, 31 August 2000).

5 See document WT/DS160/14, 18 July 2001.
6 See document WT/DS136/13, WT/DS162/16, 18 July 2001.
7 The request of the US that the DSB approves their proposal to modify the time-period in the FSC case so
as to expire on 1 November 2000 could also be considered as an implicit application of paragraph (a). See
document WT/DS108/11, 2 October 2000.
8 See panel report, United States—Sections 301–310 of the Trade Act of 1974, WT/DS152/R, adopted 27
January 2000, para. 7.38.
108
THE TIME TO COMPLY WITH AN ADVERSE WTO RULING 827
suggestions about RPT. It has indeed occurred once, with the panel India—
Quantitative Restriction suggesting “that a reasonable period of time be granted to India
in order to remove the import restrictions which are not justified under Article
XVIII:B”.9 In this case, however, the panel just mentioned some factors to take into
account without determining any specific RPT.
We can also mention in passing that periods of times have been twice postponed. In
the Canada—Milk case, parties agreed to modify their agreements pursuant to Article
21.3(b) DSU in order to extend the RPT from 31 December 2000 to 31 January 2001.10
In the FSC case, the DSB acceded to the request of the United States to postpone the
time-period from 1 October 2000 to 1 November 2000. In the United States—Copyright
and United States—1916 Act cases, the DSB approved, on 24 July 2001, that the RPTs be
modified so as to expire on 31 December 2001, or on the date of which the current
session of the United States Congress adjourns, whichever is earlier.
The second fundamental question that should be addressed is when Article 21.3
DSU is actionable. It seems that two cumulative conditions must be fulfilled:
– First, the defending Member should have been found guilty under WTO rules.
We can deduce from the 301 dispute that there is no RPT when a Member has
been found compatible with WTO obligation on the basis of undertakings that,
“should they be repudiated or in any other way removed, the findings of
conformity would no longer be warranted”.11 But a successful non-violation
complaint of the type described in paragraph 1(b) of Article XXIII of GATT
1994 seems to give rise to an RPT determination even if, pursuant to Article
26.1(b) DSU, “there is no obligation to withdraw the measure”. This case has
never arisen, but Article 26.1(c) DSU refers explicitly to “the arbitration provided
for in paragraph 3 of Article 21 DSU”. Finally, it can be inferred from certain
provisions of Article 26.2 DSU12 that RPT does not exist for “other situation”
complaints described in paragraph 1(c) of Article XXIII of GATT 1994.
– Second, the inconsistent measure shall not be “no longer in existence”. In its
Bananas Retaliations report, the Appellate Body stated:
“there is an obvious inconsistency between the finding of the Panel that ‘the 3 March
Measure is no longer in existence’ and the subsequent recommendation of the Panel that
the DSB request that the United States bring its 3 March Measure into conformity with its
WTO obligations. The Panel erred in recommending that the DSB request the United
States to bring into conformity with its WTO obligations a measure which the Panel has
found no longer exists”.13
9 See panel report India—Quantitative Restrictions on Imports of Agricultural, Textile and Industrial Products, WT/
DS90/R, adopted 22 September 1999, para. 7.5.

10 See documents WT/DS103/13 and WT/DS113/13, dated 13 December 2000.
11 See note 8, above, para. 8.1.
12 “The procedures of this Understanding shall apply only up to and including the point in the proceedings
where the panel report has been circulated to the Members”.

13 See Appellate Body report, United States—Import Measures on Certain Products from the European
Communities, WT/DS165/AB/R, adopted 10 January 2001, para. 81.
109
On several occasions, the implementation of an inconsistent measure did not give

rise to a RPT because the implementing Member was deemed to have complied
before the adoption of the report(s)14 or soon after.15
A third and more ancillary question, which arises from the determination of RPT,
is to know whether or not different RPTs could coexist for the same incompatible
measure.
Theoretically, nothing would prevent multiple complainants from agreeing
different RPTs with the defending Member or one of the complainants agreeing a
period which differs from the RPT fixed by an award asked by other complainants.
But neither has ever happened. In the cases Canada—Milk (DS103 v. United States and
DS113 v. New Zealand) and Korea—Beef (DS161 v. United States and DS169 v.
Australia), all the parties agreed the same RPT. The United States—Gasoline case is the
only dispute where an RPT was agreed with one complainant (Venezuela) but not the
other (Brazil). In addition, under the procedure set forth by Article 21.3(c),
complainants have always participated jointly in the arbitration procedure16
Moreover, in the India—Patent cases (DS50 v. US and DS79 v. EC), the EC and
India agreed on an RPT expiring on the same day as the RPT agreed by India and the
US in the context of an earlier dispute relating to the same matter. Since the two sets
of reports were adopted by the DSB at different meetings,17 the two RPT were
different; however, they logically expired on the same day.
On the other hand, we can notice that in several cases18 parties have agreed a kind
of timetable, settling different RPTs on a measure-by-measure or step-by-step basis.
But the legal value of intermediate RPTs remains to be seen. Specifically, the right to
retaliate under Article 22.2 before the end of the “last” RPT is questionable. At this
stage, we can simply note, first, that complainants have never tried to challenge an
implementing measure at the end of an intermediate RPT, either under Article 22 or
Article 21.5; and second, that arbitrators have never given such multiple RPTs.
14 See United States—Measure Affecting Imports of Woven Wool Shirts and Blouses from India (DS33), United
States—Imposition of Countervailing Duties on Certain Hot-Rolled Lead and Bismuth Carbon Steel Products Originating in
the United Kingdom (DS138).
15 See United States—Restrictions on Imports of Woven Cotton and Man-Made Fibre Underwear (DS24),
Argentina—Safeguard Measures on Import of Footwear (DS121/123), Guatemala—Definitive Anti-Dumping Measures on

Grey Portland Cement from Mexico (DS156).
16 The five complainants in EC—Bananas (DS27); Canada and the US in EC—Hormones (DS26/48); Japan,
the EC and the US in Indonesia—Cars (DS54/55/59/64); the EC and the US in Korea—Alcoholic Beverages (DS75/
84); Japan and the EC in Canada—Automotive (DS139/142) and US—1916 Act (DS136/162). In Japan—Alcoholic
Beverages (DS8/10/11), notwithstanding that only the US had requested binding arbitration, it was agreed that
Canada and the EC could participate in the arbitration process.
17 16 January 1998 for the “US” panel and appellate reports (DS50) and 22 September 1998 for the “EC”
report.
18 In the cases Argentina—Measures Affecting Imports of Footwear, Textiles, Apparel and other Items (document
WT/DS56/14, dated 7 July 1998), Canada—Measures Affecting the Importation of Milk and the Exportation of Dairy
Products (documents WT/DS103/10 and WT/DS113/10, dated 7 January 2000), and India—Quantitative
Restrictions on Imports of Agricultural, Textile and Industrial Products (document WT/DS90/15, dated 17 January
2000).
110
B. ARTICLE 4.7 ASCM: “WITHOUT DELAY”
We have already noted that Article 4.7 of the Agreement on Subsidies and
Countervailing Measures (ASCM), which is listed as a special or additional rule and
procedure in Appendix 2 of the DSU, states that “If the measure in question is found
to be a prohibited subsidy, the panel shall recommend that the subsidizing Member
withdraw the subsidy without delay. In this regard, the panel shall specify in its
recommendation the time-period within which the measure must be withdrawn”.
Contrary to ordinary cases giving rise to an RPT, the defending Member has no
choice but to withdraw the subsidy. So this time it is up to the panel to specify the time
period even if this recommendation has to be approved by the DSB with the panel
report in order to acquire legal force. In its first Brazil—Aircraft case, the Appellate Body
seemed to express the view that such a specification is an issue of law subject to appeal,
seeing on this occasion “no reason to disturb the Panel’s recommendation that, in this
case, ‘without delay’ means 90 days and, therefore, Brazil must withdraw the export
subsidies for regional aircraft under PROEX within 90 days”.19
II. THE DURATION OF THE TIME PERIOD FOR COMPLIANCE
A. THE THEORY
1. The Reasonable Period of Time
The starting point of the RPT is the adoption of the report(s). Hence the
arbitrator or the parties can either fix a period or a date of expiry of this period: usually
they do both. The only indication of duration in the DSU is 15 months. Article
21.3(c) of the DSU states: “a guideline for the arbitrator should be that the reasonable
period of time to implement panel or Appellate Body recommendations should not
exceed 15 months from the date of adoption of a panel or Appellate Body report.
However, that time may be shorter or longer, depending upon the particular
circumstances”.
These provisions identify two key notions: “guideline” and “maximum”.20 As
Mr Lacarte pointed out in the Hormones award:
“The ordinary meaning of the terms of Article 21.3(c) indicates that 15 months is a ‘guideline for the
arbitrator’, and not a rule. This guideline is stated expressly to be that ‘the reasonable period of time …
19 See Appellate Body report, Brazil—Export Financing Programme for Aircraft, WT/DS46/AB/R, adopted 20
August 1999, para. 194.

20 Also important is the concept of “case-by-case basis” which seems to be derived from the ordinary
meaning of the word “reasonable” as suggested by the Appellate Body in its Hot-Rolled Steel report about the
“reasonable time” under paragraph 1 of Annex II to the Antidumping agreement. It states that “the word
‘reasonable’ implies a degree of flexibility that involves consideration of all of the circumstances of a particular case.
What is ‘reasonable’ in one set of circumstances may prove to be less than ‘reasonable’ in different circumstances.
This suggests that what constitutes a reasonable period or a reasonable time, under Article 6.8 and Annex II of the
Anti-Dumping Agreement, should be defined on a case-by-case basis, in the light of the specific circumstances of
each investigation”. See WT/DS184/AB/R, para. 84, adopted 23 August 2001.
111
should not exceed 15 months from the date of adoption of a panel or Appellate Body report’. In other
words, the 15-month guideline is an outer limit or a maximum in the usual case. For example,
when implementation can be effected by administrative means, the reasonable period of time
should be considerably shorter than 15 months. However, the reasonable period of time could
be shorter or longer, depending upon the particular circumstances, as specified in Article
21.3(c)”.21 (emphasis added)
Some arbitrators insist on the notion of guideline, underlying the fact that the
RPT could be longer than 15 months. For example, Mr Feliciano in the Chilean
Alcoholic Beverages award said:
“What Article 21.3(c) of the DSU provides arbitrators with is a ‘guideline’, not a fixed
command, that the reasonable period should be not more than 15 months from the date of
adoption by the DSB of the pertinent Panel and Appellate Body Reports. Article 21.3(c)
evidently contemplates a case-specific approach and authorizes the consideration of the
‘particular circumstances’ of a given case, which may warrant a longer or shorter period”.22
In the same vein, India—Quantitative Restrictions analysed the 15 months as a
guideline, suggesting that “the ‘reasonable period’ in this case could be longer than
fifteen months”.23
However, the overwhelming majority of arbitrators insist on the fact that
15 months is a maximum because the context of Article 21.3(c) emphasises that
immediate compliance shall be the benchmark. In the recent award Canada—Patent,
Mr Ehlermann summarized the “guideline v. maximum” dilemma:
“when the ‘reasonable period of time’ is determined through arbitration, the guideline for the
arbitrator is that this period should not exceed 15 months from the date of adoption of the panel
report and/or the Appellate Body report. This does not mean, however, that the arbitrator is
obliged to grant 15 months in all cases. Article 21.3(c) makes clear that the ‘reasonable period of
time’ may be shorter or longer, depending upon the ‘particular circumstances’. (…) The
meaning of Article 21.3(c) is elucidated by its context. This context includes the introductory
language of Article 21.3, which recognizes that the question of a ‘reasonable period of time’ for
implementation only comes into play if ‘it is impracticable to comply immediately’; Article
21.1, which stresses that ‘[p]rompt compliance … is essential in order to ensure effective
resolution of disputes to the benefit of all Members’; and Article 3.3, which also recognizes that
the ‘prompt settlement of situations … is essential to the effective functioning of the WTO and
the maintenance of a proper balance between the rights and obligations of Members’. Thus, the
DSU explicitly emphasizes the importance of ‘prompt’ compliance”.24
But the practice shows that the 15 months functions increasingly as an outer limit.
In fact, the 15 months guideline has been progressively translated into a zero delay
benchmark. In addition, statistics show that agreements under Article 21.3(b) are
tending to fall into line with the jurisprudence of arbitrators.
21 See 21.3(c) award EC—Hormones, WT/DS26/15, WT/DS48/13, 29 July 1997, para. 25.
22 See 21.3(c) award Chile—Taxes on Alcoholic Beverages, WT/DS87/15, WT/DS110/14, 23 May 2000,
para. 39.
23 See panel report India—Quantitative Restrictions on Imports of Agricultural, Textile and Industrial Products, WT/
DS90/R, adopted 22 September 1999, para. 7.5.

24 See 21.3(c) award Canada—Term of Patent Protection, WT/DS170/10, 28 February 2001, paras 36–38.
112
2. The Time Period within which a Prohibited Subsidy Must be Withdrawn

The SCM agreement does not give any guideline about the expression “without
delay” of Article 4.7 ASCM. In the Leather panel, Australia argued that, since the
“normal” period of time for implementation of panel decisions under the DSU is
15 months, and time periods in export subsidy disputes are halved pursuant to Article
4.12, a period of seven and one-half months would be appropriate. But the panel
objected:
“Even assuming Australia is correct in its consideration of fifteen months as the ‘normal’ period
of time for implementation of panel decisions, a question we do not reach, we do not agree that
one-half of that period is appropriate in a dispute involving export subsidies. In the first place,
Article 4.12 specifically provides that ‘except for time periods specifically prescribed in this
Article’ the time periods otherwise provided for in the DSU should be halved in export subsidy
disputes. Article 4.7, which provides that the subsidy shall be withdrawn ‘without delay’, and
that the panel shall specify the time-period for withdrawal of the measure in its
recommendation, in our view establishes that the time-period for withdrawal is ‘specifically
prescribed in this Article’, that is, in Article 4 of the SCM Agreement itself. Moreover, we do
not, as a factual matter, believe that a period of seven and one-half months can reasonably be
described as corresponding to the requirement that the measure must be withdrawn ‘without
delay’”.25
B. THE PRACTICE
1. The Reasonable Period of Time
The annexed Table lists the 31 cases where an RPT was established, whether by
agreement or by binding arbitration. The RPT averages 10.6 months, with a slightly
shorter average for agreed RPTs than those granted by arbitrators.
The evolution of the RPT is interesting. The figure at the end of this article
shows that the 15-month guideline was standard until mid-1998 and the Hormones case
(DS26/48), whether under 21.3(b) agreements (US—Gasoline (DS2), Canada—
Periodicals (DS31), India—Patent (DS50)) or under 21.3(c) awards (Japan—Alcoholic
Beverages (DS8/10/11), EC—Bananas (DS27), and EC—Hormones (DS26/48)). A
second set of arbitration was granted in 1998 and 1999 with RPTs shorter than 15
months. This tendency was confirmed by the seven awards granted since 2000. Now, a
WTO Member cannot hope more than eight months for administrative implementing
measures or 12 months for a legislative action.
The figure also shows interactions between 21.3(b) and (c) RPTs. The threat by
complainants to resort to arbitration often forces the implementing Member to accept
an RPT rather than going to an arbitration procedure, which is ever more risky.
25 See panel report Australia—Subsidies Provided to Producers and Exporters of Automotive Leather, WT/DS126/R,
adopted 16 June 1999, para. 10.6.
113
2. The Time Period within which a Prohibited Subsidy Must be Withdrawn

Panels specified the time period within which the prohibited subsidy had to be
withdrawn on five occasions. Four of them26 gave 90 days. But the FSC panel stated:
“Given that implementation of the Panel’s recommendation will require legislative action (a fact
recognized by the European Communities), that the United States fiscal year 2000 starts on
1 October 1999, and that this Report is not scheduled for circulation to Members until
September 1999 (and, if appealed, might not be adopted until as late as early spring 2000), it is
not in our view a practical possibility that the United States could be in a position to take the
necessary legislative action by 1 October 1999. That being so, and acting in good faith, there is
no way that this could be described as a ‘delay’. However, this objective timing constraint
would not be present with effect from the following fiscal year (2001), which commences on
1 October 2000. As this would be the first practicable date by which the United States could
implement our recommendation, it satisfies the ‘without delay’ standard set forth in Article 4.7.
Accordingly, we specify that FSC subsidies must be withdrawn at the latest with effect from
1 October 2000”.27
This reasoning of the FSC panel looks strange for two reasons. First, the panel
specified a date, whereas the SCM agreement is explicitly referring to a time period.28
We have already quoted Article 4.7, but Article 4.10 is even more explicit about the
time period requirement: “In the event the recommendation of the DSB is not
followed within the time-period specified by the panel, which shall commence from
the date of adoption of the panel’s report or the Appellate Body’s report ….” The
specification of a date raises several concerns, in particular the obligation for the panel
to make some assumptions about the date of adoption of the report(s), which depends
on factors out of the panel’s control (such as translation, appeal, and so forth).
Second, we see in the next section that the “first practical date” criterion adopted
by the FSC panel looks very like the “shortest period possible” definition laid down by
arbitrators for a reasonable period of time.
III. THE APPLICATION OF THE 21.3(C) ARBITRATION PROCEDURE

Before entering into the substance of the case law, it is useful to describe the
procedural steps leading to an Article 21.3(c) award.
26 See note 25, above, para. 10.7; Brazil—Export financing programme for aircraft, WT/DS46/R, adopted 20
August 1999, para. 8.5; Canada—Measures Affecting the Export of Civilian Aircraft, WT/DS70/R, adopted 20 August
1999, para. 10.4; Canada—Certain Measures Affecting the Automotive Industry, WT/DS139/142/R, para. 11.7.
27 See panel report United States—Tax Treatment for “Foreign Sales Corporations”, WT/DS108/R, adopted
20 March 2000, para. 8.8.

28 To be fair to the panel, it should be noted that both parties adopted the same “date” approach.
114
A. THE PROCEDURE
We have already pointed out the different conditions that must be met in order to
give rise to an RPT, emphasising in particular that the whole procedure starts from the
adoption of the report(s).
The first procedural step happens at a DSB meeting held within 30 days after the
date of adoption of the report(s), when the Member concerned informs the DSB of its
intentions in respect of implementation of the recommendations and rulings of the
DSB. Usually, the Member announces that it will comply but needs a reasonable
period of time to do so. Pursuant to Article 21.3, “If it is impracticable to comply
immediately with the recommendations and rulings, the Member concerned shall have
a reasonable period of time in which to do so.” It could be logically inferred from
these provisions that a defending Member which neither asks for nor proposes an RPT
under Article 21.3(a) is supposed to have complied immediately or to have given up its
right to one.
Then runs the 45-day period of Article 21.3(b) to achieve a mutually agreed
RPT. Members do not consider the 45-day period binding since most of the
agreements were settled after the expiration of the 45-day period.
If the negotiations between parties have failed, the RPT shall be determined
through binding arbitration within 90 days after the date of adoption of the
recommendations and rulings. It is usually the complainant which requests
arbitration.29 After negotiations between parties, an arbitration is appointed. We can
note that the appointment procedure by the Director-General of the WTO set forth in
footnote 11 of the DSU was applied in the first three arbitrations, i.e. Japan—Alcoholic
Beverages, Bananas and Hormones.
The annexed Table shows that the 13 appointed arbitrators are Appellate Body
Members.30 This practice does not derive from the DSU, whose footnote 13 only
stipulates that “The expression ‘arbitrator’ shall be interpreted as referring either to an
individual or a group”. An Appellate Body Member was appointed even when the panel
report at stake had not been appealed.31 Originally, appointed arbitrators belonged to the
section of the three Appellate Members who dealt with the original case.32 But this
29 But the defending Member can also request arbitration, as Chile did, for example, in its Alcoholic Beverages
case (document WT/DS87/13, WT/DS110/12, 17 March 2000), or the European Communities in the Hormones
case (document WT/DS26/14, WT/DS48/12, 16 April 1998).
30 Ambassador Lafer was appointed by the Director-General in the Hormones arbitration but was unable to
accept his nomination.

31 Mr Beeby (in Indonesia—Certain Measures Affecting the Automobile Industry, WT/DS54/15, WT/DS55/14,
WT/DS59/13, WT/DS64/12, 7 December 1998), Mr Bacchus (in Canada—Patent Protection of Pharmaceutical

Products, WT/DS114/13, 18 August 2000), Mr Lacarte (in United States—Section 110(5) of the US Copyright Act,
WT/DS160/12, 15 January 2001), and Mr Feliciano (in Argentina—Measures Affecting the Export of Bovine Hides
and the Import of Finished Leather, WT/DS155/9, 12 June 2001).
32 Mr Lacarte (in Japan—Taxes on Alcoholic Beverages, WT/DS8/15, WT/DS10/15, WT/DS11/13,
14 February 1997), Mr El Naggar (in European Communities—Regime for the Importation, Sale and Distribution of
Bananas, WT/DS27/15, 7 January 1998 and in Australia—Measures Affecting Importation of Salmon, WT/DS18/9,
23 February 1999), Mr Ehlermann (in Korea—Taxes on Alcoholic Beverages, WT/DS75/16, WT/DS84/14, 4 June
1999), and Mr Feliciano (in Chile—Taxes on Alcoholic Beverages, WT/DS87/15, WT/DS110/14, 23 May 2000).
115
practice seems to have lapsed, since the last three arbitrators did not belong to the section
which heard the appeal.33
Once appointed, the arbitrator draws up a working schedule, fixing a date for
parties to file their written submissions (simultaneously) and for an oral hearing. The
rules of conduct for understanding on the rules and procedures governing the
settlement of disputes apply to the arbitration procedure.34 It is worth noting that the
procedure looks more like an Appellate Body than a panel procedure. For example,
parties are not allowed to ask each other direct questions during the oral hearing, in
addition to which the award remains secret during its translation into French and
Spanish and is communicated to the party only a few hours before its publication on
the public website of the WTO. It is worth noting that this decision is neither adopted
by the DSB nor even put down in its agenda for remarks. All the 21.3(c) procedure is
outside the DSB’s control.
The 90-day period set forth in Article 21.3(c), which runs from the date of
adoption of the recommendations and rulings to the award publications, has never
been respected. The average period is 131 days. Arbitrators cannot be blamed for this
delay as the average working time period is only 34 days. The remaining 97 days are
divided between the consultation period (63 days between the adoption of the
report(s) and the request for arbitration), the arbitrator selection process (24 days) and
the translation phase (10 days).
A few words should be said about the burden of proof. For a long time, it was
supposed that the complainant had to prove “particular circumstances” justifying a
period of time shorter than 15 months, whereas it was up to the defending Member to
demonstrate “particular circumstances” justifying a 15-month period.35 However, in
the Canada—Patent award, Mr Bacchus refuted this theory, stating that
“as immediate compliance is clearly the preferred option under Article 21.3, it is, in my view,
for the implementing Member to bear the burden of proof in showing—‘if it is impracticable to
comply immediately’—that the duration of any proposed period of implementation, including
its supposed component steps, constitutes a ‘reasonable period of time’. And the longer the
proposed period of implementation, the greater this burden will be”.36
This pro rata temporis burden of proof shows once again that the “immediate
compliance” concept has progressively superseded the “15-month guideline” model.
33 Mr Lacarte (in Canada—Certain Measures Affecting the Automotive Industry, WT/DS139/12, WT/DS142/
12, 4 October 2000), Mr Ganesan (in United States—Anti-Dumping Act of 1916, WT/DS136/11, WT/DS162/14,
28 February 2001), and Mr Ehlermann (in Canada—Term of Patent Protection, WT/DS170/10, 28 February 2001).
34 See document WT/DS/RC/1, adopted by the DSB on 3 December 1996.
35 See, for an illustration, what Australia advocated in the award Australia—Measures Affecting Importation of
Salmon, WT/DS18/9, 23 February 1999, para. 11.

36 See Canada—Patent Protection of Pharmaceutical Products, WT/DS114/13, 18 August 2000, para. 47.
116
B. THE JURISPRUDENCE
1. The Definition of the Reasonable Period of Time
The RPT was authoritatively defined by Mr Lacarte in the Hormones award as
“the shortest period possible within the legal system of the Member to implement the
recommendations and rulings of the DSB”.37 Since then, arbitrators have based their
decisions on this definition.
However, Mr Feliciano has pointed out that this definition should not be read in
absolute terms. In the Chile—Alcoholic beverages award, he noted: “the shortest period
of time theoretically possible for the completion of the legislative process … is not the
sole criterion that I should take into account in determining the reasonable period.”38
It is not surprising to observe that this passage was systematically quoted by the
defending Members during the last arbitration.
Of course, the first task of the arbitrator is to know what measure is at stake before
examining which particular circumstances should be taken into account.
2. The Implementing Measure

As Mr Lacarte pointed out in the Hormones award, it is not within the mandate
of the arbitrator “to suggest ways or means to the European Communities to
implement the recommendations and rulings of the Appellate Body Report and
Panel Reports”.39 Mr Bacchus added in the Canada—Patent decision: “My
responsibility does not, however, include in any respect a determination of the
consistency of the proposed implementing measure with the recommendations and
rulings of the DSB. The proper concern of an arbitrator under Article 21.3(c) is with
when, not what.”40
In this particular case, the European Communities claimed that implementation of
the DSB recommendations required a legislative action, and not the regulatory process
proposed by Canada. Mr Bacchus, after noting that the Panel Report had not specified,
or even mentioned, how Canada should implement the recommendations and rulings,
replied to EC argument that “it has been left for Canada to decide what form the
implementing action should take”.41
It would be interesting to see what would happen in a case where the panel
specified or mentioned how the defending Member should implement the
recommendations and rulings. But panels are usually reluctant to go so far, even
under their ability to “suggest ways in which the Member concerned could
37 See note 21, above, para. 26.

117
implement the recommendations” set forth in the second sentence of Article 19.1
DSU.42
Nevertheless, the prerogative of the implementing Member is not without
limit. For example, in the Hormones case, Mr Lacarte refused to grant a further time-
period “to conduct the risk assessment that was required as of 1 January 1995”
because it “would not be consistent with the provisions of the DSU requiring
prompt compliance with DSB recommendations and rulings, nor with the
obligations of the European Communities under the SPS Agreement”. In the words
of Mr Lacarte, “An implementing Member … has a measure of discretion in
choosing the means of implementation, as long as the means chosen are consistent
with the recommendations and rulings of the DSB and with the covered
agreements”.43 In the same vein, Mr Feliciano dismissed Argentina’s argument that
compliance with the DSB recommendations and rulings adopted with the Bovine
Hides Panel Report necessitates the enactment of a statute by the federal legislature
of Argentina, and that a new Resolución General by the administrative branch would
not be sufficient.44
Then, arbitrators “need to look first at the type of measure proposed to be used
for implementation”.45 The jurisprudence draws a line between legislative and
administrative actions.46 As Mr Bacchus noted in the Canada—Patent award:
“if implementation is by administrative means, such as through a regulation, then the
‘reasonable period of time’ will normally be shorter than for implementation through legislative
means. It seems reasonable to assume, unless proven otherwise due to unusual circumstances in
a given case, that regulations can be changed more quickly than statutes. To be sure, the
administrative process can sometimes be long; but the legislative process can oftentimes be
longer”.47
Once the implementing Member has chosen its means, the debate focuses on the
“particular circumstances” upon which “RPT may be shorter or longer”.
42 But panels are increasingly using this ability. For example, the Guatemala—Definitive Anti-Dumping Measures
on Grey Portland Cement from Mexico panel suggested that Guatemala revoke its anti-dumping measure even if it
declined Mexico’s request that Guatemala refund the anti-dumping duties collected (WT/DS156/R, adopted
17 November 2000, para. 9.6 and 9.7).
43 See note 21, above, paras 41 and 38. The same reasoning was followed by Mr El Naggar in the Salmon
award (WT/DS18/9, 23 February 1999, paras 36–37).

44 See 21.3(c) award Argentina—Measures Affecting the Export of Bovine Hides and the Import of Finished Leather,
WT/DS155/10, 31 August 2001, paras 42–50.

45 See 21.3(c) award United States—Anti-Dumping Act of 1916, WT/DS136/11, WT/DS162/14, 28 February
2001, para. 34.

46 Nonetheless, Mr Feliciano underlines in a footnote that “the non-conforming measure might also assume
other forms: e.g., an executive or administrative practice actually carried out but not specifically mandated or
authorized by statute or administrative regulation; or a ‘quasi-judicial’ determination by an administrative body”.
See note 44, above, footnote 14.
118
3. The “Particular Circumstances”

Arbitrators have elaborated step by step a very abundant case law about the
meaning of the words “particular circumstances”, set forth in Article 21.3(c) DSU. We
have already noted that the jurisprudence does not consider the time required for
conducting a risk assessment as “particular circumstances”.48 In the Canada—Patent
award, Mr Bacchus defined the “particular circumstances” as “those that can influence
what the shortest period possible for implementation may be within the legal system of
the implementing Member”.49
In the context of this definition, “external” factors such as the “contentiousness”50
of the issue, the economic prejudice accruing from the persistence of the incompatible
measure,51 or structural adjustments to affected industries52 were not considered by
arbitrators as “particular circumstances”.
By contrast, the complexity of the implementing measure is a particular
circumstance.53
For administrative actions, administrative guidelines, even if not binding, have
been taken into consideration by arbitrators,54 but the argument that the summer
vacation should be taken into account in calculating the “reasonable period of time”
for implementation was dismissed as irrelevant.55
For legislative actions, the implementing Member can use ordinary procedures.
Mr Ehlermann, in the Korea—Alcoholic beverages award, believed that it was “reasonable
to allow Korea to follow its normal legislative procedure”.56 But this does not mean
that an implementing measure is a law like others: the legislative Branch “should try to
comply with the international obligations as soon as possible, taking advantage of the
flexibility that it has in its normal legislative procedures”.57
In this context, the following factors have been found irrelevant:
– the enormous volume of legislation introduced and the small percentage of
legislation actually passed;58
– the date foreseen by the Parliament calendar for the beginning of the summer
recess;59

50 See note 36, above, para. 50 (for an administrative measure), and note 24, above, para. 58 (for a legislative
process).
52 See 21.3(c) award Indonesia—Certain Measures Affecting the Automobile Industry, WT/DS54/15, WT/DS55/
14, WT/DS59/13, WT/DS64/12, 7 December 1998, para. 23; see note 44, above, para. 41.
54 See note 36, above, para. 54 and Canada—Certain Measures Affecting the Automotive Industry, WT/DS139/
12, WT/DS142/12, 4 October 2000, para. 46.

56 See 21.3(c) award Korea—Taxes on alcoholic beverages, WT/DS75/16, WT/DS84/14, 4 June 1999, para. 42.
57 See 21.3(c) awards United States—Section 110(5) of the US Copyright Act, WT/DS160/12, 15 January 2001,
para. 39 and note 24, above, para. 64.

58 See note 57, above, para. 40 and note 45, above, para. 38.
119
– political considerations such as distribution of seats among parties;60

– the need for a period of transition to a new President, a new Administration,
and a new Congress.61
This treatment of case law would not be complete without mentioning two
problematical questions: differential treatment for developing countries, and the legal
value of what has been done by the implementing Member since the adoption of the
report(s).
(a) Developing countries

In the Indonesia—Automotive award, Mr Beeby gives life to Article 21.2 DSU,62
stating that:
“Although the language of this provision is rather general and does not provide a great deal of
guidance, it is a provision that forms part of the context for Article 21.3(c) of the DSU and
which I believe is important to take into account here. Indonesia has indicated that in a ‘normal
situation’, a measure such as the one required to implement the recommendations and rulings of
the DSB in this case would become effective on the date of issuance. However, this is not a
‘normal situation’. Indonesia is not only a developing country; it is a developing country that is
currently in a dire economic and financial situation. Indonesia itself states that its economy is
‘near collapse’. In these very particular circumstances, I consider it appropriate to give full
weight to matters affecting the interests of Indonesia as a developing country pursuant to the
provisions of Article 21.2 of the DSU. I, therefore, conclude that an additional period of six
months over and above the six-month period required for the completion of Indonesia’s
domestic rule-making process constitutes a reasonable period of time for implementation of the
recommendations and rulings of the DSB in this case”.63
It is difficult to assess in what other particular circumstances this additional
semester would be granted to developing countries. One could just note that the
India—Quantitative Restrictions panel report referred explicitly to this passage64 and that
Mr Feliciano seemed to have implicitly applied this “semester free” rule to Chile since
in Chile—Alcoholic Beverages he gave an RPT of 14 months and 9 days, exactly six
months longer than that requested by the EC.65 More recently, in its Bovine Hides
report, Mr Feliciano also took into consideration the point that Argentina “is a
developing country confronted by severe economic and financial problems”.66

62 “Particular attention should be paid to matters affecting the interests of developing country Members with
respect to measures which have been subject to dispute settlement.”

63 See 21.3(c) award Indonesia—Certain Measures Affecting the Automobile Industry, WT/DS54/15, WT/DS55/
14, WT/DS59/13, WT/DS64/12, 7 December 1998.

120
(b) Implementation
How the arbitrator should take into account the time passed since the adoption of
the report(s) is probably the most inextricable problem faced by arbitrators. The
arbitrator is reduced to a dilemma when the implementing Member failed to
implement in a timely manner during the period preceding the arbitration. There are
two possible approaches:
1) A pragmatic and realistic approach would be to consider that the arbitrator
cannot carry back the implementing steps. It is the way Mr Bacchus seemed
implicitly to adopt when he noted: “some of the administrative steps Canada
has identified to me, and has already taken toward implementation, could have
been accomplished sooner, simultaneously, or both. That said, and given what
has already transpired toward implementation, I point now to one of the
additional steps identified by Canada that has not yet been completed”.67
2) Another approach would be to take into account the fact that the Member has
not implemented in a timely manner. In this respect, Mr Lacarte pointed out:
“an implementing Member must use the time after adoption of a panel and/or Appellate
Body report to begin to implement the recommendations and rulings of the DSB.
Arbitrators will scrutinize very carefully the actions an implementing Member takes in
respect of implementation during the period after adoption of a panel and/or Appellate
Body report and prior to any arbitration proceeding. If it is perceived by an arbitrator that
an implementing Member has not adequately begun implementation after adoption so as
to effect ‘prompt compliance’, it is to be expected that the arbitrator will take this into
account in determining the ‘reasonable period of time’”.68
Neither of these solutions is satisfactory. The first rewards the implementing
Member which has done nothing until the arbitration process, the second could lead to
an RPT which it is impossible for the implementing Member to meet. This problem is
not insuperable considering that the arbitration process takes more or less one third of
the RPT.
Whatever the approach chosen, another problem is to decide the point in time up
to which or from which the arbitrator will or will not take into account what the
implementing Member has done. In an ideal world, it would be the date of the
circulation of the award. But of course, such a solution is unrealistic even if the arbitrator
can hazard some conjectures—as, for example, Mr Ehlermann did in the Canada—Patent
award when he noted that it is unnecessary to examine a pre-legislative phase which is
likely to be practically completed by the time the award is made public.69 The second
best benchmark would be the day of the hearing, since this is the last opportunity for a
contradictory debate on the matter. But we have already seen that Mr Lacarte seems to
take the view that he can only scrutinize what happened “prior the arbitration
69 Seenote 24, above, para. 62.
121
proceeding”. In the same manner, Mr Ganesan noted in the 1916 Act award: “In light of
the fact that all the parties in this case have, by a joint letter dated 19 December 2000,
agreed to this binding arbitration, I do not consider it necessary to deal with the
arguments that the United States could have enacted the required legislation by the close
of the second session of the 106th Congress in December 2000.”70
This very last topic about timing appreciation cannot hide the fact that the Appellate
Body’s arbitrators have built up, award after award, a very consistent and homogeneous
case law. The jurisprudence has been considerably elaborated and refined since 13
February 1997, when Mr Lacarte signed the first award with a decision taking only a
single paragraph.71 Nowadays, awards look more like a small judgment even if the final
decision, i.e. the duration of RPT finally granted, is left to the arbitrator’s discretion.
The final question is how the “shortest period possible” could possibly fit with the
“without delay” necessary to withdraw a prohibited subsidy.
IV. THE RELATIONSHIP BETWEEN THE REASONABLE PERIOD OF TIME AND

“WITHOUT DELAY”
The Appellate Body has clearly ruled that

“With respect to implementation of the recommendations or rulings of the DSB in a dispute
brought under Article 4 of the SCM Agreement, there is a significant difference between the
relevant rules and procedures of the DSU and the special or additional rules and procedures set
forth in Article 4.7 of the SCM Agreement. Therefore, the provisions of Article 21.3 of the
DSU are not relevant in determining the period of time for implementation of a finding of
inconsistency with the prohibited subsidies provisions of Part II of the SCM Agreement”.72
The question is then to determine what the “significant difference” is about. On
two occasions, a prohibited subsidy case has been involved with other violations.
In the Canada—Automotive case, the panel found, on the one hand, measures
incompatible with GATT and GATS, and on the other, that different measures were
prohibited subsidies that had to be withdrawn within 90 days. Complainants argued before
the arbitrator that this 90-day delay should equally apply for the RPT since Canada was
able to implement the same kind of measure for the DSB’s recommendations for the
prohibited subsidy. Mr Lacarte did not accept this argument, stating that:
“Canada’s ability to take ‘extraordinary action’ to withdraw the export subsidy ‘without delay’,
in accordance with the provisions of Article 4.7 of the SCM Agreement and pursuant to the
recommendation of the DSB, is not relevant for determining the ‘reasonable period of time’
under Article 21.3(c) of the DSU for implementation of the recommendations of the DSB
relating to Article I:1 and Article III:4 of the GATT 1994 and Article XVII of the GATS”.73

71 See 21.3(c) award Japan—Taxes on Alcoholic Beverages, WT/DS8/15, WT/DS10/15, WT/DS11/13, 14
February 1997, para. 27.
73 See 21.3(c) award Canada—Certain Measures Affecting the Automotive Industry, WT/DS139/12, WT/
DS142/12, 4 October 2000, para. 53.
122
It could be inferred from this passage that extraordinary legislative action could be
necessary to withdraw a subsidy. For an administrative action, preparatory steps or
non-binding administrative guidelines would probably not be taken into account. The
Canada—Automotive case offers a good opportunity to make a comparison since the
panel gave some explanation of its 90-day award:
“in examining what time-period would represent withdrawal ‘without delay’ in a particular case,
we consider that we may take into account the nature of the steps necessary to withdraw the
prohibited subsidy. We do not, however, agree with Canada that we should take into account the
existence or absence of adverse or trade-distorting effects resulting from the prohibited subsidy,
nor the time required to design replacement measures, as these factors are not related to the
consideration of what time-period would represent withdrawal ‘without delay’”.74
If the RPT’s arbitrator would not take into account trade-distorting effects, he
would probably (to the “shortest possible” extent) grant the time required to design
replacement measures.
In the FSC case, the same measure was found to be both a prohibited subsidy and
in breach with the agreement on agriculture. We have seen that the panel specified
1 October 2000 for withdrawal of the prohibited subsidy, but no RPT was determined
for agricultural subsidies.
Could an arbitrator under Article 21.3(c) DSU determine an RPT different from
the time period specified by the panel and adopted by the DSB for the same measure?
This question has not yet been answered, since parties have found a way to fill the gap,
inventing the new concept of “time period for compliance”,75 a hybrid animal born
from RPT and “without delay”.
V. CONCLUSION
The current system works efficiently, though there can be difficulties in

prohibited subsidy cases. It is not clear how matters could easily be improved. A
solution could be to give to panels the authority to determine the time period for
compliance in every case as is already done for prohibited subsidies. The interim
review stage of the panel could then be usefully devoted to a debate about the time
period. At the end, the matter would be reviewed by the Appellate Body.
Such a solution would avoid the difficulties arising from the determination after
the adoption of the panel. It does not mean that the arbitrator would disappear. First,
there can be occasions when the reasonable period of time specified by the panel
should be revisited in the light of circumstances such as the date of the adoption of the
report(s) or when the defending Member changes its mind about its implementing
measure. But the task of the arbitrator would be much easier. Then, it will allow him
to rule on other matters related to implementation. For example, we recall that Article
75 See document WT/DS108/11, 2 October 2000.
123
26.1(c) lays down: “the arbitration provided for in paragraph 3 of Article 21, upon
request of either party, may include a determination of the level of benefits which have
been nullified or impaired, and may also suggest ways and means of reaching a
mutually satisfactory adjustment; such suggestions shall not be binding upon the parties
to the dispute.” The level of nullification and impairment could indeed be mulled over
by an arbitrator during the RPT as is shown by the Article 25 DSU arbitration
procedure agreed by the EC and the US in the United States—Copyright case.76
76 See document WT/DS160/15, 3 August 2001.
124
Seeking Compliance with WTO Rulings:
Theory, Practice and Alternatives
Brendan P. McGivern*
I. Introduction
One of the most critical issues to emerge in the WTO recently is that of compliance
with the decisions of dispute settlement panels and the Appellate Body. Once a decision
has been rendered, how can a complaining party seek to ensure that the defending party
faithfully implements it? What recourse is available to the complainant if the defendant
refuses to comply?
These issues have emerged in a relatively small number of high profile and usually highly
politicized disputes. Under the rules of the WTO, where a defending party fails to imple-
ment a panel or Appellate Body decision, a complaining party ultimately has the right to
respond with retaliatory trade sanctions. Yet the imposition of sanctions often hurts the
commercial interests of the complaining party as much as the defendant, and the ability of
such retaliation to induce compliance is far from certain.
This article examines key issues of compliance and retaliation under the WTO Agree-
ment. It is divided into three parts. First, it examines the theory of retaliation. This includes
a discussion of the remedies provided for in the WTO Agreement, and the purpose of
retaliation under the GATT and the WTO.
Second, it examines the practice of retaliation. It reviews the cases to date in which the
retaliation provisions of the DSU have been invoked. It considers issues such as who de-
termines non-compliance, the means by which a complaining party can obtain the right to
retaliate, the quantification of the level of authorized sanctions, and the termination of
retaliation. It also examines problems associated with the application of sanctions. Third,
it considers alternative and additional means of seeking compliance, including compensa-
tion, fines, rotating or "carousel" retaliation, collective retaliation, and punitive sanctions.
*Counsellor, Permanent Mission of Canada to the World Trade Organization, Geneva, Switzerland. An
earlier version of this paper was presented at the May 3-4, 2001 conference of the Canadian Bar Association
in Ottawa, "The Practice of International Law in the 21st Century." The views expressed in this paper are
those of the author in his personal capacity.
141
125
142 THE INTERNATIONAL LAWYER
n. The Theory of Retaliation

A. The WTO Dispute Settlement System: Hierarchy of Remedies
The WTO dispute settlement system contains three key features that are absent from
the dispute settlement provisions of most international treaties: (1) panels that have com-
pulsory jurisdiction to examine complaints about violations of the Agreement; (2) appellate
review of panel decisions, again with compulsory jurisdiction; and (3) the ability to render
binding decisions.1
Decisions of panels and the Appellate Body are adopted virtually automatically by the
WTO Dispute Settlement Body (DSB), pursuant to the so-called "negative consensus"
rule.2 Where a country has been found to have breached its commitments, the panel and
Appellate Body reports normally recommend that the offending country bring itself into
conformity with its WTO obligations. Upon adoption of the reports, the recommendations
of the panel and Appellate Body become DSB rulings. In order to comply with a DSB
ruling, the defendant may have to repeal or amend its laws to eliminate the WTO-
inconsistency.
How does the complaining party enforce such WTO rulings?
It should be noted at the outset that DSB rulings are not self-executing. It remains up
to each defending party - a sovereign government - to choose how, and even if, it will
implement the ruling.
Article 21.3 of the DSU provides that if it is "impracticable" for the defending country
to comply immediately with the DSB rulings, it is given a "reasonable period of time" for
implementation.3 The issue of retaliation may arise if the defending party has not imple-
mented the DSB rulings at the end of this compliance period.
Article 3.7 of the DSU sets out the hierarchy of remedies in WTO dispute settlement.
It seeks to encourage, in descending order of preference: (1) bilateral settlement; (2) with-
drawal by the defendant of the WTO-inconsistent measure; (3) compensation and (4) as a
"last resort" retaliation.4
1 . Understanding on Rules and Procedures Governing the Settlement of Disputes, Apr. 1 5, 1994, Marrakesh
Agreement Establishing the World Trade Organization [hereinafter WTO Agreement] Annex 2, Legal In-
struments-Results of the Uruguay Round, reprinted in 33 I.L.M. 12 (1994) [hereinafter DSU].
2. The "negative consensus" or "automaticity" rule provides that certain specified requests to the DSB must
be granted unless all WTO Members at the DSB meeting - including the Member that made the request -
choose to reject it. The negative consensus rule applies to panel establishment, to the adoption of panel and
Appellate Body reports, and to requests to retaliate. DSU, supra note 1, art. 6.1, 16.4, 17.14, 22.6 and 22.7.
3. If the disputing parties cannot agree on the "reasonable period of time" for implementation, DSU Article
21.3(c) provides that it can be determined through arbitration. DSU, supra note 1, art. 21(3)(c).
4. DSU, supra note 1, art. 3.7. Article 3.7 provides in part as follows:
The aim of the dispute settlement mechanism is to secure a positive solution to a dispute. A solution
mutually acceptable to the parties to a dispute and consistent with the covered agreements is clearly
to be preferred. In the absence of a mutually agreed solution, the first objective of the dispute settlement
mechanism is usually to secure the withdrawal of the measures concerned if these are found to be
inconsistent with the provisions of any of the covered agreements. The provision of compensation
should be resorted to only if the immediate withdrawal of the measure is impracticable and as a tem-
porary measure pending the withdrawal of the measure which is inconsistent with a covered agreement.
The last resort which this Understanding provides to the Member invoking the dispute settlement
procedures is the possibility of suspending the application of concessions or other obligations under
the covered agreements on a discriminatory basis vis-à-vis the other Member, subject to authorization
by the DSB of such measures.
VOL. 36, NO. 1
126
SEEKING COMPLIANCE WITH WTO RULINGS 143
A complaining party retaliates when it "suspends concessions or other obligat

the covered agreements."5 In other words, the complainant denies benefits tha
otherwise be legally required to extend to the other WTO Member.
Most retaliation requests seek to suspend concessions. One example of a con
tariff binding, which is a commitment not to impose duties on a product abov
specified rate. By "suspending" such a concession, the complaining party co
duties above the bound rate on products from the defending party. It would do
authorization of the DSB. It would also suspend concessions on a non-most-
nation basis - the retaliatory rates would be imposed solely on the imports of th
party.
More recently, the EC and Japan sought to suspend "other obligations" by requesting
DSB authorization to adopt laws to "mirror" the WTO-inconsistent U.S. Anti-Dumping
Act of 1916. This is discussed below in the "Cases to Date" section {infra Part ELLA).
B. Retaliation under the GATT
The WTO retaliation system is based in part on the rules of the WTO's predecesso
organization, the GATT. An important reason why the drafters of the original GATT 19
included a provision on retaliation was to curb the customary international law right
unilateral reprisal, and to replace it with multilaterally-authorized retaliation.6 As stated
one of the drafters:
We have asked the nations of the world to confer upon an international organization the rig
to limit their power to retaliate. We have sought to tame retaliation, to discipline it, to ke
it within bounds. By subjecting it to the restraints of international control, we have endeavoure
to check its spread and growth, to convert it from a weapon of economic warfare to an instr
ment of international order.7
This notion of "taming retaliation" was carried into the WTO. DSU Article 23 prohib
unilateral, unauthorized action by the complaining party.8
Article XXÜL2 of GATT 1947 provides, in part, that if the members of the GATT
acting collectively, considered that the circumstances were "serious enough," they co
"authorize a contracting party or parties to suspend the application to any other contract
party or parties of such concessions or other obligations under this Agreement as they m
determine to be appropriate in the circumstances."9
However, this provision was subject to the normal "positive consensus" rule of the GAT
which meant that any request by a complaining party to retaliate required the consent
all GATT Contracting Parties, including the defendant itself. This effectively ensured th
the retaliation provisions of the GATT remained a dead letter. Indeed, in the entire histo
of the GATT (1948-1994), the GATT General Council approved a request for retaliat
only once, and this was never implemented.10
5. Id.
6. Ernst-Ulrich Petersmann, The GATT/WTO Dispute Settlement System: International Law,
International Organizations and Dispute Settlement 82 (Kluwer Law International 1997).
7. Id.
8. DSU, supra note 1, art. 23.
9. General Agreement on Tariffe and Trade, Oct. 30, 1947, 61 Stat. A-l 1, T.I.A.S. 1700, 55 U.N.T.S. 194
[hereinafter GATTI.
10. See Netherlands- Measures of Suspension of Obligations to the United States, Nov. 8, 1952, GATT B.I.S.D.
(1st Supp.) at 32 (1953).
SPRING 2002
127
C. Purpose of Retaliation
Under the GATT, one of the primary purposes of retaliation was to allow for a restoration
of the "balance of concessions" between the complaining and defending parties under the
Agreement. Each party to the GATT received commercial benefits from the legally binding
market access commitments (concessions) made by the other parties. It was argued that
since a violation of the GATT by a country upset this equilibrium of commercial benefits
retaliation would permit the restoration of the balance. As one informed observer wrote:
"retaliation is fair because it re-establishes the balance of concessions between the two
parties, a balance that is thrown into disequilibrium when one party has violated GATT's
rules."11
Under the WTO, however, it is clear that the main objective of the complaining party
is not a restoration of the balance of concessions, but is rather to induce the defending
party to comply with its obligations. As noted by the arbitral panel in Bananas: "the au-
thorization to suspend concessions or other obligations is a temporary measure pending
full implementation by the Member concerned. We agree with the United States that this
temporary nature indicates that it is the purpose of countermeasures to induce compliance"12
(Emphasis in original)
A similar statement was made by the arbitral panel in Brazil Aircraft, in seeking to define
the term "appropriate countermeasures" in Article 4 of the Agreement on Subsidies and
Countervailing Measures:
While the parties have referred to dictionary definitions for the term "countermeasures", we
find it more appropriate to refer to its meaning in general international law and to the work
of the International Law Commission (ILC) on state responsibility, which addresses the notion
of countermeasures. We note that the ILC work is based on relevant state practice as well as
on judicial decisions and doctrinal writings, which constitute recognized sources of interna-
tional law. When considering the definition of "countermeasures". . . we note that counter-
measures are meant to "induce (the State which has committed an internationally wrongful
act) to comply with its obligations
in the EC-Bananas (1999) arbitration made a similar statement. We conclude

measure is "appropriate" inter alia if it effectively induces compliance [origina
11. William J. Davey, Dispute Settlement in GATT, 11 Fordham Int'l LJ. 51 (1987), rep
Jackson et al., Legal Problems of International Economic Relations: Cases, Mat
the National and International Regulation of Transnational Economic Relations
Publishing Co. 1995).
12. European Communities - Regime for the Importation, Sale and Distribution of Ban
Arbitration by the European Communities under Article 22.6 of the DSU: Decision by th
DS27/ARB (Apr. 9, 1999), at http://www.wto.org. The statement cited above was quoted w
arbitrators in the Hormones case: see European Communities - Measures Concerning Mea
(Hormones). Original Complaint by Canada. Recourse to Arbitration by the European
Article 22.6 of the DSU: Decision of the Arbitrators, WT/DS48/ARB Quly 12, 1999), at h
see abo European Communities - Measures Concerning Meat and Meat Products (Hormone
plaint by the United States. Recourse to Arbitration by the European Communities under
DSU: Decision of the Arbitrators, WT/DS26/ARB Quly 12, 1999), at http://www.wto.org
13. Brazil - Export Financing Programme for Aircraft: Recourse to Arbitration by Brazi
of the DSU and Article 4.1 1 of the SCM Agreement: Decision by the Arbitrators, WT/D
2000), at http://www.wto.org.
VOL. 36, NO. 1
128
If the purpose of WTO retaliation is to induce compliance, has it met this

Have the threat and use of sanctions led defending parties to comply with WT
These issues are considered in the next section.
IH. The Practice of Retaliation
A. Cases to Date14
Since the WTO Agreement entered into force in January 1995, there have been a number
of instances in which complaining parties have sought, and in some cases received, author-
ization to retaliate.
1. EC- Bananas
There were two retaliation requests that arose from the WTO-inconsistent import re-
gime for bananas, one from the United States and the other from Ecuador. In April 1999,
the DSB authorized the United States to suspend concessions to the EC and its Member
States in the amount of U.S.S191.4 million per year for the failure of the EC to comply
with the DSB rulings in Bananas. iS In May 2000, the DSB authorized Ecuador to suspend
concessions in the amount of U.S. $201.6 million.16 (The Ecuador case was the first time
in which "cross retaliation" was authorized, i.e., retaliation in a different sector, or under a
different WTO agreement than that in which the violation occurred.) The United States
implemented its authorized retaliation, although Ecuador did not. In April 2001, the United
States and the European Union reached a provisional settlement in this dispute. Under the
terms of this agreement, U.S. sanctions were suspended on July 1, 2001, and will be "de-
finitively lifted" once the EU carries out the agreed changes to its banana import regime.17
2. EC - Hormones
There were two retaliation requests made against the EC (one from the United States,
and the other from Canada) arising from the failure of the EC to withdraw its WTO-
inconsistent ban on hormone-treated beef. On July 26, 1999, the DSB authorized the
United States to suspend concessions against the EC and its Member States in the amount
of U.S.$1 16.8 million per year.18 At the same DSB meeting, Canada received authorization
to suspend concessions in the amount of Cdn$l 1.3 million per year.19 Both Canada and the
United States implemented this retaliation by imposing 100 percent duties on certain ag-
ricultural imports from the EC. However, the EC has not lifted its ban on hormone-treated
beef, and the U.S. and Canadian sanctions remain in place.
14. This section is current as of January 2002.

15. Bananas: Decision by the Arbitrators, WT/DS27/ARB.
16. European Communities - Regime for the Importation, Sale and Distribution of Bananas - Recourse to
Arbitration by the European Communities under Article 22.6 of the DSU: Decision by the Arbitrator!, WTV
DS27/ARB/ECU (Mar. 24, 2000), at http://www.wto.org.
17. See Press Release, Office of the United States Trade Representative, U.S. Government and European
Commission Reach Agreement to Resolve Long-Standing Banana Dispute (Apr. 11, 2001), available at http://
www.ustr.gov/releases/2001/04/index.shtml. See Press Release, European Union, EU and Ecuador Reach
Agreement to Resolve WTO Banana Dispute (Apr. 30, 2001), available at http://europa.eu.mt/rapid/start/cgi/
guesten.ksh?reslist.
18. Hormones: Decision by the Arbitrators. WT/DS26/ARB, at http://www.wto.org.
19. Hormones: Decision by the Arbitrators. WT/DS48/ARB, at http://www.wto.org.
SPRING 2002
129
3. Australian Salmon
In July 1999, Canada made a request to the DSB to retaliate against Australia for Cdn$45
million in response to Australia's failure to lift its WTO-inconsistent ban on imports of
Canadian salmon.20 Canada and Australia put in place the first ad hoc "sequencing" arrange-
ment in the WTO, to allow for a prior ruling by an Article 21.5 compliance panel on the
WTO-consistency of Australia's implementing measures. The Article 21.5 panel subse-
quently ruled that Australia had not complied.21 However, this dispute has since been largely
settled between the parties, and no sanctions have been imposed.
4. U.S. Foreign Sales Corporations (FSC)

In November 2000, the EC put in a retaliation request for U.S.$4.043 billion in the FSC
dispute.22 This is the largest retaliation request ever made in the history of the multilateral
trading system. The EC and the United States negotiated a "sequencing" agreement, to
allow for a prior ruling on the WTO-consistency of the U.S. FSC implementing legislation
through the compliance panel process. The August 2001 report of the Article 21.5 com-
pliance panel concluded that the United States had not implemented the DSB rulings; the
Appellate Body affirmed this in January 2OO2.23 At the time of writing (January 2002), the
United States is expected to challenge the level of retaliation sought by the EC through
arbitration under DSU Article 22.6.
5. Brazil Aircraft
In December 2000, the DSB authorized Canada to suspend concessions to Brazil in the
amount of Cdn$344.2 million as "appropriate countermeasures" within the meaning of
Article 4 of the Agreement on Subsidies and Countervailing Measures, following Brazil's failure
to withdraw WTO-inconsistent subsidies in the regional aircraft sector.24 To date, Canada
has not yet implemented these countermeasures.
6. Canada Dairy
In February 2001, the United States25 and New Zealand26 each sought to suspend con-
cessions in the amount of U.S.$35 million per year in the Dairy dispute. In July 2001, an
Article 21.5 panel concluded that Canada had not implemented the DSB rulings.27 This
20. Australia - Measures Affecting Importation of Salmon - Recourse by Canada to Article 22.2 of the DSU,
WT/DS18/12 (July 15, 1999), at http://www.wto.org.
2 1 . Australia - Measures Affecting Importation of Salmon - Recourse to Article 2 1 .5 by Canada: Report of
the Panel, WT/DS18/RW(Feb. 18, 2000), at http://www.wto.org.
22. United States - Tax Treatment for "Foreign Sales Corporations" - Recourse by the European Com-
munities to Article 4.10 of the SCM Agreement and Article 22.2 of the DSU, WT/DS1O8/13 (Nov. 17, 2000),
at http://www.wto.org.
2 3 . United States - Tax Treatment for "Foreign Sales Corporations" - Recourse to Article 2 1 .5 of the DSU
by the European Communities: Report of the Panel, WT/DS108/RW (Aug. 20, 2001), at http://www/wto.org.
See also Report of the Appellate Body, WT/DS108/AB/RW Qan. 14, 2002), at http://www.wto.org.
24. Brazil - Export Financing Programme for Aircraft - Decision by the Arbitrators, WT/DS46/ARB, at
http://www.wto.org.
2 5 . Canada - Measures Affecting the Importation of Milk and the Exportation of Dairy Products - Recourse
by the United States to Article 22.2 of the DSU. WT/DS1O3/17 (Feb. 19, 2001), at http://www.wto.org.
26. Canada - Measures Affecting the Importation of Milk and the Exportation of Dairy Products - Recourse
by New Zealand to Article 22.2 of the DSU, WT/DS113/17 (Feb. 19, 2001), at http://www.wto.org.
27. Canada - Measures Affecting the Importation of Milk and the Exportation of Dairy Products - Recourse
to Article 21.5 of the DSU by New Zealand and the United States: Report of the Panel, WT/DS103/RWand
WT/DSlB/RWQuly 11, 2001), at http://www.wto.org.
VOL. 36, NO. 1
130
finding of inconsistency was overturned on appeal in December, although th

Body also found that it did not have a sufficient factual basis to make a defin
mination on Canadian compliance.28 A second Article 21.5 panel is now re-ex
Canadian measures.29 The disputing parties put in place (and later extended)
agreements to ensure a prior ruling on the WTO-consistency of Canada'
through the compliance panel process.
7. U.S. 1916 Anti-Dumping Act

In January 2002, the EC and Japan took the unprecedented step of seekin
thorization to adopt laws to "mirror" the U.S. Anti-Dumping Act of 1916.30 Th
Body had earlier found that this law - which, among other things, allows for tr
against dumped imports - violates U.S. obligations under GAIT Article VI an
Anti-Dumping Agreement.31 When the United States failed to implement the
in this dispute by the expiration of the compliance period, the EC and Japan s
iation in the form of DSB authorization to adopt parallel legal regimes.32 T
Japanese laws would replicate the substantive and procedural aspects of the 191
would apply solely to U.S. imports. This unusual request seemed to be based
that the U.S. legislation has not been applied, and so a monetary-based request
tariff concessions could be vulnerable to attack on the ground that the level of
or impairment was either negligible or non-existent. (The Japanese retaliat
states, "it is not practical to indicate the level of nullification or impairment i
monetary value."33) The EC and Japanese requests were referred to arbitration
Article 22.6. The three disputing parties also announced that they would imme
pend the arbitration to allow additional time for the U.S. Congress to pass th
implementing legislation.
8. U.S. Copyright Act

In January 2002, the EC sought authorization to retaliate against the Unite
the dispute over Section 1 10(5) of the U.S. Copyright Act.34 This law, which exem
smaller bars, restaurants, and other public places from copyright liability w
music, was found to violate U.S. obligations under the TRIPS Agreement.35
28. Canada - Measures Affecting the Importation of Milk and the Exportation of Dairy Produc
to Article 2 1.5 of the DSU by New Zealand and the United States: Report of the Appellate Body
AB/RWand WT/DS113/AB/RW(Dec. 3, 2001), at http://www.wto.org.
29. Canada - Measures Affecting the Importation of Milk and the Exportation of Dairy Prod
Recourse by the United States to Article 21.5 of the DSU: Requests for the Establishment of a
DS1 13/23 and WT/DS1O3/23 (Dec. 6, 2001) at http://www.wto.org.
30. Unfair Competition Act (Anti-dumping Act of 1916), 15 U.S.C. §§ 71-76 (2001).
3 1 . United States - Anti-Dumping Act of 1 9 1 6, Report of the Appellate Body, WT/DS 1 3 6/AB
DS162/AB/R (Aug. 28, 2000), at http://www.wto.org.
32. United States - Anti-Dumping Act of 1916, Recourse by the European Communities to Ar
the DSU, WT/DS 136/1 5 Quii. 7, 2002), at http://www.wto.org. See also United States -Anti-D
1916, Recourse by Japan to Article 22.2 of the DSU, WT/DS162/18 Qan. 7, 2002), at http://ww
33. United States - Anti-Dumping Act of 1916, Recourse by Japan to Article 22.2 of the DSU
18 Qan. 7, 2002), at http://www.wto.org.
34. Copyright Act, 17 U.S.C. §§ 101 etseq (2001).
35. United States - Section 1 10(5) of the U.S. Copyright Act, Report of the Panel, WT/DS160
2000), at http://www.wto.org.
SPRING 2002
131
unusual, in that before the expiration of the compliance period, the parties
vance ruling on the level of nullification or impairment through an ad hoc ar
DSU Article 25.36 This arbitration facilitated the negotiation of a temporary
agreement, in the form of U.S. contributions to projects for EU musicians fo
period.37 The EC then put in a retaliation request to preserve its rights pen
sional funding for the compensation arrangement. The EC sought DSB au
suspend its obligations under TRIPS "to permit the levying of a special f
nationals in connection with border measures concerning copyright goods."3
fee" would apparently be imposed on U.S. companies seeking the enforcement
in the EU. The EC request was referred to arbitration under DSU Article
disputing parties also announced that they would immediately suspend the a
allow additional time for Congressional approval for the compensation packa
B. Who Determines Non-Compliance?
One of the most difficult compliance issues that has confronted the WTO has been th
proper "sequencing" between the determination of WTO-consistency of the implementin
measures and the recourse by the complaining party to retaliation. In principle, the DSU
provides for a multilateral process to adjudicate disputes over whether a defending part
has implemented the DSB rulings. Article 21.5 provides in part as follows:
Where there is disagreement as to the existence or consistency with a covered agreement o

measures taken to comply with the [DSB] recommendations and rulings such dispute shall b
decided through recourse to these dispute settlement procedures, including wherever possi
resort to the original panel. The panel shall circulate its report within 90 days after the da
of referral of the matter to it.39
Thus, Article 21.5 sets up a ninety-day procedure to determine compliance. However,

is possible to read Article 22 to require that retaliation requests must be approved by th
DSB within thirty days of the expiration of the compliance period, or else the complaini
party would lose its right to have the request approved by negative consensus. This is t
U.S. position. If this interpretation is correct, then the ninety-day compliance process wo
be completed long after the thirty-day deadline for retaliation.40
This problem has arisen principally because of poor drafting in the DSU. WTO Mem
bers have responded to this ambiguity in two ways.
First, as noted earlier, there has been a growing number of cases in which the disputin
parties have entered into bilateral "sequencing" agreements to provide that the Article 2 1
panel process must precede any retaliation under Article 22. These have been ad hoc, cas
specific solutions.
36. United States - Section 1 10(5) of the U.S. Copyright Act, Recourse to Arbitration under Article 25
the DSU, Award of the Arbitrators, WT/DS160/ARB25/1 (Nov. 9, 2001), at http://www.wto.org.
37. Press Release, European Commission, EU and U.S. Agree on Temporary Compensation in Copyrig
Dispute (Dec. 19, 2001).
38. United States - Section 1 10(5) of the U.S. Copyright Act, Recourse by the European Communities to
Article 22.2 of the DSU, WT/DS160/19 (Jan. 7, 2002), at http://www.wto.org.
39. DSU, supra note 1, art. 21.5.
40. See Cherise M. Valles & Brendan P. McGivern, The Right to Retaliate under the WTO Agreement: Th
'Sequencing' Problem, 34(2) J. World Trade 63-84 (Apr. 2000).
VOL. 36, NO. 1
132
Second, in an effort to find a more durable solution to the sequencing problem

of Members, including Canada, made a proposal to amend the DSU to provide c
a determination of WTO-inconsistency must precede any request to suspend
This proposal was made at the December 1999 Seattle Ministerial Conference
again in November 2001 Doha Ministerial Conference.42 However, any amendm
DSU requires the consensus of all Members, and such a consensus has not bee
The November 2001 Ministerial Conference also mandated new negotiatio
dispute settlement, and the sequencing issue will likely be part of those negot
C. Obtaining Authorization to Retaliate
The complaining party triggers its right to retaliate by invoking DSU Article 22.2:
If the Member concerned fails to bring the measure found to be inconsistent with a cover
agreement into compliance therewith or otherwise comply with the recommendations a
rulings within the reasonable period of time . . . such Member shall, if so requested, and n
later than the expiry of the reasonable period of time, enter into negotiations with any par
having invoked the dispute settlement procedures, with a view to developing mutually accep
able compensation. If no satisfactory compensation has been agreed within 20 days after th
date of expiry of the reasonable period of time, any party having invoked the dispute settlemen
procedures may request authorization from the DSB to suspend the application to the Memb
concerned of concessions or other obligations under the covered agreements.43
The drafting of this provision leaves much to be desired. Nevertheless, the procedure se
out in Article 22.2 is as follows:
• the complaining party makes an initial determination that the defending party has failed t
comply with the DSB rulings within the implementation period;
• the complaining party may - but is not required to - seek compensation from the defendan
• while the text is not completely clear on this point, if the complaining party chooses not t
seek compensation, it then waits until twenty days after the expiration of the complian
period before seeking to retaliate; and
• if the complaining party does seek compensation, but the negotiations fail to produce s
isfactory compensation within twenty days after the expiration of the compliance period,
may request authorization from the DSB to retaliate.44
Article 22.6 provides that the DSB shall, upon request, grant authorization to suspe
concessions or other obligations within thirty days of expiration of the compliance perio
unless the defending party invokes its right to arbitration. The defending party may s
arbitration on the grounds that the retaliation sought is excessive. In a situation of "cros
retaliation," the defendant may also argue that certain principles, as set out in Article 22
have not been followed.45 The arbitration is supposed to be completed within sixty day
during which time no retaliation may occur.
41. Ministerial Conference Proposed Amendment to the Understanding on Rules and Procedures Govern
ing the Settlement of Disputes, WT/MIN(99)/1 5 (Dec. 3, 1999), at http://www.wto.org.
42. Amendment of Certain Provisions of the Understanding on Rules and Procedures Governing the Set
tlement of Disputes: Amendment Submission by Bolivia, Canada, Chile, Colombia, Costa Rica, Ecuador, Japa
Korea, New Zealand, Norway, Peru, Switzerland, Uruguay and Venezuela, WT/MIN(01)/W/6 (Nov. 1, 2001)
at http://www.wto.org.
43. DSU, supra note 1, art. 22.2.
SPRING 2002
133
Article 22.7 provides that the arbitrator cannot "examine the nature of the
or other obligations to be suspended" but must instead "determine whethe
such suspension is equivalent to the level of nullification or impairment." Th
may also examine whether the principles of Article 22.3 have been followed.
The decision of the arbitrator is final and is not subject to appeal. The DSB
promptly" of the decision, but does not adopt it. When requested to do so, th
authorization (by negative consensus) to suspend concessions or other obligati
tent with the decision of the arbitrator. In practice, this means that where t
has reduced the level of suspension requested, the complaining party puts in
ation request, matching the level of nullification or impairment as determin
bitrator. This revised request will be approved virtually automatically by the
D. Quantifying the Nullification or Impairment
Article 22.4 provides that the level of suspension of concessions or other obligations
authorized by the DSB shall be "equivalent to the level of the nullification or impairment."
Arbitral panels have generally determined that the date from which nullification or im-
pairment should be measured is the date on which the defending party was to have been in
compliance, i.e., the end of the "reasonable period of time" for implementation. They have
not chosen the date of the adoption of measure that has subsequently been found to be
WTO-inconsistent.46 This is consistent with the well-established principle that the GATT/
WTO dispute settlement system generally does not provide damages for past losses. Instead,
as noted above, the preferred remedy is prospective, not retrospective: the defending party
is to withdraw the WTO-inconsistent measure.47
Arbitrators have also determined the level of nullification or impairment through the use
of so-called "counterfactual" situations (i.e., a hypothetical state of affairs that would have
ciple" is that the complaining party should "first seek to suspend concessions or other obligations with respect
to the same sector(s) as that in which the panel or Appellate Body has found a violation." However, if the
complainant "considers that it is not practicable or effective to suspend concessions or other obligations with
respect to the same sectors)," then "it may seek to suspend concessions or other obligations in other sectors
under the same agreement" Going further, if the complainant considers that "it is not practicable or effective
to suspend concessions or other obligations with respect to other sectors under the same agreement," and that
"the circumstances are serious enough," it may "seek to suspend concessions or other obligations under another
covered agreement."
46. For example, the arbitrators in Hormones stated:
[W]e consider that our starting point is as follows: what would annual prospective . . . exports of
hormone-treated beef and beef products to the EC be if the EC bad withdrawn the ban on 13 May 1999}
13 May 1999 is the date of expiration of the reasonable period of time granted to the EC to implement
the panel and Appellate Body reports . . . We cannot assume that the EC from 1989 onwards, i.e. from
the time it imposed the ban, was under a legal obligation to withdraw the ban. We note, in this respect,
that the violations found were violations of the SPS Agreement, an agreement only in existence from
1 January 1995 onwards [original emphasis].
Hormones, supra note 12.
47. Since the special arbitration in Copyright Act was conducted during the implementation period, the
arbitrators found that it was not feasible to use the end of the implementation period as the point at which
nullification or impairment should be assessed. Instead, they chose the date on which the matter was referred
to them. See United States - Section 1 10(5) of the U.S. Copyright Act, Recourse to Arbitration under Article
25 of the DSU, Award of the Arbitrators, WT/DS160/ARB25/1 (Nov. 9, 2001), at http://www.wto.org.
VOL. 36, NO. 1
134
SEEKING COMPLIANCE WITH WTO RULINGS 1 5 1
existed if the defending party had complied with its obligations). The use of "cou
tuals" to establish the level of nullification or impairment sustained by the compla
involve a fair degree of subjective assessment by the arbitrators.
For example, the arbitrators in Bananas considered a range of counterfactual sit
in order to determine the monetary value of the nullification or impairment sus
the United States. They compared the value of EC imports from the United State
the then-existing, WTO-inconsistent regime with the value under various alt
WTO-consistent (counterfactual) regimes. Each counterfactual was based on a d
import regime, such as a tariff-only import system, a licensing system, or a tar
The arbitrators selected what they considered to be "a reasonable counterfactual"
their calculations on nullification or impairment based on this selection. There is
no analysis in the report as to why the counterfactual they chose was a reasonabl
calculate the nullification or impairment suffered by the complainant.
The counterfactual approach has been followed in other cases.48 For example,
mones, the arbitrators made a series of assumptions on what U.S. and Canadian ex
hormone-treated beef to the EC would have been if the EC had withdrawn its ban
counterfactual included assumptions on issues such as trade volumes, price, and th
of market share between Canada and the United States. A change in any one
assumptions could have led to very different conclusions regarding the level of nu
or impairment sustained by the two complaining parties.
Robert Hudec has suggested that although the methodology followed by such arb
may be flawed, their overall result may be politically acceptable. Writing about t
of the arbitrators in Bananas, Hudec stated:
The report contained a lengthy description of the voluminous and complex informat
arguments submitted by the parties, but at the end the panel explained its conclusio
paragraph, announcing which of several alternative methods of measurement it had
and then simply announcing the 'correct' number with little further explanation. As a sc
explanation, the panel's report was disappointingly obscure. As a political solution to t
however, it obviously looked objective enough to persuade the relevant audience
countries that a neutral tribunal had made an objective assessment of equivalence. As
the relevant audiences were willing to accept the decision, it did all that was politica
essary.49
Given the rather subjective nature of this process, it remains open to question whether
the "relevant audiences" will always be willing to accept such arbitral decisions. To the
extent that the arbitrators are rendering decisions that do not easily withstand analytical
scrutiny, affected parties may not be willing to concede the legitimacy of the results in all
future cases.
E. Ending Retaliation
The DSU contains only skeletal provisions on how and when retaliation should be
minated. Article 22.8 states in part:
48. The arbitrators in the 2000 EC/Ecuador arbitration chose the same counterfactual as was used in the
1999 US/EC arbitration. See Bananas: Decision by the Arbitrators. (2000 Ecuador complaint), supra note 16, para.
166, at http://www.wto.org.
49. Robert E. Hudec, Broadening the Scope of Remedies in WTO Dispute Settlement, in Improving WTO
Dispute Settlement Procedures: Issues and Lessons from the Practice of Other International Courts
and Tribunals 390-91 (Friedl Weiss ed., 2000).
SPRING 2002
135
1 52 THE INTERNATIONAL LAWYER
The suspension of concessions or other obligations shall be temporary and shall onl
until such time as the measure found to be inconsistent with a covered agreemen
removed, or the Member that must implement recommendations or rulings provid
to the nullification or impairment of benefits, or a mutually satisfactory solution
Thus, as noted above, retaliation should apply only until one of three thing
(1) the WTO-inconsistent measure has been removed; (2) the defending party
"solution" to the nullification or impairment, perhaps by providing temporary
tion; or (3) there is a mutually satisfactory settlement.
There are a number of problems with this provision. First, it provides no clear
whereby the defendant can seek to have the retaliation lifted. Presumably, the
party would make a request to the DSB to withdraw the authorization to suspe
sions, but Article 22.8 does not specifically say this.
Second, it provides no mechanism to resolve a dispute over whether the defen
has, in fact, complied with the DSB rulings. What if the defending party asserts
removed the WTO-inconsistent measure, but such an assertion is contested b
plainant? What if the defending party in feet withdraws the WTO-inconsisten
but immediately puts in place a new, equally restrictive replacement measur
nothing in Article 22.8 on how such disputes could be resolved.50
Third, there is no express language imposing on the DSB the obligation to wi
authorization if the conditions set out in Article 22.8 have been met. To date, t
the withdrawal of the authorization to suspend concessions has not been bro
the DSB. Nevertheless, the scope and application of Article 22.8 could well be
of dispute in future.
F. Problems with Retaliation
There are a number of well-known problems associated with the use of retaliation, most
notably the following. First, sanctions are usually imposed at a punitive rate, which has the
effect of stopping trade in the product that is the subject of the retaliation. Many have
argued that compliance issues should be resolved in a manner that promotes, rather than
harms, liberalized trade. As noted by the Report of the International Financial Institution
Advisory Commission: "Retaliation is contrary to the spirit of the WTO. Sanctions increase
restrictions on trade and create or expand groups interested in maintaining the restrictions.
Domestic bargaining over who will benefit from protection weakens support for open trad-
ing arrangements."51
Second, the use of retaliation may provide little or no relief to the affected industry in
the complaining party. Such an industry has been denied market access because of WTO-
illegal measures maintained by another country. To hit back against imports from the de-
fendant does not, from the standpoint of the affected industry, address this central problem.
50. In practice, given the lack of precision in Article 22.8, the disputing parties would likely reach znadhoc
agreement to resolve such disagreements through the compliance panel process under Article 21.5.
5 1 . The Commission was established by the U.S. Congress to make recommendations on future U.S. policy
toward a number of international institutions, including the WTO. The Commission's report is available on-
line at http://www.house.gov/jec/inu7meltzer.htm. See also Steve Charnovitz, Should the Teeth Be Pulled? A
Preliminary Assessment of WTO Sanctions (Sept. 15, 2000) (on file with the University of Minnesota Law
School).
VOL. 36, NO. 1
136
Moreover, abstract theories about the "rebalancing" of the bilateral commercial rel
will likely be of scant interest to the industry whose goods or services still fa
inconsistent market access barriers abroad.
Third, retaliation invariably imposes economic costs on the consumers and busin
the complaining party. For example, where tariffs are raised on the consumer good
defendant, the consumers of the complaining party will face higher prices. W
imported products are used as intermediate goods in the production process, certa
nesses in the complaining party will similarly face commercial hardship. Given th
grated nature of many economies, particularly among developed countries, it is oft
difficult to design a retaliation list that does not inflict harm on the complainant.5
Fourth, and perhaps most importantly, there is little hard evidence that retalia
and does achieve its objective of inducing compliance by the defendant. In Bana
sanctions were in place for over two years before a settlement was reached. The Ap
U.S.-EU settlement was based on a number of factors, including the desire of both
to remove this significant irritant from the agenda in the run-up to the Novemb
WTO Ministerial Meeting in Qatar. The U.S. sanctions may have played a part
moting a resolution of the dispute, but few observers would argue that retaliation
principal cause of the settlement. Moreover, as noted above, the EC has still not c
with the DSB rulings in Hormones, and the U.S. and Canadian sanctions impose
1999 remain in place.
Given these problems with retaliation, what are some possible alternatives?
IV. Additional and Alternative Means of Seeking

Compliance
A. Political Pressure
Political pressure is invariably the sub-text that accompanies all trade disputes, and
rarely mentioned as supplementary means of seeking compliance. Nevertheless, it is w
recalling briefly that a non-compliant defending party will usually face at least three sep
forms of political pressure.
First, it will face pressure from its own exporters, whose products are facing prohi
sanctions in the market of the complaining party. Members of the affected industry
defending party will press their own government to resolve the dispute in order to
about the removal of the sanctions. This pressure will usually commence before the
sanctions are put in place, particularly if the complaining party publishes an initial r
ation list, from which it will prepare a narrower final list of products subject to sanc
Indeed, engendering an adverse reaction from a range of industries in the defending
is one of the reasons that a complainant will release such an initial retaliation list.
Second, the defending party may face growing bilateral political pressure in its rela
with the complainant. The failure of a defending party to comply with its obligation
cloud bilateral relations, and what started out as a commercial dispute can develop
52. Petros Mavroidis writes: "WTO countermeasures, the ultima ratio of the system, fail on both eff
ness and impartiality grounds ... In all cases, economic theory suggests that the country adopting them
itself in the foot' by imposing costs on its society." Petros Mavroidis, Remedies in the WTO Legal System: B
a Rock and a Hard Place, 11(4) European J. Int'l Law 763 (2000).
SPRING 2002
137
major irritant in the broader political relationship. This will often generate add
sure on the defendant to resolve the dispute.
Third, the defendant will face multilateral political pressure within the WTO
provides that issues of implementation are subject to the multilateral surveilla
DSB. This may require the defendant to explain to all WTO Members at DSB
why it has not complied.
B. Compensation and Fines
As noted above, voluntary compensation is currently provided for in the DSU as a tem-
porary measure pending full implementation. Such compensation could take the form of
increased access to the market of the defending party. For example, the defending party
could agree to lower its bound rate on specified products as a form of interim compensation
pending the withdrawal of the illegal measure. It would do so on a most-favoured-nation
basis.53
This has the advantage of expanding trade, rather than restricting trade through sanc-
tions. Nevertheless, this has been used rarely, in part because the complaining and defending
parties may have widely divergent views on the appropriate level of compensation. Since
the entry into force of the WTO Agreement, formal compensation has been agreed in only
two disputes.54
Another form of compensation - not provided for in the current rules - would be the
imposition of fines on the non-implementing defending party. Indeed, the International
Financial Institution Advisory Commission recommended the use of fines:
If countries do not accept WTO decisions, injured parties have the right to retaliate by putting
restrictions on imports from the offending country or region. The injured country then suffers
twice - once from the restrictions on its exports, imposed by foreign governments, and again
when tariffs or duties raise the domestic cost of the foreign goods selected for retaliation. To
compensate for the injury done by others, we impose costs on ourselves as well as them. The
Commission proposes that, instead of retaliation, countries guilty of illegal trade practices
should pay an annual fine equal to the value of the damages assessed by the panel or provide
equivalent trade liberalization.55
While the concept of fines for non-compliance is not unknown in the international treaty
system,56 it does give rise to a number of issues. If a defending party insists that it has
complied with its WTO obligations, it may be completely unwilling to pay a fine, which
in turn would raise the issue of the enforceability of such a remedy. In addition, to whom
53. Since all WTO Members would have a right to benefit from such lower rates, the complaining party
would presumably seek compensation in sectors or tariff lines in which it is the principal supplier.
54. The first was in Japan - Taxes on Alcoholic Beverages: Mutually Acceptable Solution on Modalities for
Implementation, WT/DS8/17 Q'úy 30, 1997), at http://www.wto.org. The second was in the U.S. Copyright
Act dispute. As noted above, in the latter case temporary compensation has been negotiated, although (as of
January 2002) not yet provided.
55. Meltzer Commission Report, Mar. 3, 2002, available at http://www.house.gov/jec/imfymeltzer.pdf, at
107.
56. For example, the North American Agreement on Environmental Cooperation and the North American
Agreement on Labour Cooperation provide for fines ("monetary enforcement assessments"). Similar mecha-
nisms are found in the Canada-Chile Agreement on Environmental Cooperation and the Canada-Chile Agree-
ment on Labour Cooperation. To date, none of these provisions have been applied.
VOL. 36, NO. 1
138
would such a fine be paid? To all WTO Members? To the complaining party a
the government of the complaining party be entitled to distribute the proceed
fine to the affected industry?
In any event, in many disputes the provision of compensation or the paymen
may do little to resolve the core basis of the dispute, which is the continued m
of a WTO-inconsistent measure.
C. Rotating the Sanctions List: "Carousel" Retaliation
In May 2000, the U.S. Trade and Development Act 200057 was signed into law.
legislation included a provision (section 407) that requires the U.S. Trade Represe
(USTR) to modify regularly the list of goods to which retaliation would apply. The
has dubbed this law the "Carousel Retaliation Act," since it requires the list of pr
subject to sanctions to be rotated - like a carousel.
In revising the retaliation list, the law requires the USTR to "act in a manner
most likely to result in the country . . . implementing the [DSB] recommendations.
number of Members, particularly the EC, strongly oppose the notion of "carousel,"
cipally on the basis that it could lead to the imposition of sanctions above the le
thorized by the WTO. They note that when punitive tariffe are applied, it may not
additional customs duties for the importing country - instead, it can stop trade
product. Therefore, they argue that a rotation of sanctions may have a cumulative ef
excess of the level of suspension authorized by the DSB.
In June 2000, the EC requested WTO consultations on carousel, the first step
dispute settlement process.59 However, this dispute has not yet moved to the panel
in part because, to date, the United States has not applied the legislation. No rota
sanctions has taken place.60
D. Collective Retaliation
The notion of "collective retaliation" was first raised in the GATT in the mi
when a number of developing countries argued that they lacked the ability to infl
mercially meaningful retaliation on the economies of large developed countries. Th
the adoption of a new system that would require all countries, not just the compla
retaliate against the non-implementing party.61
The idea of collective retaliation raises a number of problems, including estab
system to ensure that the sanctions applied by the entire WTO community do no
the specific amount authorized by the DSB (the "double-counting" problem). M
57. Trade and Development Act 2000, Pub. L. 106-200, 114 Stat. 251.
58. Id.
59. United States - Section 306 of the Trade Act of 1974 and Amendments Thereto: Request for Consul-
tations by the European Communities, WT/DS200/1 and G/L/386 (June 13, 2000), at http://www.wto.org.
60. Robert Zoeluck, United States Trade Representative, had earlier threatened to rotate sanctions if the
EU proceeded with its plan to implement a "first come, first served" transitional mechanism for the importation
of bananas. However, given the April 2001 U.S.-EU settlement of the Bananas dispute, and the concomitant
EU decision not to implement a "first come, first served" system, any rotation of sanctions has now presumably
been placed on the back burner.
61. Hudec, supra note 49, at 392-93.
SPRING 2002
139
there could be possible trade diversion effects, in that imports may flow to a count
despite its obligations - has declined to impose the retaliation.
More to the point, however, is that an amendment would be required to the
order to implement the principle of collective retaliation, and many countries wo
agree to this. As argued above, it is extremely difficult for a complaining party to
sanctions in a way that does not hurt its own consumers and businesses. Many gov
would be even more reluctant to impose commercial harm on their economies in o
resolve a completely unrelated trade dispute.
E. Punitive Retaliation
Some have argued that in order to give sanctions more commercial force, the W
should have the ability to impose sanctions at a punitive level, instead of having them sim
match the level of nullification or impairment sustained by the complaining party. Pun
retaliation is not permitted under current WTO rules.62
However, punitive retaliation would significantly compound the problem of the c
mercial harm that sanctions impose on the economy of the complaining party, wit
assurance that it would actually induce compliance. Moreover, punitive sanctions may
gender sufficient resentment in the defending party so as to stiffen its resolve not to
to this form of pressure.63
V. Conclusions
At this relatively early stage in the development of the WTO dispute settlement system -
slightly over seven years since the entry into force of the Agreement - a number of conclu-
sions may be drawn about the compliance and retaliation in the WTO. First, in general
terms, the record of compliance with WTO rulings has been very good. Most defending
countries, in most cases, have fully implemented the panel and Appellate Body reports.
Many countries clearly view it in their interest to promote the integrity of a rules-based
system, even if this means occasionally having to accept and implement an unpalatable
decision.
Second, the record of compliance tends to be better where the dispute is exclusively or
primarily a commercial one. Implementation problems are more likely to arise in cases that
also involve non-trade concerns. For example, the EC commitment to assist the economic
development of former European colonies in the Africa, Caribbean, and Pacific (ACP)
region significantly complicated efforts to seek full EC compliance with the WTO rulings
in the Bananas dispute. Thus, the so-called "trade and" disputes have resulted in a higher
number of compliance problems.
Third, there has been relatively little use of compensation as an alternative to retaliation.
In principle, it is widely accepted that compensation is clearly preferable to retaliation.
However, compensation negotiations have often been unsuccessful. In many cases, the dis-
62. The arbitrators in the 1999 Bananas case noted that there was "nothing in Article 22.1 of the DSU, let
alone in paragraphs 4 and 7 of Article 22, that could be read as justification for counter-measures oí* punitive
nature." (Emphasis in original) Bananas: Decision by the Arbitrators (1999 U.S. complaint), supra note 12, para
6.3.
63. See Hudec, supra note 49, at 392.
VOL. 36, NO. 1
140
pute will remain unresolved until the illegal measure has been withdrawn. T
of compensation, particularly in unrelated sectors, may do little to address this
Moreover, even where there is a will to negotiate compensation, the comp
defending parties may have markedly different ideas as to what would constitu
priate level of compensation. The compensation negotiations typically begin af
iation request has been made, when there is usually no agreement as to the lev
lification or impairment" sustained by the complainant. However, in the Copyrig
as noted above, the United States and the EC held a special arbitration durin
mentation period on the level of nullification or impairment. This facilitated th
negotiation of a temporary compensation agreement between the parties. This
esting and constructive precedent, which will likely be followed in the future.
Fourth, the nature of retaliation has undergone a certain evolution since its
cation by the United States in Bananas in 1999. In most cases, complainants
what might be regarded as "traditional" retaliation, in the form of punitive tariff
imported goods of the defendant. However, there have been cases in which
tional" retaliation has been sought, such as the requests made by the EC and Ja
laws to "mirror" the WTO-inconsistent U.S. Anti-Dumping Act of 1916. Sim
proposed EC retaliation in the Copyright Act dispute - invoked in the event th
Congress fails to fund the compensation package - would target U.S. copyright
imposing an additional fee for enforcement of intellectual property rights in
both these cases, it would seem that the complaining parties have sought to
withdrawal of the WTO-inconsistent measure by targeting the industries or s
they regard as the principal beneficiaries of it.
Fifth, it is fair to say that retaliation has been less successful in inducing com
anticipated by the drafters of the DSU. At the conclusion of the Uruguay R
officials were optimistic that the mere threat of automatic retaliation wou
powerful tool to force recalcitrant defending parties to comply with their oblig
ever, there have been disputes in which retaliation has either been threatened
implemented, and yet the defending party has done nothing to comply. To dat
have been relatively few in number. However, the effective enforcement of DS
essential to ensure the continuing integrity of the WTTO dispute settlement sy
The Uruguay Round brought about a number of significant improvements to
of trade dispute settlement. Nevertheless, it is evident that the much-heralded
"automaticity" has its limitations. The automatic adoption of panel and App
reports, and the automatic approval of retaliation requests, are no guarantee o
compliance.
SPRING 2002
141
Compliance Proceedings under Article 21.5 of DSU
and Doha Proposed Reform
Tsai-Yu Lin*
I. Introduction
While the Dispute Settlement Body (DSB) within the World Trade Organization (WTO)
dispute settlement system lacks coercive powers to enforce recommendations or rulings, it
is generally accepted that the credibility of a DSB judgment itself, rather than any direct
remedy, promotes compliance.1 Such observance has, to a large extent, reflected the pre-
vailing view of WTO Members that the Dispute Settlement Understanding (DSU) has
generally functioned well to date. Existing practices demonstrate that most of the recom-
mendations and rulings of the DSB have been implemented by the Member concerned,
without the occurrence of compliance disputes.2 Nevertheless, there have been ever-
increasing numbers of disagreements with respect to implementation of recommendations
and rulings since the first article 2 1.5 panel convened in 1998.3 Some troublesome disputes
involving bananas and Foreign Sales Corporations (FSC), for instance, which are relevant
for the relationship between the United States and the European Community (EC), have
in particular received enormous publicity and caused unnecessary delay. Implementation
problems are likely to become some of the important challenges to the efficacy of today's
entire dispute settlement process.
*Professor, Department of International Business, Soochow University (Taiwan); LL.B., National Chengchi
University (Taiwan); LL.M., Edinburgh University (U.K.); Ph.D., National Chengchi University (Taiwan).
Email: kry.tylin@msa.hinet.net.
1 . John H. Jackson, Designing and bnplementing Effective Dispute Settlement Procedures: WTO Dispute Settle-
ment, Appraisal and Prospects, in The WTO as an International Organization 170 (Anne O. Kruege ed.
1998).
2. According to statistical information on recourse to WTO dispute settlement procedures provided by the
Secretariat, between January 1, 1995 and September 30, 2004, there were eighty- two panel reports and 125
Appellate Body reports adopted by DSB individually. JOB (O3)/225/Rev.l22 (Oct. 2004).
3. Panel Report, European Communities - Regime for the Importation, Sale and Distribution of Bananas, Recourse
by Ecuador, Guatemala, Honduras, Mexico, United States to Article 21. S by the European Communities, WT/DS27/
RW/EEC (Apr. 12, 1999) [hereinafter European Communities Bananas Panel Report].
915
142
9 1 6 THE INTERNATIONAL LAWYER
The compliance proceeding under article 21.5 of the DSU was introduce
creation of the WTO.4 Its mandate is to assess whether compliance measures
implementing Member bring about conformity with WTO law. Because t
set forth in article 21.5 are lacking in several respects, proposed amendments
have been advanced by certain Members in their attempts to reach successf
of ongoing DSU negotiations under the Doha Mandate.5 Of particular import
has proposed the inclusion of an article 2 Ibis on Determination of Complian
more comprehensive disciplines.
This paper will address the legal issues of DSU article 21.5 compliance pr
terms of current case law, as well as the proposed reform in the context of the
Section II will discuss issues of implementation of the DSB recommendations
Section III provides an overall review of article 21.5 proceedings. Since ar
not adequately set forth the procedures to be conducted, section IV focu
procedural issues arising from the jurisprudence created by the WTO. Sectio
alyze the foremost problem of sequencing between the relationship of ar
article 22. The concluding remarks of this paper will be provided in section
II. Implementation of the DSB Recommendations and

Rulings
A. Implementation Obligations and the DSB Recommendations and Rulings
The primary obligation for respondent Members to comply with recommendations an

rulings of the DSB has been established under the DSU. There are certain pertinent pr
visions in this regard. DSU article 3.7 states that
[i]n the absence of a mutually agreed solution, the first objective of the dispute resolution
mechanism is usually to secure the withdrawal of the measures concerned if these are foun
to be inconsistent with the provisions of any of the covered agreements. The provision
compensation should be resorted only of the immediate withdrawal of the measure is imprac
ticable and as a temporary measure pending the withdrawal of the measure which is inconsistent
with a covered agreement/'
4. Understanding on Rules and Procedures Governing the Settlement of Disputes, Apr. 15,1 994, Marrakes
Agreement Establishing the World Trade Organization, Annex 2, Legal Instruments - Results of the Urugua
Round, 33 I.L.M. 1125 (1994), available at http://www.wto. org/english/docs-e/legal-e^S-dsu.pdflhereinaft
DSU].
5. Paragraph 30 of November 2001 Doha Ministerial Declaration provides that
[wje agree to negotiations on improvements and clarifications of the Dispute Settlement Understand-
ing. The negotiations should be based on the work done thus far as well as any additional proposals
by Members, and aim to agree on improvements and clarifications not later than May 2003, at which
time we will take steps to ensure that the results enter into force as soon as possible thereafter.
World Trade Organization, Ministerial Declaration of 14 November 2001, WT/MIN(01)/DEC/l, 41 I.L.M

746 (2002). The DSU review started in 1997. Following the deadline that had been set for the end of M
2003 under the Doha Mandate lapsed, Members subsequently agreed to extend the deadline for the review
until the end of May 2004. However, the failure of the Fifth Ministerial Conference held in Cancun, in mi
September 2003, caused a further setback to overall negotiations, including DSU review negotiations, causin
the May 2004 deadline to be missed again. As part of the subsequent decision adopted by the General Counc
on August 1, 2004 on the Doha Work Programme - the so-called July Package - the mandate to continue th
negotiations has been renewed, but without a new target date being set.
6. DSU art. 3.7.
VOL. 39, NO. 4
143
COMPLIANCE PROCEEDINGS UNDER ARTICLE 21.5 OF DSU 917
Article 21.1 provides that prompt compliance with the recommendations

the DSB is essential in order to ensure the effective resolution of dispute
article 22.1 indicates that compensation or suspension is not to be preferre
mentation of the recommendation. By inference, article 26.1(b), which stat
no obligation to withdraw the WTO-consistent measure in the event of
complaint, may suggest that there is an obligation to withdraw the inconsist
of a violation complaint.8 For these reasons, one may say that the DSU has c
its intention behind those provisions - that recommendations and rulings
concluding that there is a WTO-violation, are binding upon the parties to t
shall be fully implemented.
As to the means of implementation of recommendations and rulings, t
Member would have the discretion to choose its preferred means among po
Although article 19 of the DSU provides that the panel or the Appellate Bod
ways in which the Members concerned could implement the recommendati
or the Appellate Body has usually been reluctant to suggest ways and meth
mentation.9 As a result, the real connotation of the DSB recommendations a
often be less clear and may cause a different version of interpretations for
general, the more room the WTO gives to Members in determining the scop
for implementation, the easier for the respondent Member to implement t
the less likely for the complaining Member to accept the result of the imp
is especially true when the proposed implementation measure involves a bu
islation process domestically, affecting activities of various sectors and inv
These factors may give the respondent Member more difficulties in ful
recommendations or rulings within a reasonable period of time. In controv
related to food, safety, or public health, the attitude of the general public tow
of the DSB might also become the crucial key in determining implement
At the extreme, the respondent Members may prefer to face retaliatory tr
adopted by the complaining Member or leave the issues unsolved and endles
B. Surveillance by the DSB
Article 21.3 of the DSU states that the Member concerned "shall inform the DSB of its
intentions in respect of implementation of recommendations and rulings of the DSB"
within thirty days of adoption of the final report.10 To enhance the efficiency of the dispute
settlement mechanism, the DSB shall continue to keep under surveillance the implemen-
tation of adopted recommendations or rulings. According to article 21.6 of the DSU, the
issue of implementing the recommendations or rulings may be raised by any Member at
7. Id. art. 21.1.

8. WTO, A Handbook on the WTO Dispute Settlement System 89 ( 2004).
9. DSU art. 19. In the United bates-DRAMS case, for example, the panel refused to suggest how the United
States should comply with the antidumping agreement, stating that "in light of the range of possible ways in
which we believe the United States could appropriately implement our recommendation, we decline to make
any suggestion in the present case." Panel Report, United States Anti-Dumping Duty on Dynamic Random Access
Memory Semiconductors (DRAMS) of One Megabit or Above from Korea, % 7.4, WT/DS99/R (Jan. 29, 1999); see
also Terence P. Stewart & Amy S. Dwyer, Handbook on WTO Trade Remedy Disputes: The First Six
Years (1995-2000) 394-95 (2001).
10. DSU art. 21.3.
WINTER 2005
144
9 1 8 THE INTERNATIONAL LAWYER
the DSB (including the winning Member) at any time following their adop
the DSB decides otherwise, the issue of implementation of the recommendati
shall be placed on the agenda of the DSB meeting after six months followin
establishment of the reasonable period of time pursuant to article 2 1.3. 12 It is
the DSB's agenda until the issue is resolved. Additionally, at least ten day
DSB meeting, the Member concerned is required to provide the DSB with a
report of its progress in the implementation.13
These status reports are expected to ensure transparency and give an incen
respondent to advance implementation.14 Other members, particularly the
theoretically may take the opportunity to observe the progress of impleme
decide whether the implementing Member has fully complied with the reco
or rulings. Nevertheless, the functioning of status reports is more limited th
should be in practice. In the Bananas case, for instance, the EC refused from
to be specific about its implementation plans when called upon to state its i
intentions. Despite the fact that the United States, the prevailing Member,
that the new measures proposed by the EC could not be regarded as implem
EC still maintained its implementation. The status reports of the EC to th
noted that significant progress was being made towards implementation.15 L
Beef Hormones case, the EC argued that it could hold a law, which was seen to b
with a requirement, for an adequate risk assessment by retrospectively en
necessary risk assessment and refused to remove its ban.16 The EC's first st
the DSB noted only that the Community had decided to launch a number
studies "with a view to assessing the implications thereof for the Commu
prohibition."17
Although requiring the respondent Member to provide the status report se
only way that the DSB can actually carry out its job, which is to survey the i
of adopted recommendations or rulings, it apparently has little efficiency.
DSB rules, the respondent Member is not required to identify the measures
11. Id. art. 21.6.
12. Id. art. 21.3. If it is impractical to comply immediately with the recommendations and rulings, the
Member concerned shall have a reasonable period of time in which to do so. A reasonable period of time can
be decided three ways: (1) Members concerned propose a period that is approved by the DSB; (2) If the DSB
does not approve the period proposed by the Members concerned, both Members should negotiate a certain
period of time within forty-five days after the DSB passes the recommendations and rulings of the panel or
the Appellate Body; and (3) If both Members fail to agree on a period, within ninety days after the DSB has
adopted the recommendations and rulings of the panel or the Appellate Body, the case will be rendered to
arbitration and the arbitrator with binding power will decide a reasonable period of time. The arbitrator
normally has ninety days from the date of the ruling to issue its ruling. WTO rules state that a guideline for
the arbitrator should be that the reasonable period should not exceed fifteen months from the date of the
ruling, although this can vary depending on the particular circumstances. Id.
13. DSU art. 21.6.
14. Handbook on WTO Dispute Settlement, supra note 8, at 79.

15. See Carolyn B. Gleason & Pamela D. Walther, The WTO Dispute Settlement Implementatio
A System in Need of Reform, 31 Law & Pol'y Int'l Bus. 721 (2000) (citing EC -Bananas- Status R
European Communities, WT/DS27/17 (July 13, 1998)).
1 6. Panel Report, EC Measures Concerning Meat and Meat Products (Hormones): Complaint by the
WT/DS26/R/USA(Aug. 18, 1997).
17. European Communities - Measures Concerning Meat and Meat Products (Hormones) - Status Rep
European Communities, WT/DS2 6/17, WT/DS48/ 15 (Jan. 14, 1999).
VOL. 39, NO. 4
145
remove or implement, nor is it required to specify any sort of implement

Imposing more stringent requirements for the respondent Member in the
implementation of DSB rulings may contribute an improvement in the surveill
III. An Overview of DSU 21.5 Proceeding

A. Functioning and Features of Article 21.5 Proceedings
As discussed in previous sections, disagreements as to compliance can easily arise de

pending upon the complexity of individual cases. When the respondent Member believ
that it has fully or partially followed the recommendations and rulings of the DSB, but th
original complaining party disagrees, an expedited procedure should be available. Thi
procedure is provided for in article 21.5.
Article 21.5 of the DSU states that
[w]here there is disagreement as to the existence or consistency with a covered agreement of

measures taken to comply with the recommendations and rulings such dispute shall be decided
through recourse to these dispute settlement procedures, including wherever possible resort
to the original panel. The panel shall circulate its report within 90 days after the date of referral
to the matter to it. When the panel considers that it cannot provide its report within this time
frame, it shall inform the DSB in writing of the reasons for the delay together with an estimate
of the panel within which it will submit its report.18
Based on the provision of article 21.5, the compliance panel must decide (1) whether
there is disagreement with a covered agreement of measures taken to comply with the
recommendations and rulings; (2) the existence of the implementation of measures; and (3)
whether the measures is consistent with the WTO rules.19 Wherever possible, the DSB will
bring the matter back to the original panel to review the quality of compliance. The panel
should decide the matter in an expedited fashion, normally within ninety days. If the panel
finds that it cannot provide its report within this time frame, it shall inform the DSB in
writing of the reasons for the delay, together with an estimate of the period within which
it will submit its report.
In summary, besides having different tasks, an article 21.5 panel procedure has other two
basic differences from the normal proceedings.20 First, the original panel has a six-month
period to issue its final report under article 12.8, whereas an article 21.5 panel has to
circulate its report within ninety days.21 Second, under the normal procedure, the respon-
dent Member may resort to article 21.3 to gain a reasonable period of time to implement
original recommendations and rulings. In contrast, article 21.5 does not provide such grace
periods.22 The concerned Member does not seem to be entitled to any further period of
time for implementation following the DSB's adoption of the compliance panel report. In
the aftermath of the conclusion that a WTO-inconsistent measure exists, the complainants
may directly request for the authorization to suspend concessions toward the respondent
under article 22.
18. DSU art. 21.5.
19. Jason E. Kearns & Steve Charnovitz, Adjudicating Compliance in the WTO: A Review of DSU Article
5 J. Int'l Econ. L. 331, 332-34 (2002).
20. Id.
21. Mat 333.
22. Id.
WINTER 2005
146
92 O THE INTERNATIONAL LAWYER
The features described above reveal that time is of the essence for the en
cedure. This is in line with the fundamental requirement of prompt comp
recommendations and rulings expressed in both article 3.3 and article 21.1
It may also help the DSB fulfill its article 21.6 duties and re-establish th
predictable relationship between Members sooner.24 By utilizing article 21
complainant, after having prevailed in an original dispute, should not have
the entire DSU process once again to address the new measures for imple
the other hand, if recourse to lodge new claims is through normal proceedin
complainant may again be involved in the new process for the same matter
parties to the dispute under a longer time frame. The respondent Member m
process strategically to continue practices that the original panel found to
such as gaining another compliance period for new recommendations and r
an instance, unnecessary postponement of the settlement of disputes may
effective functioning of the WTO.
B. Some Facts Regarding Article 21.5 Disputes
As of July 2005, there have been twenty-five disputes rendered to article 21.5 proceeding
since the first case was requested in 1998. Disputed issues usually involve areas such as an
dumping, subsidies and countervailing measures (SCM), agriculture, and sanitary and ph
tosanitary measures (SPS). The average time between matter referral to the original pan
and circulation of the final report is estimated around 181.36 days,25 which is far long
than the ninety days required in article 21.5. The United States, Canada,26 and the EC a
regular users of article 21.5 as complainants.27
As noted in the introduction of this paper, the implementation disputes between t
United States and the EC have particularly drawn the most attention in the media, includ
United States - Countervailing Measures Concerning Certain Products from the European Com
munities (DS212); United States - Tax Treatment for "Foreign Sales Corporations'1'' (DS10
and European Communities - Regime for the Importation, Sale and Distribution of Banan
(DS16, DS27, DS158).28In the FSC and Bananas cases, in particular, a second article 21
23. DSU arts. 3.3, 21.1. Article 3.3 states that
[t]he 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 by measures taken by another
Member is essential to the effective functioning of the WTO and the maintenance of a proper balance
between the rights and obligations of Members.
24. Kearns & Charnovitz, supra note 19, at 333-37.
25. World Trade Law Home Page, http://www.worldtradelaw.net (last visited Oct. 30, 2005) (subscriptio
required).
26. The disputes for which Canada sought recourse, for instance, include United States - Final Dumping
Determination on Softwood Lumber from Canada (DS264); United States - Investigation of the International Trade
Commission in Softwood Lumber from Canada (DS277); United States - Final Countervailing Duty Determination
with Respect to Certain Softwood Lumber from Canada (DS257); Brazil - Export Financing Programme for Aircraft
(DS46); and Australia - Measures Affecting Importation of Salmon (DS18). See WTO, Dispute Settlement: Dis-
putes by Country, httpy/www.wto.org/english/tratop-e/dispu-e/dispu-by-country-e.htm (last visited Oct.
30, 2005) [hereinafter WTO Dispute Settlement].
27. Argentina, India, New Zealand, Malaysia, Korea, Brazil, Ecuador, Guatemala, Honduras, and Mexico
were also complaints in other cases. Id.
28. Id.
VOL. 39, NO. 4
147
COMPLIANCE PROCEEDINGS UNDER ARTICLE 2 1 .5 OF DSU 92 1
panel was appointed.29 In the FSC case, for instance, the EC argued that the Am
Creation Act of 2004 enacted by the United States for implementation - wh
the tax exclusion of the FSC Repeal and Extraterritorial Income Exclusion Act
to grant and maintain export subsidies and allow companies to benefit from the
beyond the compliance deadline.30 The EC thus views the Act as a failure of
States to fully implement the D SB recommendations and rulings and has soug
recourse in a second article 2 1.5 procedure.31 In addition, in the dispute betwee
Canada over export subsidies for the regional aircraft industry, the parties also
second article 21.5 procedure.32 Likewise, in the Canada-Dairy cases, New Zeala
United States requested an article 21.5 panel for the second time, as they thoug
Appellate Body had failed to settle definitively the dispute between them.33
Another noteworthy point is that, due to the anomaly in DSU rules, not
involving implementation have to undergo a compliance review of article 2
many cases, such as Beef Hormones (DS26),34 Anti-dumping Act of 1 91 6 (DS 1 62), 3
110(5) of the US Copyright Act?6 in which the United States and the EC are inv
requested authorization to suspend concessions without first seeking recour
21.5.
IV. Issues Relating to DSU Article 21.5 Proceedings
Since article 21.5 is relatively brief, there are a number of uncertainties that inevitably
flow from it. For instance, what does the phrase "through recourse to these dispute settle-
ment procedures" in article 21.5 refer to?37 To what extent do other procedural aspects of
the DSU apply to this process? Who may seek the compliance procedure? What are the
time-limits for a request for implementation review? What range of claims can be consid-
ered in compliance procedure? Before trade retaliation has been attempted to be authorized,
should the compliance procedure be taken in advance? In examining certain procedural
issues arising from the jurisprudence created by the WTO, the following sections will
analyze standing, scope of review, consultation, appeal, and the relationship between article
21.5 and article 22.
29. See Panel Report, United States - Tax Treatment for "Foreign Sales Corporations'" - Second Recourse to Article
21.5 of the DSU by the European Communities, WT/DS108/RW2 (Sept. 30, 2005) [hereinafter United States
Tax Treatment Panel Report]; European Communities Bananas Panel Report, supra note 3.
30. United States Tax Treatment Panel Report, supra note 29, at 1-3.
31. id. at 2-3.
32. Panel Report, Brazil - Export Financing Programme for Aircraft - Second Recourse by Canada to Article 21.5
of the DSU, WT/DS46/RW2 (July 26, 2001).
33. Panel Report, Canada - Measures Affecting the Importation of Milk and the Exportation of Dairy Products -
Second Recourse to Article 21.5 of the DSU by New Zealand and United States, WT/DS1O3/AB/RW2, WT/DS1 13/
AB/RW2 (Dec. 20, 2002).
34. See WTO Dispute Settlement, supra note 26, Dispute DS26 - European Communities - Measures Con-
cerning Meat and Meat Products (Hormones).
35. See WTO Dispute Settlement, supra note 26, Dispute DS162 - United States-Anti-Dumping Act of 1916.
36. See WTO Dispute Settlement, supra note 26, Dispute DS160- US-Section 110(5) of the US Copyright
Act.
37. DSU art. 21.5.
WINTER 2005
148
A. Standing
Article 21.5 does not indicate who may request review. Most cases thus far have been
commenced by the original prevailing complainant, except one. In the Bananas case, the
EC, being the respondent member, sought review through its own initiative.38 In consid-
ering the issue of parties to an article 21.5 proceeding, the panel noted that whether the
original respondent in a panel proceeding is, or should be, permitted under the DSU to
initiate an article 21.5 proceeding is not clear; however, even assuming such an action i
permitted, the panel did not believe that a finding of WTO consistency could be made
based on the submission made by the EC.39 In this sense, the panel indicated its position
that the language of article 21.5 itself does not preclude respondents from using the com-
pliance procedure to ensure their implementing measure of WTO consistency. Since article
21.5 merely states that where there is any disagreement, "such dispute shall be decided
through recourse to the dispute settlement procedures . . . ," and does not refer to any
requirement to have a nominator, the respondent should be allowed to initiate an article
21.5 process.40 In contrast, the compliance report of the Bananas case initiated by the EC
was never adopted by the DSB.41 Thus, the issue of whether the defaulting Member can
commence an article 21.5 proceeding remains questionable.
In its proposal for reform of the DSU, the EC suggested limiting the complaining party
as the qualified party to request an article 21.5 panel, excluding the possibility of the re-
spondent Member's bringing the case against itself.42 Japan holds the same position.43 In
this author's view, unless it does nothing at all, the implementing Member should have a
right to request an article 2 1.5 proceeding to rule whether it now complies with the WTO
This would not only facilitate the resolution of disputes, but also contribute to the estab-
lishment of judicial predictability and security sooner, which is the essence of prompt set-
tlement of disputes under the DSU. This is especially true if the sequencing of article 21.5
and article 22.2 proceedings can be expressly provided for in the revised DSU by requiring
38. See European Communities Panel Report, supra note 3.

39. The panel stated that
it is not clear from the provisions of Article 21.5 whether the original respondent in a panel proceeding
is, or should be, permitted under the DSU to initiate an Article 21.5 proceeding for the purpose of
establishing the WTO consistency of measures taken to implement DSB rulings and recommendations.
Assuming such an action is permitted, we note that in this proceeding, the European Communities
presents in its written submission only one summary paragraph (paragraph 2.15) listing aspects of its
prior banana import regime that it has changed in order to comply with the DSB's recommendations
and rulings. We do not believe that a finding of WTO consistency could be made on the basis of the
submission made by the European Communities in this case, as there is an insufficient discussion of
how the previously found WTO inconsistencies have been eliminated in a WTO-consistent manner.
A/., H 4.14.
40. DSU art. 21.5.
41. WTO Dispute Settlement, supra note 26, Dispute DS27 - European Communities - Regime for the Im-
portation, Sale and Distribution of Bananas.
42. DSB Special Session - Communication from the EC, Contribution of the European Communities and its
Member States to the Improvement and Clarification of the WTO Dispute Settlement Understanding, TN/DS/W/1
(Mar. 13, 2002) [hereinafter EC Proposal TN/DS/W/1]; TN/DS/W/38 (Jan. 23, 2003) [hereinafter EC Pro-
posal TN/DS/W/3 8].
43. DSB Special Session - Proposal of Japan, Amendment of the Understanding on Rules and Procedures Con-
cerning the Settlement of Disputes, TN/DS/W/32 (Jan. 22, 2003) [hereinafter Japan Proposal TN/DS/W/3 2].
VOL. 39, NO. 4
149
article 21.5 procedures to be completed before article 22 procedures are initi

event that the original complainants are reluctant to request an article 21.5 proc
implementing Member may have more incentives to seek recourse in an article
ceeding to reconfirm the WTO-legality of its modified implementing measure,
to gaining legal certainty faster. On the other hand, as was the case in Bananas,
original complainants refused to be involved and the implementing Membe
deemed non-compliant, there might not be any Member that would be willing
place an item on the DSB's agenda for adoption of the report, rendering the
adopted.44 Such technical difficulties remain to be further resolved.
The issue of whether the respondent Member's right to initiate an article 21.5
could apply to the situation of article 22.8 is also being considered under the Do
Article 22.8 states that "[t]he suspension of concessions or other obligations sha
porary and shall only be applied until such time as the measure found to be in
with a covered agreement has been removed."45 The proposal indicated in parag
article 22 in the revised DSU text put forward by the EC suggests that
[a]fter the DSB has authorized the suspension of concessions or other obligations p
paragraph 6 or 7 of this article, the Member concerned may request a terminat
authorization on the grounds that it has eliminated the inconsistency or the nulli
impairment of benefits under the covered agreements identified in the recomme
rulings of the DSB.46
In the event that the complaining party objects to withdrawal of such authori
procedures provided for in article 21.5 (proposed article 2'bis) shall apply. By r
Member concerned, the EC proposes that the respondent Member can reque
mination of retaliatory measures when it views that it has complied with the D
This proposal is quite commendable. The availability of such expedited proced
provide an incentive for implementing Members to terminate trade retaliation
pecially when the original complainants insist that the implementing measures
cient and are not willing to initiate compliance procedures.47
Nevertheless, before the proposed provision - which would apply the 21.5 pro
determine the lifting of retaliation - has been formally inserted into future DS
it can be equally applicable under the context of the DSU remains unclear. In th
Hormone dispute relating to the termination of retaliation, despite its judgmen
EC's new Directive was inconsistent with the SPS Agreement, Canada refused
an article 21.5 procedure to obtain a review of the new Directive's consistenc
recommendations and rulings of the DSB.48 Canada maintained the suspensio
ligations in relation to imports from the EC and continued to impose impor
excess of bound rates on imports from the EC.4<; Failing to reach agreemen
44. Jeff Waincymer, WTO Litigation: Procedural Aspects of Formal Dispute Settlem
Although, article 16.4 provides that the report shall be adopted unless the DSB decides by cons
do so, as a practical matter, some must seek to place an item on the DSB's agenda.
45. DSU art. 22.8.
46. EC Proposal I N/U5/W/1, supra note 42, at 18.
47. Kearns & Charnovitz, supra note 19, at 341-42.
4ö. Request tor Consultation by the EC, Lanada - Continued suspension of Uvligations in EL-normones Dispute,
p. 2, WT/DS321/1, G/L/714 (Nov. 10, 2004).
49. Id.
WINTER 2005
150
consultation, the EC thus requested the establishment of a panel pursuan

and 6 of the DSU and article XXIII of the General Agreement on Tar
Rather than seeking recourse directly to article 21.5 procedures to termin
the EC, being an implementing member, chose to re-initiate the normal d
procedure in relation to the continued suspension of obligations toward it
pletely new dispute settlement procedure might trigger another potenti
of proceedings and unexpected delay, continuously adding more perplexit
dous dispute that has been pending since the adoption of the original
and rulings of the DSB in 1998.
The issue of whether other WTO Members, other than parties to the
directly or indirectly affected by implementing measures, can launch an
cedure also requires clarification. The current language of article 21.5 on
agreement as to the existence or consistency with a covered agreement."51 L
not to limit this disagreement specifically to disagreement between parti
Moreover, article 21.6 also states that any Member may raise the issue of
of the recommendations or rulings at the DSB "at any time following th
Because they are intrinsically multilateral in nature, the recommendatio
the DSB are also the matters that may concern other Members.53 Thus, o
article 21.5, under the current text, does not prevent other WTO Membe
a compliance procedure to challenge the WTO-conformity of implem
taken by respondent members, including third parties that have never p
previous dispute settlement process. Nevertheless, as indicated in the
the complaining party may initiate 21.5 proceedings, excluding the possi
Member commencing the proceeding. In the author's view, article 21.5
mainly related to disagreement on measures taken by the respondent Me
with rulings. Only the original parties to disputes can truly be aware of t
disputes. If any Member, other than parties to the dispute, can launch th
review under the article 2 1 .5, it might make it difficult to go forward wit
As a new party to the dispute in the specific dispute of implementation, it s
how the third party that has never participated in the previous dispute sett
can proceed to the whole proceeding. Does the compliance report have a d
if any at all, on the original complainant if the original complainant in th
settlements refuses to get involved in the new proceeding? The author
proposal raises valid points.
In considering the related issue about whether other Members can partic
21.5 proceeding as a third party, including those that have never attend
settlement procedures, the WTO jurisprudence seems to have established i
practice. In the case oí Australia-Leather, the EC and Mexico became the n
50. Id. at 1.
51. DSU art. 21.5.
52. Mart. 21.6.
53. Joost Pauwelyn, Proposals for Reforms of article 21 of the DSU, in Preparing the Doha Developmen
Round: Improvements and Clarifications of the WTO Dispute Settlement Understanding 75 (Ernst-
Ulrich Petersmann ed., 2003); also in The WTO Dispute Settlement System 1995-2003 51-59 (Federic
Ortino & Ernst-Ulrich Petersmann eds., 2004).
VOL. 39, NO. 4
151
COMPLIANCE PROCEEDINGS UNDER ARTICLE 2 1 .5 OF DSU 925
parties.54 Similarly, Australia became the third party in the Brazil-Aircraft case.55 L
Australia, Canada, India, Jamaica, and Japan became third parties in the FSC case
The EC proposes that a third party may request to join the consultations and b
"an adequate opportunity to express its views" as to the dispute.57 However, the
does not specify whether those Members that have never attended any prior dispu
ment procedure can be joined. It is noteworthy that, in the remaining parts o
proposal to article 21.5, there is no provision on third parties. The EC seems to li
right of third parties to consultation, not extending third party rights to the pa
pellate review process. This may be unduly restrictive, especially given that thir
have an opportunity to attend various stages of normal dispute settlement proce
cluding consultation, panel, and appellate review processes. Correspondingly, t
believes that the reasoning behind the exclusion of third parties in the latter two
21.5 proceedings, which are of equal importance for Members, needs greater justif
B. Disagreement or Measures to be Examined
Article 21.5 exclusively refers to disagreements on measures taken to comply with

DSB recommendations and rulings; any other measures fall outside the scope of a co
ance panel. Thus, it is quite crucial for the panel to determine the meaning of the p
"measures taken to comply with the recommendations and rulings" in article 21.5.58
cording to the view expressed by the Appellate Body in the Canada- Air craft case, the p
"measures taken to comply" refers to "measures which have been, or which shou
adopted by a Member to bring about compliance with the recommendations and r
of the DSB."59 In principle, a measure taken to comply with the recommendation
rulings will not be the same as the measure that was the subject of the original d
There would be two separate and distinct measures60 To put it differently, "measures
to comply with DSB recommendations and rulings" should refer to those measures w
are adopted or should be adopted by the respondent Member in order to impleme
recommendations and rulings of the DSB; these are different from the prior measur
gave rise to the DSB recommendations and rulings in the first place.61
54. Panel Report, Australia - Subsidies Provided to Producers and Exponéis of Automotive Leather - Recou
Article 21.5 of the DSU by the United States, <fl 1.7, WT/DS126/RW (Jan. 21, 2000) [hereinafter Aus
Subsidies Panel Report].
55. Panel Report, Brazil - Export Financing Programme for Aircraft - Recourse by Canada to Article 21.5
DSU, «H 1 .8, WT/DS46/RW (May 9, 2000).
56. Panel Report, United States - Tax Treatments for ^Foreign Sales Corporations - Recourse to Article 21.5
DSU by the European Communities, <fl 1.11, WT/DS108/RW (Aug. 20, 2001).
57. EC Proposal TN/DS/W/38, supra note 42, at 8-9. Japan, in its proposal, mentions that when con
tations are entered into during the reasonable period of time, before 21.5 procedures are activated, eac
to the dispute shall afford to any third party that so requests an adequate opportunity to express its views
Proposal TN/DS/W/32, supra note 43. This suggestion was also inserted into proposals put forward b
Rica and Ecuador. DSB Special Session - Communication from Costa Rica, Proposal by Costa Rica - Third
Rights, TN/DS/W/12/Rev.l; DSB Special Session - Proposal by Ecuador, Negotiations on Improvement
Clarifications of the Dispute Settle?nent Understanding, TN/DS/W/33 (Jan. 23, 2003).
58. DSU art. 21.5.
59. Appellate Body Report, Canada - Measures Affecting the Export of Civilian Aircraft - Recourse by Braz
Article 21.5 of the DSU, <fl 36, WT/DS70/AB/RW ( July 21, 2000).
6U. Id.
61. See id.
WINTER 2005
152
In the Australia-Leather case, the original panel found that payments und
tract between the Government of Australia, Howe and Company Proprieta
and Howe's parent company Australia Leather Holdings, Ltd. (ALH) wer
consistent with the SCM Agreement.62 The panel recommended that Aust
those subsidies without delay.63 To implement the rulings of the DSB, Au
Howe to repay the Australian Government $A8.065 million and had term
sisting obligations under the grant contract. However, at the same tim
provided a new $A1 3 .65 million loan to Howe's parent company, ALH.64 Th
indicated its view that the measures taken by Australia to comply with the re
and rulings of the DSB were not consistent with the SCM Agreement
Australia argued that the 1999 new loan was not part of the implementati
ruling and recommendation and thus, not within the scope of the panel's
ence.66 In the view of this article 21.5 panel, the United States' request fo
clearly had identified both the repayment by Howe and the 1999 loan as t
issue.67 The 1999 loan was "inextricably linked to the steps taken by Aust
to the DSB's ruling in this dispute" and was therefore within the panel's term
Accordingly, in this case, the compliance panel was based entirely on the c
United States' panel request for establishment as far as implementing m
cerned, which comports with the general rule that "it is the complaining M
dispute settlement which establishes the scope of the measures before a pan
In the Australia-Salmon case, after Australia announced that it had alread
the dispute rulings, Canada asserted that Tasmania Island, an island belong
had imposed an import ban toward salmonids.70 To determine whether th
port ban on salmonids would be "measures taken to comply," the complian
that it "cannot leave it to the full discretion of the implementing Member to
or not a measure is one 'taken to comply.'"71 If so, as the panel explained,
an implementing Member could simply avoid any scrutiny of certain measures

panel, even where such measures would be so clearly connected to the pane
Body reports concerned, both in time and in respect of the subject-matter, th
observer would consider them to be measures 'taken to comply.'72
Another issue associated with the Tasmanian import ban concerns t

new measures that were enacted only during the article 21.5 procedure
tioned in the complaining member's panel request. Referencing decisio
62. Australia Subsidies Panel Report, supra note 54, UH 6.1-6.7.

63. id.
64. Id. 116.3.

65. Id.
66. Id. H6.1.
67. Id. U 6.4.
68. Id. U 6.5.
69. Id. 11 6.4.
70. Panel Report, Australia - Measures Affecting Importation of Salmon - Recourse by Canada to Article 22.2 of
the DSU, H 4.26 et seq., WT/DS18/RW, (Feb. 18, 2000) [hereinafter Australia Salmon Panel Report].
71. Id. H 7.10(22).
72. Id.
VOL. 39, NO. 4
153
panels73 the Australia-Salmon compliance panel was of the view that the Tasmanian
fell within its mandate given that it was "implementing, subsidiary or so closely
measures that were specifically mentioned" and that Australia should have reason
pected that any further measures it would take to comply could be scrutinized by th
The panel did not consider that "measures taken subsequently to the establishmen
article 21.5 compliance panel should perforce be excluded from its mandate."75 R
stated that, "compliance is often an ongoing or continuous process and once it
identified as such in the panel request, as it was in this case, any 'measures taken t
could be presumed to fall within the panel's mandate."76
Unlike rulings by previous panels, the Australia-Salmon compliance panel ruling
the measures "taken to comply" in an innovative way.77 According to the view of
on the one hand, timing and the same subject-matter should be decisive in determ
measure "taken to comply" and such measures should be clearly connected to
Appellate Body reports.78 Hence, not all measures specifically indicated in the pan
would definitely fall within the terms of references of compliance panel and
examined. On the other hand, measures "taken to comply" under broad interp
should include future measures to be taken outside of the scope of the panel requ
cluding measures that occur during the article 21.5 procedures.79 An argument in
having panels consider such changes or modifications is that their rulings otherw
run behind the facts and could not provide positive solutions to the dispute.80 In the
view, this argument has its merits in terms of the objective of prompt complianc
in articles 3.3 and 21.1 of the DSU. In particular, by its nature, compliance is
ongoing or continuous process. If the measure currently in place could not be
by the panel, then its rulings might become incomprehensive and less meaningfu
versely, allowing such measures to be examined also requires consideration of
risks for the dispute settlement procedure. If the panels can go beyond the mand
time the panel was established based on panel requests, the respondent Membe
simply make subtle changes in their violations to avert the obligations arising fr
adverse ruling on it. In addition, when faced with certain measures that have
identified as such in the panel request, it would be questionable whether the resp
would be provided sufficient information in a timely manner about the scope of
Turning to proposals for reform, Korea suggests expanding the mandate of 21.
to provide that, "in the case of non-compliance, the Panel shall proceed to determ
73. See, e.g., Panel Report, European Communities - Regime for the Importation, Sale and Distribu
nanas- Complaint by the United States (Bananas III) '' 121-129, WT/DS27/R/USA(May 22, 1997
Body Report, European Communities- Bananas III, ' 140, WT/DS27/AB/R (Sept. 9, 1999); Pan
Japan- Measures Affecting Consumer Photographic Film and Paper, «fl 10.8, WT/DS/44/R (Apr. 22,
pellate Body Report, Australia- Measures Affecting Importation of Salmon, ' 90-105, WT/DS18/AB
1998); Panel Report, Argentina - Safeguard Measures on Imports of Footwear, Tffl 8.23-8.46, WT/DS12
25, 1999).
74. Australia Salmon Panel Report, supra note 70, "fl 7.10(26).
75. M «H 7.10(28).
76. Id.
77. See id. ^1.10.

/». Id.
79. Id.
80. Pauwelyn, supra note 53, at 77.
WINTER 2005
154
level of the nullification or impairment and circulate its report within 12

date of referral of the matter to it, including the level of the nullification o
One problem of such a suggestion might lie in the overlap that could a
jurisdictions of article 21.5 and article 22.6 procedures. The author believ
mandated under article 21.5 and article 22.6 should be distinguished an
the context of DSU.
In addition, the EC proposal limits the mandate of 21.5 procedures to disagreement

mainly between the complaining party and the defendants by adding the wording "dis-
agreement between the complaining party and the Member concerned."82 Given the nature
of bilateral focus, if applicable, whether such wording would affect the determination of
"measures taken to comply" in future practice remains to be observed.83
C. Consultation
The procedural issue of whether the complainant in an article 21.5 proceedin

satisfy the requirements of consultations under article 4.7 of the DSU also need
clarification. Some argue that, pursuant to the provisions of articles 4.3, 4.7, and 6
DSU, consultation should be a mandatory prerequisite for establishing the panel.84
this view, because rules governing consultations and the establishment of panel
distinguish between different types of panels, generally applicable rules must also
served in proceedings under article 21.5. However, others prefer the contrary vie
including a consultation requirement into article 21.5 would undermine prompt com
with the DSB recommendations and rulings and extend the period during which t
plaining Member suffers harm. In addition, those Members concerned must ha
aware of factual aspects of disputes during the consultations held before the establ
of the original panel, in which instance, the failure of consultation would not preju
respondent party's rights.
In most article 21.5 cases, prior consultations were not held. In the Mexico-HFCS
the Appellate Body did not confirm that this requirement applies to article 21.5 pro
leaving this issue open.85 However, it did confirm that it is possible, either explic
implicitly, to waive consultation between parties to disputes.86 In the US-FSC case
requested that the United States enter into consultations on November 17, 2000. 8
sultations were then held on December 4, 2000, in Geneva.88 Thus, the period durin
the consultation was held following the request was about twenty-seven days.89
81. DSB Special Session - Communication from the Republic of Korea, Contribution of the Republic
to the Improvement of the Dispute Settlement Understanding of the WTO, J E(13), TN/DS/W35 (Jan. 22
82. EC Proposal TN/DS/W/1, supra note 42, «fl 23 (Article Ubis).
83. See id.
84. United Nations Conference on Trade and Development, Dispute Settlement - WTO: Imple-
mentation and Enforcement 2 1 (2003), available at http://www.unctad.org.
85. Appellate Body Report, Mexico Anti-Dumping Investigation of High Fructose Corn Syrup (HFCS) From the
United States- Recourse to Article 21.5 of the DSU by the United States, Part IV, '' 35-75, WT/DS132/AB/RW
(Oct. 22, 2001).
86. See id.
87. WTO Dispute Settlement, supra note 26, DS108 - United States-Tax Treatment for "Foreign Sales C
porations": Summary of the Dispute.
88. See id.
89. See id.
VOL. 39, NO. 4
155
As discussed earlier, the EC, Japan, Costa Rica, and Ecuador, respectively, prop
consultation opportunities, during a reasonable period of time before 21.5 proced
commenced, should be given.90 Currently, each party to the dispute is only re
accord sympathetic consideration91 to any request from another party to the disp
consultation is thus not mandatory. From the perspective of facilitating prompt s
disputes, in the author's view, if the implementing party is willing to participat
consultation and sympathetically consider divergent views from another party a
implementation of DSB rulings during the reasonable period of time, mutual unde
between parties to the dispute might be achieved at an earlier stage.
As to the nature of the consultation requirement before a formal article 21.5 pa
cedure is initiated, based on the wording indicated in the proposal of the EC,
plaining party may request consultations.92 Under the EC's proposal, "[t]he M
which the request is made shall, unless otherwise agreed, reply to the request with
after the date of its receipt and shall enter into consultations in good faith withi
from the date of circulation of the request."93 In addition, Japan further clarif
proposal that consultations between the Member concerned and the complaining
desirable, yet are not required prior to a request for a compliance panel.94 In the
proposals for reform, the consultation requirement is thus not suggested as bein
datory prerequisite to the panel procedure, but rather a voluntary action by Mem
most cases show, parties to disputes are given the discretion to waive the consult
quest under the 21.5 proceedings; so far these proposals merely reflect the reality of
practice.
The author believes that consultation is valuable to the resolution of implementation
issues. The implementing measures that are the subject of consultation under the article
21.5 proceedings should be separate and different from the measures examined under the
original panel procedure. If parties to the dispute are given the opportunity to exchange
information and further narrow the scope of difference through consultations in good faith,
it would benefit the pursuit of a mutually satisfactory solution. Moreover, the respondent
party, to which the request is addressed, should not be entitled to refuse to consult or impose
any terms and conditions as a precondition to engaging in consultations when consultation
is requested. Otherwise, the non-mandatory consultation requirement might mean little,
especially in the event that the respondent party refused consultation based on any reason.
In considering the goal of faster compliance with article 21.5 proceedings, the timeframe
of consultation, as suggested by the EC, should be shorter than it is in the original pro-
cedure. Additionally, such timeframe should be given more flexibility based on mutual
agreements.
90. See supra text accompanying note 57.

91. The phrase "sympathetic consideration" is also seen in various other agreements; for instance, articles
9.2, 22, and 23 of GATT; article 4.2 of the DSU; and article 17.2 of the Anti-Dumping Agreement. General
Agreement on Tariffs and Trade arts. 9.5, 22, 23, Oct. 30, 1947, 61 Stat. A-ll, A-30, A-64, 55 U.NT.S. 194;
DSU art. 4.2; Agreement on Implementation of Article VI of the General Agreement on Tariffs and Trade
1994, Apr. 15, 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1, Legal In-
struments- Results of the Uruguay Round, 33 I.L.M. 1125 (1994), art. 17.2.
92. See EC Proposal TN/DS/W/1, supra note 42.
93. Id. <fl 23 (Article 2 Ibis).
94. See Japan Proposal TN/DS/W/32, supra note 43.
WINTER 2005
156
Another question relates to the time period for a request for consultatio
EC proposal, consultations can be requested at any time after the Member
(1) states that it does not need a reasonable period of time for complia
notification that it has complied with; or (3) thirty days before the expir
sonable period of time.95 Although the EC appropriately pointed out the
period for consultation that needs to be considered, the maximum time
cluded.96 If a new implementing measure has been adopted for a certain pe
years for example, and during that period no disagreement arises, can the co
start a consultation request after five years following the completion of
The absence of a provision on the maximum time period for consultation
create legal uncertainty, and should therefore be elaborated upon.
D. Appealing
Article 21.5 does not explicitly provide for the possibility to appeal. Whether the co
pliance panel report can be appealed was not clarified until the first appeal was accepte
the Brazil-Aircraft Appellate Body in May 2000.97 So far, the practice has shown that appe
against the compliance panel report are allowed and are used quite frequently.98 As to
timeframe of appellate review, article 21.5 only provides that the completion of proceed
is to be ninety days after the referral of the matter to the Panel - whether both the p
and Appellate Body are required to share the ninety days remains to be clarified.99
The EC supports the current case law on the issue of appeal, describing in its propos
that "in case the report of the compliance panel is appealed, the Appellate Body proceed
as well as the adoption of the Appellate Body report, shall be conducted in accordance w
article 17" of the DSU.100 The Japanese take the same position in their proposal.101 U
such proposals, all provisions of article 17 would apply to article 21.5 appellate proceedi
including article 17.5 that gives the Appellate Body sixty days to complete its review.10
the author's view, while the possibility of an appeal adds to the overall time taken to re
a dispute, transferring the sixty-day requirement of article 17.5 directly into article
situations is not in keeping with the essential time factor stressed in article 21.5 proceedin
An abbreviated appeal process should thus be taken into account. This is particularly t
because regulatory flexibility has been suggested in the proposals for consultation and
95. See EC Proposal TN/D S /VW 1, supra note 42.

97. Appellate Body Report, Brazil - Export Financing Programme for Aircraft - Recourse by Canada to Article
21.5 of the DSU, Part I, n 1-7, WT/DS46/AB/RW (July 21, 2000).
v«. lake me ivitLJUû-nt^ò case tor example, in me cover page or me panei report, me secretariat noted
that "[i]f the Panel Report is appealed to the Appellate Body, it shall not be considered for adoption by the
DSB until after the completion of the appeal." Panel Report, Mexico Anti-Dumping Investigation of High Fructose
Corn Syrup (HFCS) From the United States, WT/DS132/R (Jan. 28, 2000).
99. DSU art. 21.5.
100. EC Proposal TN/DS/W/1, supra note 42, «fl 23(8) (Article 2 ibis). Article 17.1 of the DSU provide
the "[t]he Appellate Body shall hear appeals from panel cases," and Article 17.6 provides that "[a]n appe
be limited to issues of law covered in the panel report and legal interpretations developed by the pane
arts. 17.1, 17.6.
101. See Japan Proposal TN/DS/W/32, supra note 43.
102. DSU art. 17.5.
VOL. 39, NO. 4
157
proceedings, which both differ from the original provisions of normal disput
procedures.
V. Sequencing Issues of Article 21.5 and Article 22
The DSU does not clearly specify the relationship between the article 21.
implementation and article 22.6 review of retaliation rights, creating what is k
sequencing problem. As noted above, article 21.5 merely requires a complian
complete its review within ninety days of the referral. Nothing in the provisi
the time limits for when the respondent Member should initiate such a proce
result, it is possible that the proceedings might be commenced before or after the
of a reasonable period of time for implementation of rulings. However, article
that, "the DSB . . . shall grant authorization to suspend concessions and other
within 30 days of the expiry of the reasonable period of time unless the DS
consensus to reject the request," or an objection to such request is raised and r
arbitration.103 Article 22.6 thus states an express time limit for DSB authoriz
requested and granted. Yet there is no provision to ensure that an article 21.5
is completed within thirty days following a reasonable period of time for com
the same time, article 22.6 is limited to cases where "the situation described in
2 occurs," which refers to the case where the Member concerned "fails to bring
found to be inconsistent with [the] covered agreement into compliance."104 A
the sequencing question arises as to whether an article 21.5 proceeding takes
over an article 22 proceeding. More specifically, before invoking trade sanctio
compliance authorized by the DSB, must the complaining Member first go thro
pliance panel process under article 21.5 to ensure the WTO consistency or in
of its implementing measure?105
In the Banana case, the United States asserted that it had the right to reques
zation from the DSB to suspend concessions and other obligations towards the
the EC's new banana regime adopted in response to the recommendations an
the DSB.106 According to the United States, a Member does not need to proc
article 21.5 before utilizing article 22.107 Article 22.2, on the one hand, states t
103. DSU art. 22.6.

104. Id. arts. 22.6, 22.2.
10). See bylvia A. Rhodes, lhe Article ¿l.y/ZZ Problem: Llanpcation through Bilateral Agreements?, 3 J. Intl
Econ. L. 553-58 (2000); Cherise M. Valles & Brendan P. McGivern, The Right to Retaliate under the WTO
Agreement: The "Sequencing" Problem, 34(2) J. World Trade 63 (2000); Petros C. Mavroidis, Proposals for Reform
of Article 22 of the DSU: Reconsidering the "Sequencing" Issue and Suspension of Concessions, in Preparing the Doha
Development Round: Improvements and Clarifications of the WTO Dispute Settlement Understand-
ing 78 (Ernst-Ulrich Petersmann ed., 2003).
106. See Arbitrators Decision, European Communities - Bananas III - Recourse to Arbitration by the Europe
Communities under Ankle 22.6 of the DSU, Part IV, «jffl 4.1-4.15, WT/DS27/ARB (Apr. 9, 1999) [hereinaf
Bananas Arbitrators Decision]. For a detailed analysis of the implementation problem of this dispute, see Ma
icio Salas & John Jackson, Procedural Overview of the WTO EC-Banana Dispute, 3(1) J. Int'l Econ. L. 145-6
(2000); John Jackson & Patrico Grane, The Saga Continues: An Update on The Banana Dispute and its Procedu
Offspring, 4(3) J. Int'l Econ. L. 581-95 (2001); Michelle M. Mulvena, Note, Has the WTO Gone Bananas? How
the Banana Dispute has Tested the WTO Dispute Settlement Mechanism, 7 New Eng. Int'l & Comp. L. Ann. 1
(2001).
107. See Bananas Arbitrators Decision, supra note 106, 1^ 4.1-4.15.
WINTER 2005
158
satisfactory compensation has been agreed within 20 days after [the end]
period of time, any party . . . may request authorization from the DSB t
concessions or other obligations."108 Article 22.6, on the other hand, pro
"shall grant authorization to suspend the concessions or other obligations w
[after] the reasonable period."109 Thus, if the complaining Member does
ceedings for authorization of retaliatory measures within ten days follow
period, it may miss the negative consensus deadline under article 22. Actu
proceedings were required before invoking the retaliatory right, the lat
mean little, on the ground that the compliance panel would be unlikely t
within ten days if such case was brought at the end of the reasonable per
tation. In sum, the United States contended that there is no sequencing ap
article 21.5 and article 22.110
The EC, however, strongly opposed the United States' interpretation believing that ar-
ticle 22 can only be utilized after the completion of article 21.5 procedures.111 For the EC,
if complainants were allowed to unilaterally judge the WTO-consistency of the EC's new
banana regime, and sanctions thereof were based on that judgment alone (the approach
taken by the United States), then the WTO multilateral review would be undermined.112
The DSB established two compliance procedure panels under article 21.5 to review the
legality of the EU revised regime on January 12, 1999. One was requested by Ecuador113
and the second by the EC itself.114 Instead of seeking any recourse under article 21.5, the
United States moved alone to directly invoke the article 22.2 process, requesting the DSB
to authorize the suspension of concessions towards the EC on January 14, 1999, which
involved a $US5.2 billion dollar deal.115 The EC then objected to the level of suspension
proposed and referred the matter to arbitration based on article 22.6 on January 29, 1999. 116
The Bananas case, rather bizarrely, established three separate panels - all with the same
members as the original panel and all working at the same time. The same panelists thus
made determinations under both article 22.6 and article 21.5 concurrently and strategically
delayed their article 22.6 rulings so that all three reports could be issued at the same time
in April 1999.
On April 6, 1999, the arbitrator made its decision that there had been SUS191.4 million
worth of nullification or impairment of the United States each year.117 The United States,
thus, was authorized by the DSB to suspend concessions against the EC on April 19. On
April 12, the panel also ruled that the EC's modified regime remained illegal.118
108. DSU art. 22.2.

109. DSU art. 22.6.
1 10. See Bananas Arbitrators Decision, supra note 106, Tffl 4.1-4.15.
111. Id. «114.11.
112. See id.
113. See Panel Report, European Communities - Bananas III - Recourse to Article 21.5 by Ecuador, WT/DS2
RW/ECU(Apr. 12, 1999).
1 14. òee European Communities Dananas ranel Report, supra note 3.
115. Recourse by the United States to Article 22.2 of the DSU, European Communities - Bananas HI, WT
DS27/43 (Jan. 14, 1999).
116. Request by the European Communities for Arbitration Under Article 22.6 of the DSU, Europe
Communities- Bananas III, WT/DS2 7/46 (Feb. 3, 1999).
117. Bananas Arbitrators Decision, supra note 106, % 7.8.
118. European Communities Panel Report, supra note 3, Part V.
VOL. 39, NO. 4
159
Because the Chairman of the DSB indicated that the solution to the Bananas matter
would be totally without prejudice to future cases relating to the relationship between 21.5
and 22,119 the debate does not cease after the bananas compromise solution. In authorizing
U.S. retaliation, the article 22 arbitrators did not follow the EC interpretation of article
22, rather they based their work only on the provisions of article 22.6, namely, to evaluate
the level of suspension following the arbitrators' determination of the WTO legality of the
measure in dispute. In other words, the article 22 arbitrators not only carried out work
relating to the tasks mandated by article 22.6, but also to those mandated by article 21.5.
In the meantime, there were also two other article 21.5 proceedings being processed for
the same subject of dispute.
In the author's view, without the clarification of the relationship between article 2 1.5 and
article 22, the Bananas arbitrators/panelists seemed to lean more towards the United States,
considering that the WTO compatibility determination mandated by the article 21.5 could
be performed through the article 22.6 arbitration process.120 The appropriateness of such
deliberation needs more thought. First, it might blur the line of application between article
21.5 and article 22. In the context of the DSU, besides the normal panel, panels established
under articles 21.3, 21.5, 22.6, and 25, as adjudicating bodies, perform different functions
mandated by the respective provisions. If an article 22.6 arbitration panel can perform the
tasks mandated by article 21.5, it seems to be more logical to infer that an article 21.3
arbitration panel - that is set-up to determine the reasonable period of time for implemen-
tation - can also perform the tasks under article 2 1.5 on the ground that the determination
of what constitutes measures taken to comply with the rulings of DSB should be made first
before the period of such measures taken to comply can be ensured.121 From this point of
view, can one say that an article 21.3 arbitration panel is more appropriate than an article
22.6 arbitration panel to perform the tasks of an article 21.5 procedure arbitrating panel?
If so, the functioning of each procedure in the context of the DSU would be more difficult.
The author does not consider that, based on the provisions of article 21.5 and article 22.6,
the separate tasks should be performed by the same original panel, or that individual Mem-
bers should regard one panel as a ground to mix the respective functions mandated under
other provisions. In particular, this is likely to occur in the event that the original members
are not available. In addition, the specific tasks of the arbitrator set out by article 22.7 do
not expressly include the consideration of WTO compatibility that is mandated exclusively
by article 21.5.
Secondly, to allow analysis of implementing measures under an article 22.6 arbitration
panel might deprive rights conferred on respondent Members under Article 2 1 .5. According
to article 22.6 and article 22.7, arbitration shall be completed within sixty days after the
expiration of the reasonable period of time and the parties shall accept the arbitrator's
decision as final, without the option to appeal. However, there are ninety days before the
completion of a panel process, and appeal by parties in dispute is generally accepted in the
WTO, which may extend the ninety period in practice. In such an instance, to save time
and facilitate implementation, complaining Members may be considered to have recourse
119. Dispute Settlement Body, Minutes on Meetings held on Jan. 25, 28, 29 & Feb. 1, 1999, WT/DSB/M/
54 (Apr. 20, 1999), at 30.
120. Similar consideration also appeared in another case. See Panel Report, United States - Import Measures
on Certain Products From the European Communities, WT/DS165/R (July 17, 2000).
WINTER 2005
160
under article 22.6 procedures. This would undermine the functioning ma

21.5.
In the Australia-Salmon case,122 the members concerned took the de facto sequencing
approach. While asking for a review of the article 22.2 retaliation right, Canada also re-
quested the establishment of an article 21.5 compliance panel.123 For Canada, the simul-
taneous approach may have been the effective way to prevent missing any rights conferred
from potential conflicts between article 21.5 and article 22. Following the consultation held
between Canada and Australia, both parties agreed that the article 22.6 arbitration process
initiated by Australia would be suspended until the article 21.5 panel was completed.124 If
the implementing measures were ruled as a W^TO-inconsistency by the article 21.5 panel,
both parties could then request the re-processing of arbitration regardless of whether either
party appealed. Arbitrators would make their decisions after the issuance of the compliance
panel report.125 Thus, under such initiative, an article 21.5 compliance panel may first per-
form its task of determining whether the implementing measure was consistent with W^TO
requirements, based on mutual agreements between parties.
This kind of approach may ensure that the examination of W^TO conformity is conducted
through a multilateral track, instead of unilateral judgment by the complaining party, before
the article 22.6 retaliation review into processing. However, there are certain points of the
initiative to be discussed. First, once the article 21.5 panel determines that Australia does
not fulfill the rulings of DSB, no matter whether the case is appealed or not, the arbitration
procedure would then begin to progress. Accordingly, even while an appeal of the compli-
ance panel report is pending, the arbitrators may be allowed to make decisions on the level
of suspension before the completion of appellate review. As a consequence, the lack of a
final decision relating to WTO conformity by article 2 1.5 proceedings remains an issue. It
might render the arrangements as to the sequencing of article 21.5 and article 22 less of an
issue. Secondly, article 22.6 states that arbitration shall be completed within the sixty days
following the end of a reasonable period of time. Whether the sixty days of completion of
arbitration can be counted from the date of the issuance of the article 21.5 panel report,
which is inconsistent with the conditions provide for in article 22.6, may need more justi-
fication. Thirdly, the initiative proposed in this case was for the solution of a specific dispute
and has practical implications. Nevertheless, the potential conflict arising from the legis-
lative flaws as to the relationship between article 21.5 and article 22 still remains. Also, the
rights of parties to dispute under the WTO cannot be ensured or certain in advance before
any mutual agreements are made. In the event of disagreement, further perplexity may
inevitably flow.
In the Australia - Leather case, without recourse to countermeasure granted by the DSB
based on article 4.10 of the SCM, the United States instead utilized article 21.5 and agreed
with Australia not to seek authorization to retaliate until after the article 21.5 proceedings
were completed.126 Such a model was also applied to the disputes between Canada and
Brazil regarding aircraft subsidies.127
122. See Australia Salmon Panel Report, supra note 70.

123. See id. «fl 1.3.
124. Id.
125. See id.
126. See Australia Subsidies Panel Report, suùra note 54.
127. Panel Report, Brazil - Export Financing Programme for Aircraft - Recourse by Canada to Article 21.5 of the
DSU, WT/DS46/RW(May 9, 2000); Panel Report, Canada- Measures Affecting the Export of Civilian Aircraft-
Recourse by Brazil to Article 21.5 of the DSU, WT/DS70/RW (May 9, 2000).
VOL. 39, NO. 4
161
Even though parties following the Bananas case have reached bilateral agr
how they will proceed in a practical way, ambiguity remains and may arise ag
at the proposals advanced in this context, Members appear to widely support th
between article 21.5 and 22. On September 29, 2000, Japan, supported by Cana
bia, Costa Rica, Ecuador, Korea, New Zealand, Norway, Peru, Switzerland, an
submitted a proposal to amend the provision of the DSU in order to clarify the
issue.128 The amended DSU would require a compliance panel to decide disa
over measures taken to implement a panel or Appellate Body ruling before Me
request WTO authorization to impose retaliatory trade sanctions.129 Australia h
vided clear and concise wording in its proposal, which provides that "the p
article 21.5 shall be completed before the procedures in article 22 are initiat
EC's position on this issue is that "after circulation of the compliance panel or t
Body report, a complaining Member . . . may request authorization from the D
pend the application to the Member concerned with concessions or other oblig
suant to article 22. "m Under the EC's proposal," [t] he DSB shall grant auth
such request only after the adoption of the compliance panel or the App
report."132
VI. Concluding Remarks
In the past ten years of dispute settlement practice under the DSU, an article 21.5 pro-
ceeding has become one of the pillars in the enforcement process and of vital interests for
Members. For the most part, the procedure set forth in proposed article Ubis reflects the
current practice under article 21.5. Perhaps the most significant change is the resolution of
the sequencing issue of article 21.5/22 that would allow the complainant to request DSB
authorization to retaliate only after an article 21.5 proceeding finds that the respondent has
failed to bring the measures found to be WTO-inconsistent into compliance with the WTO
Agreement. In addition, the proposal for the termination of retaliation measures through
article 21.5 proceedings may have important practical implications. Dealing with the ques-
tions of consultation and appeal on the basis of expedited process will also help prevent
undue delay in the implementation phase of dispute settlement.
Strengthening the function of article 21.5 through the elaboration and clarification in
the DSU text is a necessity. If a large number of ongoing disputes about the implementation
of the recommendations and rulings are allowed and the injuries that the prevailing com-
plainant has suffered cannot be redressed in a timely way, the effectiveness and credibility
128. Submission by Bolivia, Canada, Chile, Colombia, Costa Rica, Ecuador, Japan, Korea, New Zealand,
Norway, Peru, Switzerland, Uruguay and Venezuela for Examination and Further Consideration by the General
Council, Proposal to Amend Certain Provisions of the Understanding on Rules and Procedures Governing the Settlement
of Disputes (DSU) Pursuant to Article X of the Marrakesh Agreement Establishing the World Trade Organization,
WT/GC/W^lO/Rev. 1 (Oct. 26, 2001).
129. Id.
130. DSB Special Session - Communication from Australia, Negotiations on Improvements and Clarifications
of the Dispute Settlement Understanding, art. 21, «fl Sbis, TN/DS/W749 (Feb. 17, 2003).
131. EC Proposal TN/DS/W/i, supra note 42, «fl 23(9)(ii) (Article 21 bis).
132. Id.
WINTER 2005
162
of the dispute settlement system may thus be greatly undermined. While

DSU reform are going on under the Doha Mandate, no substantial prog
reached and no clear deadline for the conclusion of negotiations has yet
this, amendment of the DSU text along the line of proposals being advan
able. In the meantime, practical issues of article 21.5 interpretation may be r
a further development of dispute settlement practice.
VOL. 39, NO. 4
163
World Trade Review (2002), 1 : 2, 123–134 Printed in the United Kingdom
# Kym Anderson DOI : 10.1017/S1474745602001118
Peculiarities of retaliation in WTO

dispute settlement
 *
Centre for International Economic Studies, Adelaide University
Abstract : The dispute resolution procedures of the World Trade Organization allow
sanctions to be imposed when a country is unwilling to bring a WTO-inconsistent
trade measure into conformity. Apart from the fact that the procedure for triggering
the retaliation process has ambiguities that need to be removed, the retaliation itself
has some undesirable economic features. This paper looks at why compensation is
not preferred to retaliation and then examines five economic features of the
temporary trade retaliation that WTO may permit under certain conditions.
Both efficiency and equity concerns are raised. The paper concludes with some
suggestions for reforming this part of WTO dispute resolution during the review of
the Dispute Settlement Understanding that is due to be completed by May 2003.
One of the major achievements of the Uruguay Round of multilateral trade

negotiations was the creation in the new World Trade Organization of a much-
enhanced dispute settlement mechanism." Known formally as the ‘ Understanding on
Rules and Procedures Governing the Settlement of Disputes ’ (WTO, 1995), this body
of international trade law raised great expectations of much more definitive and
speedier resolution of trade disputes than under the law of the original General
Agreement on Tariffs and Trade (GATT). Even so, the possibility remains that a
WTO member found to have WTO-inconsistent policies may choose not to reform
those policies enough to satisfy complainants. If in that case a WTO Panel finds the
new policies still inconsistent with WTO rules, the complainant country is entitled
to compensation. Should an agreement on compensation not be forthcoming there
is provision, in Article 22 of the Dispute Settlement Understanding (DSU), for the
complainant to retaliate against the respondent. If the extent of retaliation is not
agreed between the parties, an arbitrator will determine the amount. The first two
* Correspondence : Centre for International Economic Studies, University of Adelaide, Adelaide SA 5005.
Tel : (j61 8) 8303 4712. Fax : (j61 8) 8223 1460. fkym.anderson!adelaide.edu.aug. The author served as a
Panelist and Arbitrator on the WTO Dispute Settlement cases involving the European Communities’ banana
import regime, 1996–2000. The views expressed here do not necessarily reflect those of the other Panel
members, the parties to the dispute, other WTO members, or the WTO Secretariat. Without implicating him,
the author is grateful for helpful comments from Petros Mavroidis on an earlier draft.
1 Jackson (1998, p. 176) went so far as to claim that the establishment of the Dispute Settlement Body in
the WTO ‘ is likely to be seen in the future as one of the most important, and perhaps even watershed,
developments of international economic relations in the twentieth century ’.
123
164
124           
uses of that provision were in the late 1990s (bananas and beef hormones) and
involved the two largest WTO members : the United States, as one of the
complainants, and the European Communities, as respondent.#
This provision to retaliate and the rules for doing so involve an apparent
procedural dilemma as well as some peculiar economic features. Now is the time to
reconsider those features of the rules, because the DSU has been under review as part
of the Uruguay Round’s built-in agenda and is now also part of the multilateral trade
negotiations launched in Doha in November 2001, which involves a commitment to
reach agreement on DSU improvements and clarifications no later than May 2003.
The present paper discusses first the apparent dilemma in the procedure for
triggering the retaliation process, then asks why compensation is not preferred to
retaliation, and then examines five economic features of retaliation. Those features
are the concept of equivalence, the inherent injustices that remain, the choice of
counterfactual to the WTO-inconsistent regime, the appropriate breadth of coverage
in the damage calculations, and the cross-retaliation issue. The paper then points to
the complications that globalization is adding to trade dispute resolution. Some
implications for reforming the WTO’s dispute settlement procedures are drawn out
in the final section.
1. The problem with the triggering process
If a WTO dispute settlement Panel or the WTO Appellate Body finds a member’s
trade policy to be not in conformity with WTO rules and the member’s
commitments, a ‘ reasonable period of time ’ is allowed for the policy to be brought
into conformity (DSU Article 21.3). In practice this period has ranged from three to
15 months. If the policy is not changed, consultations between the complainant and
respondent members should begin before that time period expires, with a view to
establishing mutually acceptable compensation. If no satisfactory agreement is
reached within 20 days after expiry of the ‘ reasonable period of time ’, the
complainant(s) can request authorization from the WTO’s Dispute Settlement Body
(DSB) to ‘ suspend concessions ’, that is, to retaliate (DSU Article 22.2).
Unless there is a consensus not to, the DSB will grant that authority, and within
30 days after expiry of the reasonable period of time. Should the respondent object
to the amount of retaliation proposed by the complainant, the matter is referred to
an Arbitrator for a decision within 60 days after expiry of the reasonable period of
time (DSU Article 22.6). The task of the Arbitrator (where possible the original Panel
that ruled on the WTO inconsistency in the first place) is to decide whether the level
of retaliation proposed is ‘ equivalent ’ to the level of damage (‘ nullification or
impairment ’). That decision by the Arbitrator is final : there is no appeal option or
2 There were no GATT-authorized retaliatory measures imposed during the 47 years prior to the
formation of the WTO in 1995. In the one case where it was authorized, for the Netherlands in 1952, the
authorized quota was never put in place (Valles and McGiven, 2000, p. 77).
165
Peculiarities of retaliation in WTO dispute settlement 125
opportunity to seek a second arbitration. The DSU will accept the decision and grant
authorization for retaliation, again unless there is a consensus not to (DSU Article
22.7).
Straightforward as that procedure may seem, it hides an apparent dilemma in the
DSU. Suppose the respondent takes the full ‘ reasonable period of time ’ before
announcing a reform of the offending policy measure. If the complainant believes the
reform is insufficient to make the policy WTO consistent, there is the opportunity to
refer the matter to a Panel (again, preferably the original one). The Panel in turn
must report within 90 days of that request (DSU Article 21.5). If the respondent is
unhappy with the Panel’s ruling, another 45 days could be required for the Appellate
Body to consider the matter. The apparent dilemma is that even if the Panel or
Appellate Body find the reformed policy still WTO inconsistent, the 20 days after the
‘ reasonable period of time ’ for a complainant to lodge a request to retaliate will have
expired. This interpretation of DSU Articles 21.5 and 22 suggests there could be an
endless loop of litigation.
An alternative interpretation is provided by Mavroidis (2000, pp. 795–799). He
argues (a) that Article 22.2 includes the possibility that the losing party does nothing
to bring its measures into compliance, in which case there is a 20-day deadline to
request countermeasures, and (b) that Article 21.5 deals with the situation where the
losing party does something to bring its measures into compliance and there is
disagreement as to the adequacy of that reform, in which case the 20-day period to
request countermeasures starts at the end of the Article 21.5 (Panel or Appellate
Body) proceedings if the finding is against the respondent. With this interpretation
there is no dilemma in law : the compliance Panel’s mandate under Article 21.5 is
simply to decide if the reform is adequate. If so, end of story ; if not, the complaining
party can request countermeasures.
Even this second interpretation leaves open the possibility for a respondent to
make an endless series of inadequate reforms – although the retaliatory measures
would remain in place until the complainant or a Panel is satisfied with the reform.
Clearly that possibility of very prolonged resolution could be tightened up through
a re-drafting of the procedural rules when the DSU is reviewed.$ The problem is now
well recognized by the WTO membership, and suggestions for remedial action have
been forthcoming.% Hence attention now turns to the economic features of
compensation and retaliation in cases where a policy measure has been found to be
still WTO inconsistent.
3 The various aspects of the sequencing problem that need to be reviewed, and the approaches adopted in
the interim by other Panels, are laid out by two Canadian officials in Valles and McGiven (2000). For
discussions by other trade law professors of how the WTO Panel and Arbitrator handled this dilemma in the
case involving the EC banana import regime, see Salas and Jackson (2000), Bhala (2000) and Komura (2000).
4 Mavroidis (2000, p. 799), for example, suggests three points for consideration by the DSU review : that
a request for countermeasures can be examined only if a compliance panel has previously pronounced the
implementing action inadequate ; that the compliance panel decision is final (not appealable) ; and that
respondents have only one ‘ reasonable period of time ’ to bring their measures into compliance with their WTO
obligations.
166
126           
2. Why is compensation not preferred to retaliation ?
A complainant unhappy with the respondent’s policy reform is entitled to seek

compensation until satisfactory reforms are implemented. If that compensation
comes in the form of a temporary lowering of the respondent’s import barriers on
some other products, the changes must be consistent with existing agreements. In
particular, they must be offered on a most-favoured-nation (MFN) basis.
In pure economic terms, compensation in this form is simply trade liberalization.
That boosts economic welfare in the respondent country, in the complainant
country, and even in third countries that export the products whose import barriers
have been lowered. Even if some third countries that import those or like products
were to lose from a terms-of-trade deterioration, we know from standard gains-
from-trade theory that the world as a whole will be better off economically.
Why, then, is compensation rarely preferred to retaliation, when the latter
typically involves the complainant raising its import barriers on products exported
by the respondent and thereby harming economic welfare in both countries and
globally ?
The answer has at least three parts. First, for reasons discussed in the previous
section, the current procedures do not guarantee a timely arbitration decision on the
extent of compensation that is warranted. Hopefully that will not be an issue
if\when the procedural rules in Articles 21.5 and 22.6 of the DSU are amended
following the review currently under way.
Second, the MFN feature of compensation means third countries also could
export more to the respondent, so the complainant would require a larger degree of
access than if access were able to be provided on a preferential basis. Even though
that greater openness would improve economic welfare for the respondent country,
the political economy of trade policy is such that the political leadership of the
country would lose from such unilateral reform – otherwise it would not have those
import barriers there in the first place (Grossman and Helpman, 1994).&
And third, if compensation were to be granted, the respondent would gain greater
control of procedures. Specifically, since the compensation is voluntary (DSU Article
22.1), the respondent can end it once the offending policy is further reformed,
pending the protracted outcome of any further action by the complainant under DSU
Article 21.5. With retaliation, by contrast, the complainant can keep that pressure on
the respondent until the latter satisfies an Article 21.5 Panel that the proposed
replacement of the offending policy is WTO-consistent and the new measure has
been implemented satisfactorily.
5 The fact that bilateral and multilateral agreements to lower import barriers are reached does not
contradict the fact that governments are unwilling to liberalize unilaterally. Those agreements can be seen as
a form of political gain from exchange in market access : Country A allows more of B’s exports into A’s market
on the condition that B allows sufficiently greater access to its market for A’s exports that the leadership in each
country gains at least as much extra political support from its exporters (and consumers) as it loses from its
import-competing producers who are exposed to greater competition from abroad (Grossman and Helpman,
1995 ; Hillman and Moser, 1995).
167
This is not to say that retaliation is desirable from the complainant’s viewpoint.
On the contrary, it is a last-resort action with peculiar economic features, five of
which are considered below.
3. Five features of the economics of trade retaliation
The concept of equivalence

As noted above, DSU Article 22.7 requires the Arbitrator to determine whether the
level of retaliation for which authorization is being sought is ‘ equivalent ’ to the
damage caused by the WTO-inconsistent measure. The type of retaliation most
commonly considered involves the complainant listing a range of products it imports
from the respondent on which it will impose prohibitively high tariffs until the
respondent’s offending measures are brought into conformity. The gross value of
the imports to be prohibited, typically the average for the three most recent
representative years for which import data are available, should match the value of
the complainant’s imports excluded by the respondent’s WTO-inconsistent measure.
That is probably the simplest way of quantifying the on-going damage to the
complainant from the WTO-inconsistent part of the respondent’s policy measure
when that measure is an excessive import restriction.
However, it needs to be recognized that ensuring equivalence between the damage
and the retaliation in terms of the gross value of trade between the respondent and
the complainant does not mean that retaliation has the same economic welfare effect
on the respondent as the initial damage is having on the complainant. The bilateral
trade value necessarily exaggerates the negative effect on both parties’ economic
welfare, but it does not do so equally (except by coincidence).
To see this, consider the case of retaliation depicted in Figure 1. Suppose Dc is the
complainant’s import demand curve for a product imported from the respondent,
and Dw is the world’s import demand for it. If Sr is the respondent’s excess supply
curve for that product, then, in the absence of distortions and assuming perfect
competition, the international price of the product will be P and the quantity traded
will by Qw, of which Qc goes to the complainant. A prohibitive tariff which reduces
Qc to zero will cause the global demand for the respondent’s exported product to
shrink to Dw. This horizontal shift will be by less than the full extent of Dc because
the complainant will demand more close substitutes in the world market, driving up
their price and hence shifting out to some extent the demand curve of the rest of the
world for the respondent’s export product. The international price of that product
will fall to Ph and the volume of the respondent’s exports will fall from Qw to Qw,
that is, by less than 0Qc. The gross value of the imports to be prohibited is the area
PbQc0, whereas the respondent’s net loss of export earnings from this product is the
smaller area PeQwQwdPh. Meanwhile, the respondent’s economic welfare loss
because of that trade prohibition is just area PedPh, since the area under the
respondent’s export supply curve between points d and e represents costs of
168
128           
Figure 1. The international market for a product that was imported by a

complainant from a respondent being retaliated against
Product
price
Sr
b e
P
P´
d
Dw
Dw´
Dc
0
Qc Qw´ Qw Product
quantity
production which would not be borne if the respondent’s export volume was
reduced by QwQw.
The economic welfare loss to the respondent is larger absolutely, and relative to
the gross value of trade curtailed, the steeper (i.e., the less price elastic) is Sr, the
respondent’s export supply curve, between the free market equilibrium point e and
the constrained equilibrium point d. Hence the ratio of PedPh to PbQc0 would vary
across products. This underscores the point that ensuring the reduction in the value
of imports from the respondent, due to retaliation, matches the reduction in the
value of imports from the complainant, due to the WTO-inconsistent measure being
used by the respondent, will not be (except by coincidence) the same as ensuring
equivalent losses in economic welfare for the other party due to the import
restriction each is imposing.
Moreover, this retaliation is not costless to the complainant : it loses area abP in
Figure 1 from choosing to forego purchasing those import products from the
respondent. That economic welfare cost is smaller the flatter (or more price elastic,
e.g. because of the availability of close substitutes) is Dc, the complainant’s import
demand curve for the respondent’s export product, to the left of b. It is also offset
169
somewhat by the gains from trading more with suppliers of close substitutes to the
respondent’s products (not shown in Figure 1) – although that necessarily is smaller
than area abP, for otherwise the complainant would have been purchasing from
those other supplying countries in the first place.
Although political sensitivity is likely to be the main criterion for selecting
products to target, the net economic cost of retaliating to the complainant relative to
the respondent would be lower the lower the listed products’ export supply elasticity
between e and d and the higher the complainant’s import demand elasticity between
a and b, ceteris paribus. That is, there is scope for the complainant to minimize its
economic welfare loss from retaliating, just as there is scope for the respondent to
mitigate its lost sales through diverting them to other countries.'
For all these reasons, trade loss equivalence would never translate into equivalent
damage to economic welfare, except by coincidence – and necessarily the com-
plainant will in addition lose economically during the retaliation period from the
import restrictions it imposes on the respondent’s trade.
More fundamentally, why does the DSU seek ‘ equivalence ’ anyway ? On the one
hand, the idea of legitimizing retaliation is contrary to the objective of reducing
impediments to trade. On the other hand, if the purpose of the retaliation is to induce
swift WTO compliance from the respondent, then perhaps some multiple of the
damage is the optimal level of retaliation to impose. Even the possibility that some
multiple of the damage might be authorized would reduce the tendency for members
to persist with WTO-inconsistent policies.
The inherent injustice of retaliation

Presumably the equivalence concept draws on notions of fairness (with parallels in
centuries-old international conventions on military wars), especially since large
economies have more scope to use this procedure effectively than smaller economies.
But that notion is violated under WTO retaliation rules in at least four respects.
First, past wrongs go uncompensated. That is, trade retaliation under WTO only
targets non-compliance after the ‘ reasonable period of time ’ has elapsed following
a Panel or Appellate Body finding against a respondent’s policy regime. The damage
done in preceding years to the complainant’s export industry is simply ignored by
DSU procedures.
Second, when the remedy chosen is a withdrawal of concessions by the
complainant towards the respondent’s exports, the complainant’s economy is not
helped but harmed by retaliation (the standard cost of protectionist barriers). For
small or developing economies confronting a much larger trader, that may well deter
them from seeking recourse through the DSU or, should the Arbitrator rule in their
6 Indeed the empirical evidence on past attempts to impose sanctions suggests the availability of close
substitutes and the scope for circumvention mean such trade bans have much less economic impact than the
gross value of trade might lead one to expect (see Hufbauer, Schott, and Elliott, 1990).
170
130           
favour, from adopting retaliatory measures (as was the case with Ecuador following
the decision reported in WTO, 2000).(
Third, retaliation does nothing to help the export industry that has been denied
market access by the respondent in the first place.) Rather, it is the complainant’s
import-competing industries that enjoy temporary assistance because of the
prohibitive retaliatory tariffs imposed.
And fourth, the respondent’s industries that are harmed by the complainant’s
retaliatory measures typically are not the industries that have been benefiting from
the WTO-inconsistent measure. Rather, they are the ones chosen by the complainant
typically with a view to having the largest negative political impact on the
respondent government.
The choice of counterfactual to the present regime

To measure the export damage, the present WTO-inconsistent import policy regime
needs to be compared with what it would be like once amended to become WTO
consistent. This raises the obvious difficulty that many alternative policy measures,
including even-more trade restrictive ones, could be WTO consistent. The
respondent is unlikely to be helpful in suggesting what the alternatives might be, for
one or other of them – including free trade – would have been implemented already
if they were not so unattractive politically. That leaves the complainant with the task
of proposing some counterfactual possibilities and estimating the damage under
those various scenarios. The estimated damage is considered to be the difference in
the value of the imports of this product set under the counterfactual regime(s) as
compared with under the present WTO-inconsistent regime.*
The breadth of coverage

Only direct exports from the complainant to the respondent of the product(s)
concerned should count. Indirect exports of the complainant’s produce embodied in
the final-good exports from another WTO member to the respondent are the
concern of that third country – otherwise, its third-party rights would be violated
(or, alternatively, there would be double counting of a portion of the damage to that
third country’s exports)."! If the product concerned is a service attached to goods
from another country, then the complainant’s trade would be measured at most
from the f.o.b. point of that third country. It may go beyond the c.i.f. point in the
7 This is not to say that small countries would be better off without a multilateral dispute settlement
process of course. On the contrary, Maggi (2001) provides a clear analysis of why bilateral agreements would
be inferior from the viewpoint of small countries.
8 This point was made as long ago as the eighteenth century by Adam Smith (1776, Book IV, Chapter II).
9 For a clear description of how such a calculation is made, see the report of the Arbitrator in the case of
the EC–US beef hormone case (WTO, 1999b).
10 Cases in point are the EC banana import regime arbitrations, whereby Ecuador was considered able to
retaliate for loss of goods trade in the total value of bananas up to the f.o.b. point while the United States was
not considered able to retaliate for loss of, say, fertilizer trade to Ecuador even though that fertilizer may have
been used in producing Ecuadorian bananas that could have been sold to the EC. See WTO (1999a, para. 6.12)
and WTO (2000).
171
respondent country, however, if some of those services exports are achieved by

commercial presence in that country (as with the United States in wholesale banana
markets within the EC – see WTO, 1999a). For the sake of equivalence, comparable
points in the marketing chain also should be used for measuring the trade flow to be
prohibited on the retaliation side.
The limitations on cross- retaliation
Another feature of economic and political significance in the retaliation procedures
has to do with DSU Article 22.3. That Article states that the complainant should
retaliate in the same sector wherever practicable. For example, retaliation against
a goods violation should be in goods. In the case of retaliation against a services
or TRIPs violation, however, DSU Article 22.3(f ) is even more precise. There are
11 services categories and nine TRIPs categories separately identified, and the
retaliation should where practicable match the services or TRIPs category of the
violation. Where there are multiple violations in more than one category, the re-
taliation is expected to be in the same categories in similar proportion.
From an economic viewpoint this raises two puzzles. First, why are services or
TRIPs so finely subdivided when their trade importance is so much less than goods,
which are treated as a single category ? And, second, why not allow the possibility of
cross retaliation so as to maximize the opportunity to encourage policy reform by the
respondent, and to minimize the limitation that within-sector retaliation poses
where the economies of the two parties are vastly different in size ?
While there are strict conditions attached to cross retaliation, that possibility is
not completely ruled out. The conditions under which it is allowed are laid out in
DSU Article 22.3. And already it has been used : in the case of the EC banana import
regime, it was recognized by the Arbitrator that Ecuador imported too few goods
from the EC to be able to retaliate in just goods, or even goods plus services, and so
a suspension of concessions relating to TRIPS was allowed"" (WTO, 2000).
4. Implications of globalization for trade dispute settlement
Globalization includes the phenomenon of economic integration resulting from the

fall in transactions costs of doing business across space, including across national
borders. The fall in barriers to trade in goods, services, financial capital and ideas has
been a consequence of declining transport and communications costs, most recently
because of the digital revolution, as well as trade and other regulatory reforms (see
WTO, 1998 ; World Bank, 1999). One consequence is that services and ideas are
growing in relative importance in the world economy and in global trade (see World
Bank, 1998).
This phenomenon is making the measurement of trade damage and hence
equivalent trade retaliation more complicated for the WTO compared with its
11 Mavroidis (2000, p. 804) points out that it is not clear from DSU Article 22.3 whether the Arbitrator
needed to be so intrusive or whether it could have been left it to the complainant to decide where to suspend
concessions.
172
132           
GATT predecessor when only goods were of significance. Even discerning a nation’s
economic interest in a dispute can become difficult now that services and TRIPs are
involved. For example, a country may have its services exports curtailed by another
country’s WTO-inconsistent policy measure in a way that restricts a third country’s
exports of goods to which those services are added, in which case the f.o.b. export
price and volume of that good fall.
A further complication arises if a country is an importer of a good, but a supplier
of services to the export of that good to third countries. As was true for the United
States in the EC banana import case, US economic welfare has been enhanced by the
lower international price of Latin American bananas, even though EC policies have
reduced US banana services exports that might have been attached to Latin
American banana sales to the EC.
Two other aspects of the spread of globalization also are relevant. One is the
growth of multinational corporations (MNCs) ; the other is the related disaggre-
gation of production into ever-more footloose intermediate production processes.
WTO trade dispute settlement procedures cause a problem for MNCs to the
extent that the share of their production outside their home country dominates, and
the number of countries in which they produce a particular product grows. This
stems from the WTO convention that a product is considered to be exported solely
from the country of final transformation, as defined by the World Customs
Organization’s classification. If the export-oriented production of an MNC is
outside its home country, then the host country’s government needs to defend it in
the WTO. If there are many such host countries, there is likely to be less interest
taken by any one of them."#
Globalization and the growth of MNCs also are generating an ever-finer
fragmentation of production processes that seek out the ever-changing lowest cost
locations globally. A growing share of the intermediate inputs and value added by
those plants is sourced from abroad (Feenstra, 1998). Hence even if a product
undergoes sufficient transformation to alter its customs classification, the vast
majority of its value could be due to the productive activities of another country. The
convention that the host country has 100 per cent of the right to retaliate against
another country’s WTO-inconsistent policy while the input-supplying country has
no right to retaliate is looking increasingly inappropriate as this aspect of
globalization proceeds.
5. Scope for reforming the DSU
The most obvious implication to come out of WTO arbitration so far is the need to
tighten current procedural rules in Articles 21.5 and 22 of the DSU. The problem is
12 In principle, there is also the possibility that the interests of a MNC from a non-WTO member country
could be defended by a host country that is a WTO member. In practice, however, this is becoming less and
less relevant as the WTO membership expands : as of early 2002, all but 5 per cent of global trade emanates
from WTO members.
173
now well recognized and alternative procedures are under consideration. Even with
those sequencing uncertainties removed, however, it is unlikely that compensation
will be preferred to the economically less-efficient temporary measure of retaliation
to entice a respondent to bring its trade regime into line with WTO obligations, for
the political economy reasons mentioned in section 2.
The economic purist might wonder why the value of imports curtailed, rather
than a valuation of the national economic welfare consequences of the import
barriers, is used to determine equivalence in retaliation to nullification or
impairment. The reason probably has much less to do with an understanding of
economics than with the relative simplicity of the traditional concept of trade
equivalence.
Meanwhile, the ethicist might wonder why the present peculiar concept of
equivalence is adhered to during the dispute settlement period. One reason to
question it is that a greater dose of retaliation could speed up the process of
becoming compliant – the apparent objective of the drafters of the DSU."$
A second equity concern is that retaliation does not help the complainant’s
exporters who have been and continue to be harmed, nor are the respondent’s
industries harmed by the retaliation the same ones that have been helped by the
WTO-inconsistent measure. Monetary compensation to the complainant from the
respondent may offer more scope for governments to target the transfers to achieve
a more-equitable outcome (and in the process to capture at least the full historical
cost of the WTO-inconsistent measure to the complainant, if not also to add a
punitive element).
A third equity concern is that small complainants are not as able as large ones to
inflict damage on large respondents. For that reason Horn and Mavroidis (1999,
page 22) suggest consideration be given to the notion that the rest of the WTO
membership contribute to the cost of small countries taking retaliatory action, so as
to reduce the risk of large traders remaining WTO inconsistent. This is part of the
bigger issue of technical and financial assistance for developing countries to
participate in the DSU more broadly, and indeed in all of the WTO’s activities.
Finally, with the ever-widening spread of multinational corporations and the
ever-greater fragmentation of production as globalization proceeds, national
governments are gradually becoming less appropriate bodies to defend the interests
of multinational corporations. Will the day come when firms will be allowed to
defend their trade interests directly at the WTO ? Meanwhile, such firms are likely to
come under increasing pressure to supplement the taxpayers’ money needed to cover
the complainant’s cost of dispute settlement activities that involve their pecuniary
interests – especially if, as was possibly the case in the US complaint against the EC
banana regime (Bhala, 2000, pp. 968–970), national welfare in the complainant’s
13 Note that the Panel considering the EC banana import regime case could not find anything in Article
22 of the DSU to justify punitive retaliation (WTO, 1999a). But note also that the International Law
Commission’s Draft of 16 July 1996 on State Responsibility does not consider compensation for damage
already incurred from an illegal activity as punitive (Horn and Mavroidis, 1999, p. 5).
174
134           
economy may have been boosted rather than harmed by the WTO-inconsistent
measure.
References
Bhala, R. (2000), ‘ The Bananas War ’, McGeorge Law Review, 31(4), Summer : 839–971.
Charnovitz, S. (2001), ‘ Rethinking WTO Trade Sanctions ’, American Journal of International Law, 95(4),
October : 792–832.
Feenstra, R. C. (1998), ‘ Integration of Trade and Disintegration of Production in the Global Economy ’,
Journal of Economic Perspectives, 12(4), Fall : 31–50.
Grossman, G. and E. Helpman (1994), ‘ Protection for Sale ’, American Economic Review, 84(4), September :
833–850.
—— (1995), ‘ Trade Wars and Trade Talks ’, Journal of Political Economy, 103(4), August : 675–708.
Hillman, A. L. and P. Moser (1995), ‘ Trade Liberalization as Politically Optimal Exchange of Market Access ’,
in M. Canzoneri et al. (eds.), The New Transatlantic Economy, Cambridge : Cambridge University
Press.
Horn, H. and P. C. Mavroidis (1999), ‘ Remedies in the WTO Dispute Settlement System and Developing
Country Interests ’, mimeo, University of Neuchatel, 16 July.
Hufbauer, G. C., J. J. Schott, and K. A. Elliott (1990), Economic Sanctions Reconsidered, 2nd edition,
Washington, DC : Institute for International Economics.
Jackson, J. H. (1998), ‘ Designing and Implementing Effective Dispute Settlement Procedures : WTO Disputre
Settlement, Appraisal and Prospects, ’ Ch. 5 in A. O. Krueger (ed.), The WTO as an International
Organization, Chicago : University of Chicago Press.
Komura, N. (2000), ‘ The EC Banana Regime and Judicial Control ’, Journal of World Trade, 34(5) : 1–87.
Maggi, G. (2001), ‘ The Role of Multilateral Institutions in International Trade Cooperation ’, American
Economic Review, 89(1), March : 190–214.
Mavroidis, P. (2000), ‘ Remedies in the WTO Legal System : Between a Rock and a Hard Place ’, European
Journal of International Law, 11(4) : 763–813.
Salas, M. and J. H. Jackson (2000), ‘ Procedural Overview of the WTO EC–Banana Dispute ’, Journal of
International Economic Law, 3 : 145–166.
Smith, A. (1776), An Enquiry into the Nature and Causes of the Wealth of Nations, New York : The Modern
Library (1994 reprint).
Valles, C. M. and B. P. McGiven (2000), ‘ The Right to Retaliate under the WTO Agreement : The
‘ Sequencing ’ Problem ’, Journal of World Trade, 34(2) : 63–84.
World Bank (1998), World Development Report 1998\99, New York : Oxford University Press.
—— (1999), World Development Report 1999\2000, New York : Oxford University Press.
WTO (1995), The WTO Dispute Settlement Procedures : A Collection of the Legal Texts, Geneva : World
Trade Organization, August.
—— (1998), Annual Report 1998, Vol. 1, Geneva : World Trade Organization.
—— (1999a), European Communities – Regime for the Importation, Sale and Distribution of Bananas :
Recourse to Arbitration by the European Communities Under Article 22.6 of the DSU, WT\DS27\
AB\R, 9 April, Geneva : World Trade Organization.
—— (1999b), European Communities – Measures Concerning Meat and Meat Products (Hormones) :
Recourse to Arbitration by the European Communities Under Article 22.6 of the DSU (US complaint),
WT\DS26\AB\R, 12 July, Geneva : World Trade Organization.
—— (2000), European Communities – Regime for the Importation, Sale and Distribution of Bananas :
Recourse to Arbitration by the European Communities Under Article 22.6 of the DSU, WT\DS27\
AB\RECU, 24 March, Geneva : World Trade Organization.
175
Compliance Problems in WTO
Dispute Settlement
William J. Davey t
This Comment surveys the problems of compliance facing the World
Trade Organization (WTO) dispute settlement system and considers
reforms that might improve compliance. In general, the WTO dispute set-
tlement system has an excellent compliance record. A recent examination
of the implementation record of WTO decisions for the first ten years of
WTO dispute settlement found a compliance rate of 83%.1 Although new
problem cases continue to arise, several of the ten problem cases outstand-
ing at the time of the study have since been resolved. 2 This compliance
rate is very good for an international state-to-state dispute settlement sys-
tem. 3 Moreover, the success rate of consultations in WTO cases that do4
not result in either adopted panel or Appellate Body reports is impressive.
The picture, however, is not so rosy if one looks beyond general statis-
tics and considers the quality and timeliness of compliance actions. By
"quality of compliance actions," I am referring to how a WTO decision was
implemented and whether the offending measure was withdrawn. If the
offending measure was withdrawn, then there probably was not a problem
with the quality of the compliance action. Alternatively, if the offending
measure was modified or replaced, the compliance action may not have
been truly satisfactory. When I refer to the "timeliness of compliance," I
am asking whether the implementing action was taken within the reasona-
ble period of time set for implementation. Timeliness also encompasses
inquiries into whether the time taken by the panel and appellate processes
t Guy Raymond Jones Chair in Law, University of Illinois College of Law.

1. William J. Davey, The WTO Dispute Settlement System: The First Ten Years, 8 J.
INT'L ECON. L. 17, 46-48 (2005) [hereinafter The First Ten Years].
2. JOHN H. JACKSON ET AL., LEGAL PROBLEMS OF INTERNATIONAL ECONOMIC RELATIONS
284 (5th ed. 2008).
3. Even though the WTO system has faced more challenging cases, the success rate
is comparable to the success rate of the GATT dispute settlement system. The First Ten
Years, supra note 1, at 48; see ROBERT E. HUDEC, ENFORCING INTERNATIONAL TRADE LAW:
THE EVOLUTION OF THE MODERN GATT LEGAL SYSTEM 344-51 (1991). The success rate in
the WTO system is clearly better than the success rate in the International Court of
Justice. See, e.g., Tom Ginsburg & Richard H. McAdams, Adjudicating in Anarchy: An
Expressive Theory of International Dispute Resolution, 45 WM. & MARY L. REV. 1229,
1308-12 (2004) (finding a 68% compliance rate, as defined by the authors, for a sample
of cases before the International Court of Justice).
4. William J. Davey, Evaluating WTO Dispute Settlement: What Results Have Been
Achieved Through Consultationsand Implementation of Panel Reports?, in THE WTO IN THE
TWENTY-FiRST CENTURY: DISPUTE SETTLEMENT, NEGOTIATIONS, AND REGIONALISM IN ASIA 98,
102-07 (Yasuhei Taniguchi et al. eds., 2007) [hereinafter Evaluating WTO Dispute
Settlement].
42 CORNELL INT'L LJ. 119 (2009)
176
Cornell InternationalLaw Journal Vol. 42
has met the standards specified in the WTO Dispute Settlement Under-
5
standing (DSU).
An examination of the quality and timeliness of compliance in the first
ten years of the WTO dispute settlement system reveals some interesting
patterns. General Agreement on Tariffs and Trade (GATT) and Trade-
Related Aspects of Intellectual Property Rights (TRIPS) cases typically
result in the timely withdrawal of the contested measure. 6 The two TRIPS
cases brought against the United States 7 and the European Communities -
Bananas8 case are the main exceptions to this trend. 9 In other words, the
desired result has generally been achieved in GATT and TRIPS cases.
There have also typically been timely withdrawals of the contested mea-
sures in safeguard and textiles cases; however, the contested measures in
these cases were often in place for all or most of the initially intended
period of effectiveness. 10 Thus, compliance was timely in terms of respect-
ing the reasonable period of time for implementation set by the WTO dis-
pute settlement process, but the overall WTO process took so long that
implementation was not very meaningful in practical terms. 1 In 75% of
trade remedy cases, the typical result has been a modification of the mea-
sure, which does not result in a significant change in the applied duty 50%
of the time. 12 Trade remedy cases often take a long time, and almost one-
half of the Article 21.5 compliance proceedings have involved trade remedy
cases. 13 Modifications and compliance disputes have also been common
14
in agriculture, subsidy, and Sanitary and Phytosanitary (SPS) cases.
Thus, in these cases-particularly in trade remedy cases-compliance is
often not timely and may not have much practical effect.
5. See Understanding on Rules and Procedures Governing the Settlement of Dis-

putes arts. 12, 16.4, 17.5, Apr. 15, 1994, Marrakesh Agreement Establishing the World
Trade Organization, Annex 2, Legal Instruments-Results of the Uruguay Round, 33
I.L.M. 1125 (1994) [hereinafter DSU].
6. Evaluating WTO Dispute Settlement, supra note 4, at 114-15, 139-40.
7. Appellate Body Report, United States-Section 211 Omnibus AppropriationsAct of
1998, WT/DS176/AB/R (Jan. 2, 2002) [hereinafter US-Section 211 AppropriationsAct];
Panel Report, United States-Section 110(5) of the US CopyrightAct, WT/DS160/R (June
15, 2000) (adopted July 27, 2000) [hereinafter US-Section 110(5) Copyright Act].
8. Appellate Body Report, European Communities- Regime for the Importation, Sale
and Distribution of Bananas, WT/DS27/AB/R (Sept. 9, 1997).
9. Evaluating WTO Dispute Settlement, supra note 4, at 115, 140. The two cases in
which the United States has been found to have violated the TRIPS Agreement required
Congressional action for implementation, which has not been forthcoming. Id. at 115.
The prevailing party in the two cases, the European Union, has never sought authority to
retaliate, perhaps because the cases involve a relatively limited amount of trade.
10. Id. at 110.
11. Id. at 113-14.
12. William J. Davey, Implementation of the Results of WTO Remedy Cases, in THE
WTO TRADE REMEDY SYsTEM: EAST AsiAN PERSPEcTIVES 33, 33-61 (Mitsuo Matsushita et
al. eds., 2006).
13. Dispute Settlement Body, Annual Report (2007), Addendum, Overview of the State
of Play of WTO Disputes, 92-95, WT/DSB/43/Add.1 (Dec. 7, 2007) [hereinafter DSB
Annual Report (2007)].
14. Id.; Evaluating WTO Dispute Settlement, supra note 4, at 114-15, 139-40.
177
2009 Compliance Problems in WTO Dispute Settlement
Overall, timely compliance occurred in about 60% of the cases.' 5

Trade remedies, subsidies, agriculture, and SPS cases were problem
areas. 16 Countries that failed to comply in a timely manner in the first
decade of WTO dispute settlement included the United States, the Euro-
pean Union, Canada, Japan, and Australia. 1 7 These countries often failed
to comply in cases against each other. 18 Accordingly, their on-time imple-
mentation rate was 50%. 19 Developing countries achieved a higher on-
time implementation rate of over 80%.20 Trends observed during the first
decade have continued during succeeding years, although trade remedy
cases now predominate in contested dispute settlement proceedings, 2 ' and
the United States has become the main source of untimely implementation
22
of WTO decisions.
Finally, with respect to the overall timeliness of the panel and appel-
late process, a detailed examination of the time taken by panels to issue
their reports shows that panels typically exceed the targets set in the DSU
23
by many months, especially in Article 21.5 compliance proceedings.
While the Appellate Body usually issues its report within ninety days of an
appeal, 24 the overall time taken by the process-especially when the "rea-
15. Evaluating WTO Dispute Settlement, supra note 4, at 137-38 (discussing country-
by-country implementation as of December 2004).
16. Id. at 114.
17. Id. at 113, 138.
18. Id.
19. Id.
20. Id.
21. See Appellate Body, Annual Report for 2007, 42-47, WT/AB/9 (Jan.30, 2008)
[hereinafter Appellate Body 2007 Annual Report].
22. The United States has reported to the Dispute Settlement Body monthly for
many years on its failure to implement in three cases. See, e.g., Dispute Settlement Body,
Minutes of Meeting Held in the Centre William Rappard on March 14, 2008, 11 1-50, WT/
DSB/M/248 (Apr. 30, 2008) (discussing Appellate Body Report, United States- Contin-
ued Dumping and Subsidy Offset Act of 2000, WT/DS217/AB/R, WT/DS234/AB/R (Jan.
16, 2003) (adopted Jan. 27, 2003) [hereinafter United States- Continued Dumping]; US-
Section 211 AppropriationsAct, supra note 7; US- Section 110(5) Copyright Act, supra note
7). Moreover, there is one case in which the European Union and Japan comment
monthly on the failure of the United States to implement or report. Id. C11 20-45 (dis-
cussing United States- Continued Dumping, supra). In addition, the United States has
recently lost three Article 21.5 compliance actions. See generallyAppellate Body Report,
United States- Subsidies on Upland Cotton, Recourse to Article 21.5 of the DSU by Brazil,
WT/DS267/AB/RW (June 2, 2008); Panel Report, United States-Laws, Regulations and
Methodology for CalculatingDumping Margins ("Zeroing"), Recourse to Article 21.5 of the
DSU by the European Communities, WT/DS294/RW (Dec. 17, 2008); Panel Report,
United States-Measures Affecting the Cross-Border Supply of Gambling and Betting Ser-
vices, Recourse to Article 21.5 of the DSU by Antigua and Barbuda,WT/DS285/RW (Mar.
30, 2007) [hereinafter US-Gambling]. Another compliance action is pending against
the United States as well. See WTO Secretariat, Update of WTO Dispute Settlement Cases,
77-78, WT/DS/OV/33 (June3, 2008) (discussing how Japan requested the creation of
an Article 21.5 panel to resolve compliance issues of United States-Zeroing (Japan)).
23. William J. Davey, Expediting the Panel Process in WTO Dispute Settlement, in THE
WTO: GOVERNANCE, DISPUTE SETTLEMENT & DEVELOPING COUNTRIES 409, 415-18,
420-21 (Merit E. Janow, Victoria Donaldson & Alan Yanovich eds., 2008) [Expediting
the Panel Process].
24. Id. at 418.
178
Cornell International Law Journal Vol. 42
sonable" periods of time for implementation are taken into account-is

quite long. 25 This Comment will not further consider the problem of the
length of proceedings, except to note that reforms are both needed and
26
feasible.
Taken together, the foregoing suggests that although the WTO dispute
settlement system may have an admirable record overall, considerable room
for improving the quality and timeliness of compliance exists. Indeed,
businesses have expressed concerns about non-compliance and delays as
reasons not to use the WTO dispute settlement system, which raises seri-
ous concerns for the future. 27 Thus, it is appropriate to consider the ques-
tion of what changes might be made to the system to address these
problems. Accordingly, I will briefly consider changes in compensation
and retaliation rules.
First, although compensation by definition does not produce compli-
ance, it is worth examining because it compensates for non-compliance.
Currently, the DSU provides for the possibility of negotiating compensa-
tion in lieu of retaliation within the twenty-day period following the expira-
tion of the reasonable period of time. 28 Though parties seldom use
compensation, there are several cases in which the parties agreed to com-
pensation. 29 This raises two questions: (1) whether it would be desirable
to promote more use of compensation and (2) how such greater use could
be facilitated in practical terms.
I think that it would be beneficial to expand the use of compensation.
It could be particularly desirable for developing countries if they find
themselves in a situation where timely implementation is not going to
occur and retaliatory action is not practical, as is typically the case.
Although I acknowledge concerns that this is tantamount to allowing coun-
25. Id. at 419-20.

26. Id. at 421-30.
27. See, e.g., Gary G. Yerkey, U.S. Poultry ProducersDo Not Plan to Urge U.S. to File
Case at WTO over EU Import Ban, INT'L TRADE DAILY (BNA), June 9, 2008 (noting EU
failure to comply in the Hormones case). In the DSU reform negotiations, Mexico has
been particularly critical of what it calls the "'fundamental problem' of the WTO dispute
settlement system, namely the 'period of time which a WTO-inconsistent measure can be
in place without the slightest consequences' to the offending party." Daniel Pruzin, Mex-
ico Presents 'Radical' Proposal for WTO Dispute Resolution Reform, 19 INT'L TRADE REP.
(BNA) 1984, 1984 (2002).
28. DSU art. 22.1-22.2.
29. Generally, compensation has been used when certain implementing measures
have been delayed beyond the end of the reasonable period of time set for implementa-
tion. See, e.g., Mutually Acceptable Solution on Modalities for Implementation, Japan-
Taxes on Alcoholic Beverages, WT/DS8/20, WT/DS10/20, WT/DS11/18 (Jan.12, 1998)
(Japan provided trade compensation to Canada); Notification of Mutually Acceptable
Solution, Turkey- Restrictions on Imports of Textile and Clothing Products, WT/DS34/14
(july 19, 2001) (Turkey provided trade compensation to India); Agreement Under Arti-
cle 21.3(b) of the DSU, United States- Definitive Safeguard Measures on Imports of Circu-
lar Welded Carbon Quality Line Pipe from Korea, WT/DS202/18 (July 31, 2002) (the
United States provided trade compensation to Korea); Notification of a Mutually Satis-
factory Temporary Arrangement, United States-Section 110(5) of the US Copyright Act,
WT/DS160/23 (June 26, 2003) (the United States provided monetary compensation for
three years of non-implementation).
179
tries-particularly rich countries-to buy their way out of obligations, I

think that, on balance, promoting this option-which would not be a
30
requirement-would be useful.
The more difficult question is how the use of compensation could be
facilitated. The most promising suggestion involves establishing a proce-
dure to set the level of nullification or impairment earlier than it is pres-
ently set.3 1 This procedure would facilitate the use of compensation
because the procedure would establish a more clear basis for conducting
compensation negotiations and there would be a concrete starting point-
and perhaps ending point-for negotiating the amount owed. The offend-
ing party, however, would still have to agree to provide compensation.
Although ideas positing ways that compensation might be "compelled"
exist, 3 2 these ideas are difficult to implement as long as the respondent is
recalcitrant.
An alternative solution contemplates improving compliance by
increasing the effectiveness of the WTO's ultimate weapon-retaliation-or
through other remedies. 33 It is worth noting that the overall effectiveness
of retaliation may be somewhat questionable. The third edition of the clas-
sic Hufbauer study on the effectiveness of economic sanctions finds that
economic sanctions work only about one-third of the time, although the
study does not look at economic sanctions taken for GATT/WTO-related
30. It is worth mentioning that although there are some notable exceptions, most
long-term non-compliance has occurred in disputes between developed countries. See,
e.g., Request to Join Consultations, Communication from Cameroon, European Commu-
nities- Regime for the Importation, Sale and Distributionof Bananas,WT/DS27/74 (Dec.
11, 2006) (dispute arising between Cameroon and European Communities); US- Gam-
bling, supra note 22 (dispute involving Antigua and Barbuda). However, poorer and
smaller developing countries might make greater use of the WTO dispute settlement
system if it seemed more probable that a useful result could be achieved. See William J.
Davey, The WTO Dispute Settlement System: How Have Developing Countries Fared?
24-40 (Ill. Pub. Law & Legal Theory Research Paper Series, Research Paper No. 05-17,
2005) [hereinafter How Have Developing Countries Fared?], availableat http://ssrn.com/
abstract=862804 (analyzing the use of dispute resolution by small and large developing
countries). So far, only large developing countries are significant users of the system.
Id.
31. Korea has proposed that Article 21.5 compliance panels set the level of nullifica-
tion or impairment in cases where they find non-implementation. See Special Session of
Dispute Settlement Body, Contributionof the Republic of Korea to the Improvement of the
Dispute Settlement Understandingof the WTO, c 9-10, TN/DS/W/35 (Jan. 22, 2003). It
would also be possible for panels to set that level in original proceedings.
32. See Joost Pauwelyn, Enforcement and Countermeasures in the WTO: Rules Are
Rules- Toward a More Collective Approach, 94 Am. J. INT'L L. 335, 337-38, 345-46 (2000)
(discussing methods to induce compensation payments).
33. Under WTO rules, if a WTO member fails to bring an offending measure into
compliance with its WTO obligations within the reasonable period of time set for imple-
mentation, the prevailing WTO member may ask the WTO Dispute Settlement Body for
authority to suspend concessions that it owes to the offending member and such author-
ity is to be granted absent a consensus to the contrary. DSU art. 22.2. Such suspension
of concessions is typically referred to as retaliation and usually consists of raising tariffs
beyond bound limits, often to prohibitive levels, on certain products from the offending
member. The DSU limits the level of suspension of concessions to the level of nullifica-
tion or impairment caused by the offending measure. Id. art. 22.4. If there is a dispute
over that level, it is resolved by arbitration. Id. art, 22.6-22.7.
180
trade reasons. 3 4 Retaliation was never formally used under GATT rules,
and retaliatory measures have been applied to date in only four cases
under WTO rules: European Community- Bananas,35 European Commu-
nity-Hormones,36 United States-Foreign Sales Corporation (FSC),3 7 and
United States Byrd Amendment 38.3 9 The Hormones retaliation has not
changed EU policy after almost nine years. 40 It is arguable that retaliation
in the other three cases, plus the possibility of retaliation under the safe-
guards agreement in the United States-Steel Safeguards case, 4 1 may have
had some effect on resolving those disputes, but it would be difficult to
establish that retaliation was the key factor in eventual implementation. 42
34. See GARY CLYDE HUFBAUER ET AL., ECONOMIC SANCTIONS RECONSIDERED (3d ed.
2007).
35. See Panel Report, European Communities- Regime for the Importation, Sale and
Distributionof Bananas, Recourse to Article 21.5 of the DSU by the United States, passim,
WT/DS27/RW/USA (May 19, 2008) (discussing the Untied States' retaliation
measures).
36. See Appellate Body Report, United States- Continued Suspension of Obligations in
the EC-Hormones Dispute, T1 12-15, WT/DS320/AB/R (Oct. 10, 2008) [hereinafter
US- Continued Suspension].
37. See generally Decision of the Arbitrator, United States- Tax Treatmentfor "Foreign
Sales Corporations",Recourse to Arbitration by the United States Under Article 22.6 of the
DSU and Article 4.11 of the SCM Agreement, WT/DS108/ARB (Aug. 30, 2002).
38. See generally Decision of the Arbitrator, United States- Continued Dumping and
Subsidy Offset Act of 2000, Recourse to Arbitration by the United States Under Article 22.6
of the DSU, WT/DS234/ARB/CAN (Aug. 31, 2004).
39. Retaliation was also authorized, but not imposed, in two of the cases arising out
of the Canada-Brazil dispute over export subsidies to regional jet makers. Arbitration
reports setting the permissible levels of retaliation were issued in the United States 1916
Act case and the United States- Gambling case, although the Dispute Settlement Body
has yet to be requested to authorize retaliation in those cases. JACKSON ET AL., supra note
2, at 365; see DSB Annual Report (2007), supra note 13, at 97, 99.
40. See US- Continued Suspension, supra note 36, '1 737 (noting that the Appellate
Body could not ascertain at this stage whether the European Communities had substan-
tially complied with Article 22.8 of the DSU).
41. Appellate Body Report, United States- Definitive Safeguard Measures on Imports of
Certain Steel Products, WT/DS248/AB/R (Nov. 10, 2003).
42. In the Steel Safeguards case, the threat of significant and imminent retaliatory
action by the European Union seemed to play a major role in the prompt removal of the
United States steel safeguards that had been found to violate WTO rules. Rossella
Brevetti & Christopher S. Rugaber, Bush Ends Steel Safeguard Tariffs in Face of Threat by
EU to Retaliate, 20 INT'L TRADE REP. (BNA) 2021 (Dec. 11, 2003). In the FSC case, some
congressional leaders and staffers cited actual and threatened European Union retalia-
tion to justify changing the contested measures, but the overall impact of the retaliatory
measures was not clear. Compare Kurt Ritterpusch & Gary G. Yerkey, Conferees Reach
Reconciliation Accord, Strip Some FSC Benefits to Offset Cost, 23 INT'L TRADE REP. (BNA)
720 (May 11, 2006), with Alison Bennett, Export Tax Repeal Conference Major Priority
with Only Days Left for Senate, Frist Says, 21 INT'L TRADE REP. (BNA) 1591 (Sept. 30,
2004) [hereinafter Export Tax Repeal], and Alison Bennett, Retaliatory EU Taxes Not Seen
Creating Expected Pressure to Pass Export Tax Bills, 21 INT'L TRADE REP. (BNA) 1193 (July
15, 2004) [hereinafter Retaliatory EU Taxes]. In the Byrd Amendment case, the role of
actual and threatened retaliation was less clear. The EC and Canada imposed retaliatory
sanctions in May 2005, while Mexico and Japan did so in August and others threatened
to do so. Rossella Brevetti & Michael O'Boyle, EC, Canada Move to Impose Retaliatory
Duties in Byrd Dispute, 22 INT'L TRADE REP. (BNA) 546 (Apr. 7, 2005); Michael O'Boyle,
Mexico Slaps Punitive Duties on U.S. Goods Due to Noncompliance with WTO Byrd Ruling,
22 INT'L TRADE REP. (BNA) 1386 (Aug. 25, 2005); Daniel Pruzin, Remaining Complainants
181
The reality is that the overall good record of the system is due mainly to the
good faith desire of WTO members to see the dispute settlement system
work effectively. 4 3 The more active users of the system are repeat players,
and they appear as both complainants and respondents. 44 It is in their
overall interest that the system functions effectively. However, there will be
cases in which such good faith cannot be relied upon, compliance becomes
an issue, and retaliation may be helpful to promote compliance.
Initially, it is important to bear in mind that the ultimate WTO remedy
of retaliation is prospective. Prospective retaliation gives the losing coun-
try an incentive to delay the time of reckoning as long as possible and prob-
ably explains the extensive delays in the system as well as the frequent use
of Article 21.5 compliance proceedings. However effective retaliation may
be when used by a major player in the WTO, retaliation is probably not an
effective remedy for a small or developing country (even if that country can
target sensitive large country sectors such as copyright holders). It is cer-
tainly the case that developing and small countries have never utilized the
45
possibility of taking such measures, even following WTO authorization.
There are several practical changes in the WTO remedy provisions that
offer hope for improving implementation of WTO dispute settlement deci-
sions. Three changes in particular should be given serious consideration.
The WTO remedies for non-implementation should incorporate (i) the pos-
sibility of substituting fines or damages as a remedy in lieu of suspension
of concessions; (ii) some degree of retroactivity, so as to help encourage
compliance within the reasonable period of time; 46 and (iii) some adjust-
ment mechanism to increase the level of sanctions over time, so as to pre-
Warn of Intent To Proceed With Byrd Sanctions by July, 22 INT'L TRADE REP. (BNA) 938
(June 9, 2005). Following a September GAO report showing that few firms benefited
from the Byrd Amendment, a House committee approved legislation repealing it, which
was finally approved in early 2006. Rossella Brevetti, Byrd Act Beset by Problems, GAO
Says; Five Firms Get 46 Percent of Distributions, 22 INT'L TRADE REP. (BNA) 1557 (Sept.
29, 2005); Rossella Brevetti, House Approves Budget Measure Containing Byrd Amendment
Repeal, 23 INT'L TRADE REP. (BNA) 184 (Feb. 2, 2006); Rossella Brevetti, House Ways and
Means Panel Approves Repeal of Byrd Law in Budget Measure, 22 INT'L TRADE REP. (BNA)
1747 (Nov. 3, 2005). While the retaliatory measures certainly played a role, the limited
number of beneficiaries of the Byrd Amendment was also important. In the Bananas
case, the United States retaliatory measures served to increase the profile of the dispute,
but other factors such as the desire by new administrations in the United States and in
the European Union to improve their bilateral relations seemed at least as important in
its resolution. Joe Kirwn, Daniel Pruzin & Gary G. Yerkey, U.S. and EU End Decade-
Long Dispute over EU's Banana Import Regime, 18 INT'L TRADE REP. (BNA) 564 (Apr. 12,
2001).
43. See generally Appellate Body 2007 Annual Report, supra note 21, at 42-47
(describing the way that repeat players use this system).
44. Id.
45. The only arguable exception is Mexico in the Byrd Amendment case where the
European Union, Canada, and Japan acted first. O'Boyle, supra note 42, at 1386; see
How Have Developing Countries Fared?,supra note 30, at 37-38.
46. 1 have argued elsewhere that the typical reasonable periods of time for imple-
mentation set by the WTO system are too long and should be systematically shortened.
See Expediting the Panel Process, supra note 23, at 426-28.
182
clude non-compliance from becoming an acceptable status quo position. I

will discuss each in turn.
First, I think it is evident that another remedy more meaningful to the
typical WTO member is needed, because only a small group of powerful
countries can be expected to effectively use retaliation. The obvious possi-
ble remedy is to allow a prevailing party to choose between suspension of
concessions and receipt of a periodic monetary payment. Two particular
problems would arise from implementing such a change. The first prob-
lem is that WTO members obviously vary in their abilities to pay fines.
The system would have to be designed to avoid the possibility that rich
members could effectively buy their way out of obligations in a way not
available to the poor members. That result might be accomplished by tying
the amount of fines to the size of the member's economy or otherwise pro-
viding for a sliding scale that would minimize "discrimination" against
poor members. The result could also be accomplished by giving the choice
between suspension of concessions and receipt of a periodic monetary
payment only to developing countries and small developed countries; how-
ever this "discrimination" against the larger, richer countries might make
the reform considerably less viable. Additionally, because this change
would resemble forced compensation, an enforceability problem arises
here as well. However, if the amount has been set by a DSU Article 26
arbitration, one could provide that the award would be enforceable under
the New York Convention on the Recognition and Enforcement of Foreign
Arbitral Awards, thus largely eliminating the enforceability issue. 4 7 Even if
this alternative is sometimes problematic, the right to receive a payment
will presumably work out in some circumstances and still be more valuable
than the never-used and probably unusable right to suspend concessions.
Second, the prospective nature of WTO remedies currently gives coun-
tries no incentive to comply promptly and may even encourage foot drag-
ging. To minimize this problem and to create incentives for prompt
compliance, any remedy (whether retaliation or money payment) should
be calculated from a date prior to the date set for implementation (e.g.,
date of adoption of the relevant report or date of panel establishment or
even earlier). Because the current rule would continue so that no remedy
would be imposed if implementation occurs within the reasonable period
of time, there would be an incentive to meet that deadline for
implementation.
Third, increasing sanctions over time would also seem to offer some
real possibilities for improving implementation. Such a procedure would
help to avoid the perception that the payment of fines or damages is simply
an alternative to compliance. In a sense, the European Community used
this concept in the FSC case when it imposed a duty on a long list of U.S.
47. See Convention on the Recognition and Enforcement of Foreign Arbitral Awards,
June 10, 1958, 21 U.S.T. 2517, 330 U.N.T.S. 38, available at http://www.uncitral.org/
pdf/english/texts/arbitration/NY-conv/XXII_l e.pdf. The vast majority of WTO mem-
bers are party to the convention. As a backup, non-parties could be required to create
funds from which fines or damages could be paid without specific legislative approval.
183
products that started at 5% and was increased by 1% each month. 48 The

monthly change focused attention on the case each month, and the
impending increase, even if small, created an incentive to act to forestall
it. 49 In U.S. congressional debates on the FSC implementation legislation,
some members of Congress made this point.5 0 The huge size of the FSC
sanctions-$4 billion-made a phase-in of the sanctions practical in any
event.5 1 However, the same concept could be used in other cases to
encourage more prompt implementation.
If the changes suggested above were made, how effective would they be
in addressing the problems of timeliness and quality of implementation?
The use of retrospective remedies and increased sanctions over time would
likely help solve the particular implementation problems in the subsidy,
agriculture, and SPS areas.5 2 In each case, the cost of non-compliance
would increase and that increase would help offset the political opposition
to implementation. However, that would not necessarily be the case for
safeguards and trade remedies. As outlined above, the assumption of the
current system (and the proposed changes thereto) is that compliance
within the set reasonable period of time would absolve a country from suf-
fering the application of any remedy. However, in the area of safeguards
and trade remedies such a rule would still allow a country to impose the
safeguard or remedy and enjoy its effect; the country would only be com-
pelled to remove the safeguard or remedy after the WTO dispute settlement
procedure had run its course. Thus, questionable safeguards and trade
remedies might continue to be imposed and then removed after a couple of
years of disrupting trade. While removal would be more timely with these
changes, significant trade disruption would continue to occur.
It seems to me that the only way to solve this problem would be to
require payment of reparations if the imposition of such measures is found
to be WTO-inconsistent. I think that other possible reforms are unlikely to
work. For example, WTO disputes are often too complex to expect that a
system of provisional remedies could make much difference. Because these
measures often stop trade-resulting in no collection of duties-a rule on
refund of duties would also not address the problem in general. Providing
for reparations for damages to trade flows might work. However, although
reparations are the standard remedy in international law for violations of a
state's obligations, 5 3 importing such a concept into the WTO would proba-
48. Ritterpusch & Yerkey, supra note 42, at 722.

49. See id.
50. See id. at 720; see also Export Tax Repeal, supra note 42, at 1591; Retaliatory EU
Taxes, supra note 42, at 1193.
51. See Ritterpusch & Yerkey, supra note 42, at 721.
52. See Isabelle Van Damme, Fifth Annual WTO Conference: An Overview, 8 J. INT'L
EcON. L. 769, 779-80 (2005).
53. See JAMES CRAwFoRD, THE INTERNATIONAL LAW COMMISSION'S ARTICLES ON STATE
RESPONSIBILITY: INTRODUCTION, TEXT AND COMMENTARIES 201-06 (2002) (noting that Arti-
cle 31(1) of the International Law Commission's Articles on State Responsibility pro-
vides: "The responsible State is under an obligation to make full reparation for the
injury caused by the internationally wrongful act").
184
128 Cornell International Law Journal Vol. 42
bly be viewed by WTO members as a more drastic and less acceptable

change than the adjustments to the current reforms proposed above.
In conclusion, the WTO dispute settlement system has an admirable
record, but one that is in need of reinforcement through thoughtful
improvements, especially with respect to remedies. In particular, as argued
above, the WTO remedies for non-implementation should incorporate: (i)
the possibility of substituting fines or damages as a remedy in lieu of sus-
pension of concessions; (ii) some degree of retroactivity, so as to help
encourage compliance within the reasonable period of time; and (iii) some
adjustment mechanism to increase the level of sanctions over time, to pre-
clude non-compliance from becoming an acceptable status quo position.
These changes would likely improve the WTO dispute settlement system's
implementation record.
185
FORUM
This article appeared as a contribution to a Forum on “The WTO after Cancún”,

in: INTERECONOMICS – Review of European Economic Policy, Volume 38, Number 5,
September/October 2003, pp. 241-245.
Heinz Hauser* and Thomas A. Zimmermann**
The Challenge of Reforming the WTO Dispute

Settlement Understanding
T he May 2003 deadline for the completion of the

negotiations on improvements and clarifications
of the Dispute Settlement Understanding (DSU) under
tomatic” adoption means that the reports are adopted
unless the parties decide by consensus (i.e. including
the party that has prevailed) not to adopt the report. If
the Doha Mandate has not been met. However, Mem- it has been found that a trade measure is in violation
bers agreed in July 2003 to extend the deadline for the of WTO law, the defendant must bring this measure
review until the end of May 2004. This article briefly into compliance with the covered agreements within a
summarises the past six years of negotiations on the reasonable period of time, normally not exceeding 15
DSU review, the most contentious issues and the sys- months. If the defendant refuses to comply, the com-
temic difficulties of the negotiations. We conclude with plainant may ask the defendant to enter into negotia-
prospects for the forthcoming negotiations until 2004. tions on compensation, or it may seek authorisation
The DSU: Rules for Settling Disputes in the WTO from the DSB to suspend concessions or other obliga-
tions vis-à-vis the defendant in an amount equivalent
The DSU contains the rules for settling disputes be- to the injury suffered. If the adequacy of implementa-
tween WTO Members that arise under the agreements tion is disputed, the implementation measures are
covered. In short, it provides for a procedure that subject to further review under the DSU. Retaliation,
starts with mandatory consultations as a negotiatory if authorised, normally takes the form of punitive tariffs
element. If the parties cannot agree to a settlement on a defined volume of the complainant’s imports from
during these consultations within a certain period, the the defendant.
complainant may request a panel to review the matter.
Panels engage in fact-finding and apply the relevant The DSU has often been praised as the “crown
WTO provisions. Their findings and recommenda- jewel” of the Uruguay Round Agreements. Key innova-
tions are published in a report, which either or both tions with regard to dispute settlement under the GATT
of the parties may appeal. The Appellate Body is then are strict time-frames and the lifting of the former
to review the issues of law and legal interpretations consensus requirement which allowed a defendant
in the panel report. It can uphold, modify or reverse to block the adoption of an adverse ruling. Since the
the panel’s findings. Subsequently, the reports are DSU entered into force on 1 January 1995, its provi-
adopted in a quasi-automatic adoption procedure by sions have been applied to the settlement of some 300
the Dispute Settlement Body (DSB) where all WTO disputes on a wide range of topics.
Members are represented by a delegate. “Quasi-au- DSU Review: Fruitless Efforts Since 1998
Originally, a 1994 Ministerial Decision had called
* upon Members to complete a full review of the DSU
Professor of international economics and director of the Swiss In-
stitute for International Economics and Applied Economic Research, by 1 January 1999, and to take a decision whether to
University of St. Gallen, Switzerland (http://www.siaw.unisg.ch).
**
continue, modify, or terminate the DSU at the Seattle
Research Associate, Swiss Institute for International Economics
and Applied Economic Research, University of St. Gallen, Switzer- Ministerial Conference. Despite intense discussions
land. and an extension of the deadline until July 1999, no
Intereconomics, September/October 2003 241
186
FORUM
result was achieved. After the failure of the Seattle as well. The compromise text would have introduced
Ministerial in December 1999, the DSU review fell into numerous amendments in other areas, including, inter
an inconclusive limbo. Efforts to further the review dur- alia, housekeeping proposals, enhanced third party
ing 2000 and 2001 remained unsuccessful. It was only rights, enhanced compensation, strengthened notifi-
at the Doha Ministerial Conference in November 2001 cation requirements for bilateral solutions, and special
that a mandate for further negotiations on improve- and differential treatment of developing countries.
ments and clarifications of the DSU was included in
the Ministerial Declaration. Between spring 2002 and Contentious Issues
May 2003, 42 specific proposals were submitted by The controversial issues that have not been in-
members, covering virtually all provisions of the DSU. tegrated into the text include, for instance, several
The “Chairman’s text” of 28 May 2003, named after elements of a proposal by the United States and Chile
the Chairman of the negotiations Péter Balás, incorpo- on “improving flexibility and member control in WTO
rates many of these proposals and was meant to serve dispute settlement”. Obviously motivated by a series
as a basis for agreement.1 However, portions of the of defeats in trade remedy cases and a surge of criti-
text are still bracketed (which indicates disagreement cism of WTO dispute settlement from US Congress, it
between the parties) and, more important, many of the
would have allowed the deletion of findings in panel
more controversial proposals have been left out. Al-
or Appellate Body reports by mutual agreement of the
though the text was not accepted by the negotiators, it
parties. Furthermore, it would have provided for the
is worth looking at in some detail.
partial adoption of panel and Appellate Body reports,
Searching for Compromise: The Balás Text and it called for “some form of additional guidance to
The Balás text contains a procedure to overcome WTO adjudicative bodies”. A majority of small and
the “sequencing issue” which arose over ambiguities medium-sized trading nations refuses any increase
(or even contradictions, as some may argue) in Art. of political control, as this would automatically benefit
21.5/22 DSU. It surfaced during the EC - Bananas the more powerful Members.
case, where the WTO-consistency of the EU imple- Another proposal that was not taken into account is
mentation measures was disputed. The key question
the EU call for a permanent panel body. Whereas pan-
is whether a “compliance panel” must first review the
elists are usually government officials or other trade
implementation measures undertaken by a defendant
specialists who are appointed ad hoc and discharge
before a complainant may seek authorisation to retali-
their tasks as panelists on a part-time basis and in
ate on grounds of the defendant’s alleged non-com-
addition to their ordinary duties, the EU wants to es-
pliance. Whereas the US initially opposed any idea of
tablish a roster of 15 to 24 full-time panelists. The EU
sequencing and favoured immediate retaliation, the
hopes that this would lead to a professionalisation of
EU and many other Members argued in favour of the
completion of such a compliance panel procedure as the panel process and help overcome problems with
a prerequisite to seeking an authorisation to retaliate. the selection of panelists, as parties find it increas-
Over time, however, this debate lost its acrimony: after ingly difficult to agree on the composition of panels.
the US had been defeated in the US – Foreign Sales Opponents of the proposal argue that members of a
Corporations (FSC) case, it found itself unable to im- permanent panel body could be more “ideological”
plement the rulings in a timely and WTO-consistent and might engage in lawmaking. They therefore feel
manner. It subsequently agreed with the EU on se- more comfortable with the current system as it draws
quencing for that particular case. This practice, which heavily on government officials who are familiar with
was also applied in other disputes, would now have the constraints faced by governments. The EU equally
been introduced into the DSU. failed in introducing a prohibition on “carousel” retali-
The Balás text would also have brought noteworthy ation. Carousel retaliation consists of periodic modi-
modifications at the appellate stage, introducing an fications of the list of products that are subject to the
interim review, and a remand procedure. In this proce- suspension of concessions if a respondent fails to
dure, an issue may be remanded to the original panel comply with an adverse ruling. The US signed such
in case the Appellate Body is not able to fully address provisions into law in 2000 as the EU did not comply
an issue due to a lack of factual information in the pan- with the adverse WTO rulings in the Hormones and
el report. Remand panel reports could be appealed Bananas cases. Although the provisions have never
been applied, the EC continues to oppose them as it
1
Document TN/DS/9, available at the WTO website. already did in the DSU review during 1999.
242 Intereconomics, September/October 2003
187
FORUM
The US, in turn, failed to have its proposals on settlement. While one might expect at first that this is
increased transparency considered. The US wants largely an issue that divides larger and smaller nations,
to make submissions of parties to panels and the things are not as simple as that: many developing
Appellate Body public, and it wants to allow public countries equally argue for strengthening the negoti-
observance of panel and Appellate Body meetings. ating mechanisms, as they are disappointed with the
Particularly developing countries oppose such in- final outcome of litigation. The Ecuadorian experience
creased transparency, as they fear “trials by media” in the Bananas case has shown that retaliation as the
and undue public pressure. Insisting on the intergov- last resort is ineffective for small developing countries.
ernmental nature of the WTO, they also resist efforts by Not only do they lack retaliatory power because of in-
the US and the EC to formalise the acceptance of ami- sufficient market size, but they would also mainly harm
cus curiae or “friend of the court” briefs. Amicus briefs
their own development prospects by shutting out im-
are unsolicited reports which a private person or entity
ports from industrial nations. Moreover, litigation is ex-
submits to the adjudicative bodies in order to assist
pensive. Finally, overall political considerations (such
(and to influence) the Court in its decision-making. The
as GSP preferences, official development assistance
issue had surfaced for the first time in 1998 when the
and many others) may prevent developing countries
Appellate Body decided in US – Shrimp/Turtle that the
from engaging in litigation with developed countries.
panel had the authority to accept unsolicited amicus
curiae briefs. That right was subsequently expanded Systemic Difficulties in Reforming the DSU
in further disputes, causing outrage among many de-
Thirdly, there are also systemic reasons for the low
veloping country Members.
success of the DSU review: the dispute settlement
Developing countries, in turn, did not manage to mechanism has a “constitutional” character, as it con-
introduce collective retaliation into the draft. It was tains the basic rules for the settlement of any dispute
meant to address the problems caused by the lack that may arise under any of the covered WTO Agree-
of retaliatory power of many small developing econo- ments. Factually, it also has a crucial function in inter-
mies, such as those experienced by Ecuador in the EC
preting the provisions. Not surprisingly, the decision
– Bananas case. With collective retaliation, all WTO
to approve amendments to the DSU shall be made by
Members would be authorised (or even obliged under
consensus, as Art. X.8 WTO Agreement provides. As
the idea of collective responsibility) to suspend con-
Rawls taught us, constitutional rules should always be
cessions vis-à-vis a non-complying Member.
agreed by actors in the “original position” and behind
Diverging Views on the Fundamental Orientation a “veil of ignorance” in order to prevent self-serving
of the DSU choices. In the reality of trade policy, however, such
There are at least three major reasons why it has a veil of ignorance does not exist, as Members, after
been so difficult to agree on a compromise text. The nearly 300 disputes, know fairly well their own and the
first one is disagreement on specific issues as outlined other parties’ vulnerabilities. The “context” in which
above, combined with a lack of political will to settle for the DSU review takes place thus creates difficulties on
a compromise. The second one is a more fundamental three levels in particular.
disagreement which blocks successful negotiations:
• Firstly, the review is conducted in the light of the
the dispute settlement system has gradually moved,
substance of the disputes that are brought to the
over the past few decades, from barely codified prac-
WTO on a continuing basis, and in particular of
tices relying heavily on diplomatic negotiations to an
the politically more controversial ones. It is in these
increasingly codified litigation mechanism with strong
emphasis on the rule of law. Currently, however, there disputes or in their respective context where all the
seems to be no consensus on whether that trend to- crucial issues in the debate have arisen (sequenc-
wards judicialisation should continue. Some proposals ing and collective retaliation in Bananas, carousel
would contribute directly or indirectly to a strengthen- in Hormones, and amicus curiae in Shrimp/Turtle)
ing of the rule of law, such as a professional perma- and where country positions have been shaped. In
nent panel body, increased notification requirements addition, country positions are not only influenced
for mutually agreed solutions, improved enforcement by past experience but also by expectations with
or strengthened third party rights. Other submissions, regard to looming disputes. For instance, the EC will
however, such as the US proposal on flexibility and have been aware during the entire review exercise of
increased Member control, aim at reversing this trend a potential challenge to its GMO regime. Indeed, a
and seek to strengthen the political element of dispute panel has now been established.
188
FORUM
• A second element of this context consists of specific Although the EC and the US are not the only par-
procedural disagreements which, at the same time, ticipants in the negotiations, much will depend on the
are the subject both of on-going disputes and of ne- evolution in (and between) Brussels and Washington.
gotiations on the DSU. For instance, the EC tried to Figures available today suggest that the US will con-
settle the sequencing dispute by making it the sub- tinue to be in a defensive position, because (as of 22
ject of specific complaints in US – Certain EC prod- July 2003) it was the complainant in only 10 active
ucts and US – Section 301. Similarly, a consultation cases, as opposed to 21 active cases where it was
request in US – Section 306 on carousel retaliation is the defendant. Of the latter, all but two concern trade
still pending. Each time such a dispute is under con- remedies.2 More adverse rulings will therefore likely
sideration and the outcome is unclear, no party has spur the criticism of Congress and prevent US ne-
an interest in prejudicing its position through a prior gotiators from consenting to any package that would
agreement on the issue during DSU review negotia- not increase political control. Interests could change,
tions. however, if the US prevailed in the new transatlantic
trade dispute on Genetically Modified Organisms
• Finally, the third layer of this context consists of on-
(GMOs). Rulings in the GMO case could finally set the
going negotiations on material WTO law. The extent
stage for a comprehensive settlement of all outstand-
to which new disciplines such as the Singapore
ing transatlantic trade disputes within a package. So
issues (e.g. investment, competition) could be sub-
far, the EU has consistently refused any package deal,
ject to dispute settlement rules has a direct impact
probably because not all issues were on the table. If
on Members’ approach to the DSU. This logic also
accounts could finally be settled, pressure from DSU
holds for the re-negotiation of existing agreements.
negotiations could be lifted and a mini-package might
For instance, there is a clear link between the current
become feasible. However, settling the many cases
Doha Round negotiations on WTO Rules (such as
involving the US would not only have to involve the
anti-dumping and countervailing measures) and the
EC but many other important WTO Members such as
on-going US debate on the “standards of review”
Japan, China, India or Brazil.
applied by the adjudicative bodies. It will clearly be
more difficult for the US to agree on new rules on As this is a tremendous task, it is unlikely that the
trade remedies in a setting where the adjudicative new May 2004 deadline for the DSU negotiations will
bodies act independently than if they operate under be met – particularly now that the failure of the Cancún
close control of Members. Ministerial has caused a severe setback to the Doha
Round negotiations in general.
Add to these considerations a dynamic aspect: this
entire context is not static, but it evolves with each A lack of progress on the DSU review does not,
new (or merely expected) development that threatens however, need to impair the functioning of the DSU.
to modify positions taken on the DSU review, thereby First of all, negotiators have missed several deadlines
making negotiations even more difficult. so far, and the DSU is still functioning relatively well.
Secondly, provisional solutions have been found
What Can We Expect from Negotiations Until the
for most practical problems in DSU practice: For
New May 2004 Deadline?
instance, countries make bilateral agreements on se-
The foregoing analysis may help us to evaluate the quencing that bridge the gaps of Articles 21.5/22. With
context under which the review will take place during regard to the amicus issue, the Appellate Body has
the following months as the new May 2004 deadline developed its own methodology which grants a lot of
comes closer. discretion. It is using this discretion wisely, displaying
As far as potential outcomes are concerned, funda- a general openness towards accepting amicus briefs
mental changes to the system must not be expected. while at the same time not giving them decisive weight
Each of the more far-reaching proposals with implica- in its decisions, at least not explicitly. And as some ob-
tions on the fundamental orientation of the DSU (rule servers privately argue, the increasing judicial restraint
versus power-orientation) will not be acceptable to a which the Appellate Body exercises in trade remedy
substantial number of Members. Therefore, a package cases helps to appease growing US concerns with
of mainly technical modifications seems to be feasible regard to an alleged anti-US bias of the system. These
at best. Such a package, however, may not enjoy suf- are just three examples which show that the system
ficient support from the large players (notably the EC
and the US) as it will do little to satisfy their ambitions 2
USTR: Snapshot of WTO Cases Involving the United States, updated
and improve their situation. 22 July 2003; http://www.ustr.gov/enforcement/snapshot.pdf.
244 Intereconomics, September/October 2003
189
FORUM
has displayed enough flexibility to deal with new is- decision-making mechanism in the WTO is very weak.
sues as they arise. This imbalance between the inefficient political deci-
sion-making mechanism and the efficient adjudication
Whereas the timely completion of the DSU review is
mechanism causes problems for the long-term sus-
therefore less urgent, the improvement of the political tainability of the WTO. Whereas the US proposal seeks
decision-making mechanism will be crucial. Adjudica- to remedy the imbalance by introducing more political
tive bodies are currently forced to issue rulings even elements into the dispute settlement procedure, legal
on provisions that have been left deliberately vague as scholars strongly advocate an improvement of po-
negotiators were unable to agree on clear treaty text. litical decision-making. Changing the traditionally con-
Interpreting these vague provisions in a legal adjudi- sensus-based decision-making in the WTO, however,
cation procedure inevitably creates political tensions. is yet a much more formidable task than the relatively
These, however, are difficult to correct as the political limited DSU review.
190
WTO Dispute Settlement: Three Years in Review
The panel was convened at 2:30 p.m., Thursday, April 2, by its Chair, Robert Hudec,*
who introduced the panelists: Andrew Shoyer, Powell, Goldstein, Frazer & Murphy LLP;
Debra Steger, World Trade Organization Appellate Body Secretariat; and Thomas Cottier,
Institute of European and International Economic Law.
The Future of WTO Dispute Settlement

by Andrew W. Shoyer"
The World Trade Organization (WTO) dispute settlement rules and procedures are
working and should be allowed to continue in force. There is no call for wholesale revisions
to the Dispute Settlement Understanding (DSU) at this time. There are, however, several
areas that warrant further attention over the next few years. First, members should consider
modifying current practice so as to protect solutions reached in settlement. Second, members
are likely to develop legislative responses in the future to controversial panel and Appellate
Body reports. Third, members must ensure greater transparency in the dispute settlement
process, and should start by making their own submissions public.
Protecting Negotiated Settlements

Mutually agreed solutions to matters raised in dispute settlement are "clearly to be
preferred" to recourse to panel procedures or other devices included in the DSU.1 More than
20 percent of the requests for consultation filed by WTO members during the first three
years of the WTO's existence led to mutually agreed solutions. Several had bumpy rides
afterward?parties to the settlement disagreed vehemently on whether the negotiated
solution was being implemented.2 One would think that the Understanding would facilitate
the implementation of settlement agreements at least as well as it does the implementation
of the rulings and recommendations of the Dispute Settlement Body (DSB). Yet the
Understanding is silent on this point.
The parties to those settlements could have helped themselves by including a clause in
the settlement agreements providing for arbitration of disputes arising out of the agreement.
Indeed, the parties could have provided for arbitration pursuant to Article 25 of the DSU.
That provision appears to be sufficiently broadly drafted to encompass arbitration over
compliance with a settlement agreement in a dispute brought under the WTO dispute
settlement provisions. Most important, paragraph 4 of Article 25 provides that arbitral
awards are subject to surveillance under Article 21 and the provisions of Article 22 on
compensation and suspension of substantially equivalent concessions.
"Professor of Law, University of Minnesota Law School.

"Partner, Powell, Goldstein, Frazer & Murphy LLP, Washington, DC. From August 1993 through
October 1997, Mr. Shoyer served as Legal Adviser in the U.S. Mission to the World Trade
Organization. The views expressed herein do not necessary reflect those of either the Office of the U.S.
Trade Representative or the firm of Powell, Goldstein, Frazer & Murphy LLP.
'DSU Article 3.7 provides, in part, that "[a] solution mutually acceptable to the parties to a
dispute and consistent with the covered agreements is clearly to be preferred [over the use of panel
procedures]."
2See, e.g., Korea?Measures Concerning the Shelf-Life of Products?Notification of Mutually
Agreed Solution?Revision, WT/DS5/5/Add.l (Apr. 22,1996); Annual Report at 11 (1997) (regarding
the U.S. complaint concerning EC tariff concessions on grains).
191
Parties should routinely include in settlement agreements a clause providing for arbi
tration pursuant to Article 25 of the DSU. In the absence of guidance in the text of the DSU,
however, parties in a relatively weak bargaining position might be unable to convince their
negotiating partners to include such a provision. To remedy that situation, members might
consider modifying either Article 3 or Article 25 (or both) to instruct parties to a settlement
to include such an arbitration clause in all such agreements to be notified to the DSB.
An even more daring solution would be to modify the DSU to provide in Article 1 that
"covered agreements" include mutually agreed solutions notified to the DSB and the relevant
councils and committees, pursuant to paragraph 6 of Article 3 of the Understanding.
Effectively, this step would transform commitments made in settlements into WTO
commitments. Such a reform would be consistent with the spirit of DSU provisions requiring
that all solutions to matters formally raised in dispute settlement be consistent with WTO
rules.3 This would eliminate any doubt that Article 25 arbitrators could hear conflicts about
compliance with a settlement agreement. It would also create a right to seek the estab
lishment of a panel to review an alleged breach of the settlement agreement. In addition, it
would create a powerful incentive for members to notify their settlements pursuant to
paragraph 6 of Article 3 of the DSU, as they have done in only some cases so far.
The danger might be, however, that in formalizing settlements, members will find
parties to a dispute less willing to settle, or less likely to use WTO dispute settlement
procedures. With several more years of experience with the DSU, members will have a
better foundation on which to assess whether this risk is a significant one.
Responses to a "High Court"

The introduction of appellate review into the dispute settlement process is one of the
most interesting innovations to emerge from the Uruguay Round. After only three years, the
Appellate Body appears to have acquired significant influence over the way that panels
perceive their role and how proceedings should be conducted. For example, the introduction
of appellate review brought much greater attention than ever before to the need for a written
record of the panel proceeding. Pursuant to Rule 25 of the Working Procedures for
Appellate Review,4 the Appellate Body stipulated that, upon the filing of a Notice ofAppeal,
the WTO Secretariat would transmit the "complete record of the panel proceeding" to the
Appellate Body. Rule 25(2) of the Working Procedures defines what the "record" includes.
As a result, panels and the Secretariat have been forced to be much more rigorous about the
written record created during the proceedings, and this has affected how proceedings are
conducted.
The development of a healthy relationship between panels and the Appellate
Body?between court of first instance and reviewing court?will be critical to the
performance of the system. Although the DSU does not require panels to adhere to decisions
of the Appellate Body, panelists (and those advocates appearing before them) treat Appellate
Body reports as high court decisions. They cite them with great deference, as is fitting. In
turn, the Appellate Body is helping panels understand how to work within the guidelines the
Appellate Body sets. In its recent decision in the dispute on the EC's ban on beef from cattle
grown with hormones, the Appellate Body corrected a panel on its application of previous
Appellate Body decisions regarding the allocation of burden of proof.5 The relationship
between upper and lower courts necessarily involves a degree of tension. But if both "sides"
3DSU Article 3.5.

4WT/ABAVP/3, circulated Feb. 28,1997.
5See Appellate Body Report on EC Measures Concerning Meat and Meat Products (Hormones),
circulated Jan. 16,1998, (WT/DS26/AB/R & WT/DS48/AB/R) at para. 99.
192
(assisted by their respective secretariats) strive to maintain a constructive dialogue, then a

better jurisprudence will emerge.
Although the Appellate Body can correct the errors of panels, the Dispute Settlement
Body is not in an effective position under the DSU to correct the errors of the Appellate
Body. In light of the "negative consensus" rule of Article 17.14 of the DSU, and the
requirement that Appellate Body reports shall be adopted "unconditionally," there is little
chance for the Dispute Settlement Body to fix problems in legal interpretation that arise in
Appellate Body reports. Although members can comment on problems, the Dispute
Settlement Body is unlikely to develop consensus on clarifications or corrections in reaction
to Appellate Body reports.
The role of consensus maker is more likely to fall on the Ministerial Conference or the
General Council. Indeed, perhaps the most interesting dynamic that might emerge from the
post-Uruguay Round dispute settlement mechanism will be the "legislative" response by
WTO members to decisions taken by the WTO "judiciary." In the past, General Agreement
on Tariffs and Trade (GATT) "law" was developed principally through the work of panels.
Even after the Tokyo Round, there was relatively more jurisprudence than "code law" (in
the form of treaty text). It could hardly be said, however, that successive trade rounds under
the auspices of the GATT 1947 constituted reactions by the contracting parties to errant
panel decisions.
Much has changed in this regard after the Uruguay Round. First, the delegations that
negotiated what became the WTO Agreement created a great deal of "code law." Second,
they created in the Appellate Body a judicial body that functions like a high court, subject
to checks and balances. Third, the Appellate Body appears to base its interpretations of the
WTO Agreement on the Vienna Convention on the Law of Treaties rather than on
negotiating history or its own understanding of negotiating intent.
The Appellate Body is under enormous time pressure to produce decisions in increasing
numbers of cases. Sometimes, the Appellate Body will interpret the WTO "code law" with
unexpected or unintended results. Perhaps members will use their authority under Article
11.2 of the WTO Agreement to adopt interpretations to correct the perceived errors of the
Appellate Body. Instead, or in addition, the decisions of the Appellate Body could elicit a
legislative response from members where previously no such legislative mechanism existed.
There might come a day when the agendas of trade rounds are influenced just as heavily by
Appellate Body decisions as they are now by external commercial and economic issues faced
by the trading community. Such a relationship between members and the Appellate Body
would be very useful because it would improve the quality of work of those who draft future
WTO treaty text and those who interpret it.
Ensuring Timely Access to Submissions

One of the debates during the final stage of negotiation of the DSU addressed the
implications of the culture of confidentiality inherited from the GATT 1947 procedures.
Some delegations argued that the results of the WTO dispute settlement process would never
be credible unless the process were opened up to participation by those nongovernmental
entities affected by those results. Accordingly, they suggested that nongovernmental entities
be permitted to attend and participate in panel and Appellate Body proceedings, submit
written briefs, and have complete access to all documents submitted to panels and the
Appellate Body. In contrast, other delegations argued that protecting the confidential,
government-to-government nature of dispute settlement was essential to the resolution of
disputes. These delegations argued that nongovernmental entities had ample opportunity to
participate in and influence the dispute settlement process through domestic advisory
procedures conducted in members' capitals.
193
The common ground that developed during the debate was not extensive, but it found
its eventual home in paragraph 2 of Article 18 of the DSU. Although no consensus emerged
on opening dispute settlement proceedings to nonmembers, delegations agreed that the
confidentiality rules should not preclude members from disclosing to the public statements
of their own positions taken in the course of dispute settlement. This notion was codified in
paragraph 2 of Article 18 of the Understanding. Article 18.2 also provides that, upon the
request of any member, a party to a dispute must provide a nonconfidential summary of the
information contained in its written submissions that could be disclosed to the public.
In practice, implementation of Article 18.2 has been disappointing. Although the
obligation of a member to provide a nonconfidential summary of its submissions is absolute,
there is no deadline by which the party must supply the summary to the requesting member.
The U.S. Government, as a matter of domestic law, must request that each party to a WTO
dispute settlement proceeding?regardless of whether the United States is itself a party to
the dispute?provide nonconfidential summaries of its written submissions.6 Accordingly,
the United States requests as a matter of course that each party to a dispute provide a
nonconfidential summary if it has not already made its submissions public. In a handful of
cases during the first three years, parties to disputes did make their submissions available to
the public, but in the vast majority of cases they did not. Parties that received requests to
provide nonconfidential summaries responded slowly, if at all.
The sensitivity of some members to the disclosure of positions taken before panels or
the Appellate Body is difficult to appreciate in light of the highly detailed descriptions of
arguments that appear in panel reports. In the descriptive portions of their reports, panels
generally defer to the manner in which parties have articulated their arguments, both in
writing and orally during panel meetings. Indeed, much of the bulk of recent panel reports
is attributable to die near-verbatim transcription of parties' arguments and counterarguments.
(For example, the parties' arguments occupied 321 pages of the roughly 500-page panel
report circulated in March 1998 in the dispute on Japan's measures affecting consumer
photographic film and paper!) Thus, the question of whether parties' arguments should be
disclosed never arises?it is merely a matter of time before these arguments are transcribed
in panel reports and thereby become accessible to the public.
The lack of timely access to the parties' arguments is of critical importance, however.
Judging from the increasing worldwide media attention to WTO dispute settlement, one may
fairly assume that there are increasing numbers of nongovernmental entities that wish to
follow closely and, if possible, influence the course of these proceedings. In the course of
their review of the DSU, members presumably will reengage in debate on private parties'
direct access to panels and the Appellate Body. In the meantime, however, the limited
success in implementing Article 18.2 of the Understanding allows members standing firmly
in the antitransparency camp to have it both ways?they insist that transparency begins (and
ends) at home, rather than in Geneva, but then effectively deny interested parties the material
with which to participate meaningfully in domestic procedures conducted by member
governments in capitals.
I suggest that the current situation is untenable. Without prejudice to members' views
on direct participation of nongovernmental entities in WTO dispute settlement proceedings,
members should adopt a practice now in the Dispute Settlement Body to ensure that all
submissions to panels and the Appellate Body are drafted as public documents and made
available to the public at the time they are submitted. Members could eventually modify the
text of Article 18.2 of the Understanding to require all parties to a dispute to make their
6Uruguay Round Agreements Act ? 127(d), 19 U.S.C. ? 3537(d) (1994).
194
submissions available to interested parties on a "real-time" basis. Business confidential

information has generally not played a role in WTO proceedings, but parties should clearly
retain the right to protect such information from disclosure.
wto Dispute Settlement: Emerging Practice and Procedure

by Debra P. Steger and Peter Van den Bossche*
Introduction
In the more than fifty years of General Agreement on Tariffs and Trade/World Trade
Organization (GATT/WTO) history, the dispute settlement system has been remarkable for
its pragmatism and effectiveness. The dispute settlement system of the GATT 1947 evolved
over time from the very sparse foundation provided by Articles XXII and XXIII. Although
procedural rules were codified in certain Decisions and Understandings agreed to by the
contracting parties in 1966, 1979 and 1989,* the Understanding on Rules and Procedures
Governing the Settlement of Disputes (DSU), which emerged from the Uruguay Round
negotiations, represents the first extensive, negotiated agreement reforming and revitalizing
the GATT dispute settlement system. It has been argued that some of the detailed
innovations in the DSU will drive the system ever more into a judicialized model of dispute
settlement. Foremost among these innovations was the establishment of the Appellate Body
in late 1995.
Since that time, the Appellate Body has heard and decided ten appeals. The Appellate
Body consists of seven members, selected from different legal systems. The members serve
for fixed terms, in most cases of four years, which are renewable for another four years. The
Appellate Body sits in divisions of three members, selected on a rotational basis, to hear and
decide individual appeals. The Appellate Body has developed a practice of collegiality
whereby all seven Appellate Body members come to Geneva in the deliberation phase of
each appeal to exchange views on the matters at issue in the appeal. This practice has
promoted consistency in the early decisions of the Appellate Body, which has been
especially important for issues of a systemic or general nature.
It is perhaps somewhat anomalous that the WTO dispute settlement system is now
characterized by, on the one hand, an Appellate Body, which is judicial?indeed, it is much
like an appellate court?and, on the other, panels that function more on the arbitral model.
Appellate Body proceedings are governed by the detailed Working Proceduresfor Appellate
Review,2 which were drawn up by the Appellate Body in early 1996, and which resemble,
in their complexity and specificity, rules of procedure of some courts. An Appellate Body
hearing is a formal judicial hearing, much of which is taken up with penetrating questioning
of the participants and the third participants by the Appellate Body members. In contrast,
panels do not have detailed, standard rules of procedure; rather, there are some very general
provisions set out in Appendix 3 of the DSU. However, panels are free to develop their own
working procedures in individual cases, after consulting with the parties. Panels have
'Debra Steger is Director, and Peter Van den Bossche is Counselor, with the Appellate Body
Secretariat, World Trade Organization, Geneva. The views expressed in this article are the views of
the authors and do not represent the views of the Appellate Body or the World Trade Organization.
1 Procedures under Article XXIII (decision of 5 April 1966) GATT Basic Instruments and Selected
Documents (BISD) 14S/18; Understanding Regarding Notification, Consultation, Dispute Settlement
and Surveillance (adopted on 28 November 1979) BISD 26S/210; Improvements to the GATT Dispute
Settlement Rules and Procedures (decision of 12 April 1989) BISD 36S/61.
2 WT/AB/WP/3,28 Feb. 1997. This is a consolidated, revised version ofthe Working Procedures
for Appellate Review, and replaces WT/AB/WP/1, dated 15 Feb. 1996.
195

Reading Material

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Reading Material

Transféré par

Droits d'auteur :

Formats disponibles

NATIONAL LAW UNIVERSITY ODISHA,

Law on Dispute Resolution

B.A. /B.B.A. LL.B

1. The WTO and Rules-Based Dispute Settlement: Historical Evolution, Operational

The WTO and Rules-Based Dispute

the WTO as a functioning international organization essentially depends on ensuring

13 See generally Croome, above n 2.

of this.27 There is peer pressure to implement WTO Dispute Settlement Body

37 See generally Ibid.

B. Dispute resolution: results of consultations and DSU recommendations

C. The performance of the Appellate Body

D. Cautionary notes: lack of timeliness, limited results, and gap-filling

A. The USA, the European Union, and China: do

74 See text accompanying n 49 above.

B. Useful reforms to improve operation of

2. Encouraging prompt implementation

3. Strengthening WTO remedies for nonimplementation

PRABHASH RANJAN The purpose of the dsu is given in Article

in diction to settle trade disputes

than the most

panels have to examine

A combined international law or non-WTO law the ba-

Ééfs EndGdvour ^.scs^mS

Articles/papers for the Journal 'Endea

■ Leadership ■ Personality Develo

Rajiv Gandhi National Institute of

cannot Report of the Appellate

Economic & Political weekly B3Q april n, 2009 vol xliv no 15 27

submissions available to interested parties on a "real-time" basis. Business confidential

wto Dispute Settlement: Emerging Practice and Procedure

representation by counsel of a government's own choice may well be a matter of

3Adopted 25 Sept. 1997, WT/DS27/AB/R, paras. 132 and 133.

7 Id. para. 142.

The Appellate Body has also ruled that

22EC Hormones, supra note 18, at para. 116.

Sources of Law and Principles of Treaty Interpretation

^Argentina?Measures Affecting Imports of Footwear, Textiles, Apparel and Other Items

"US?Gasoline, supra note 38, at 17.

The WTO Dispute Settlement System: New Horizons

'Professor of Law, University of Bern, Switzerland (thomas.cottier@iew.unibe.ch). I am indebted

2. TIME TAKEN TO ISSUE PANEL REPORTS

Figure 1. Time Taken to Issue Panel Reports

Year of panel composition

3. DSU TIME FRAMES FOR PANEL REPORTS

4. TIME FOR PARTIES’ SUBMISSIONS

be accommodated, with difficulty, in fixing an original timetable and in dealing with

Zeroing.106 Mr Harsha Singh accepted appointment as a WTO Deputy Director-General

7. SCOPE AND COMPLEXITY OF DISPUTES

responses on the panel’s timetable. A complementary strategy is to ensure effective docu-

10. SCIENTIFIC AND TECHNICAL EXPERTS

12. CIRCULATION OF THE FINAL REPORT

THE QUESTION OF REMAND AUTHORITY

In this chapter, the author addresses the possibility of

II PROBLEMS ARISING FROM THE CURRENT SYSTEM

A Powers of the Appellate Body

B Panels’ Use of Judicial Economy

Nothing … in previous GATT practice requires a panel

Nonetheless, it also imposes certain shortcomings when the Appellate

C Circumstances in which Remand Authority Could be Used

(b) the revival of subsidiary claims or defences before the panel.

D Existing Alternatives for the Appellate Body in the Absence of

(c) leave the issues undecided (although this alternative is negated

1 Completing the Analysis

(a) The Practice and its Features