Dispute Settlement in European Union Free Trade Agreements Lessons

Document Sample
Dispute Settlement in European Union Free Trade Agreements Lessons Powered By Docstoc
     Dispute Settlement in European
     Union Free Trade Agreements:
           Lessons Learned?
                      ignacio garcia bercero *

                                    i. introduction

Within the framework of the WTO, the European Union (EU) has become a
major proponent and user of a ‘quasi-judicial’ approach to the solution of
trade disputes. Until recently, however, EU bilateral agreements relied, almost
exclusively, on traditional diplomatic means for dispute resolution. A funda-
mental shift occurred with the conclusion of the free trade agreements with
Mexico (2000) and Chile (2001), both of which incorporate a quasi-judicial
model of trade adjudication, which is largely inspired by the WTO Dispute
Settlement Understanding. Since then, the EU has sought to introduce similar
dispute settlement procedures to all ongoing negotiations for free trade
   This chapter explores the reasons for the increasing EU preference for a
quasi-judicial model of adjudication as far as trade disputes is concerned.
Section II makes a brief presentation of the current network of EU free trade
agreements. Section III describes the traditional ‘diplomatic’ dispute settlement
mechanism, which until recently characterized all EU free trade agreements.
Section IV discusses the main characteristics of the quasi-judicial model of
trade adjudication as reflected in the free trade agreements with Mexico and
Chile. Section V examines how these two agreements have tried to address

  * The author wishes to thank Edna Ramirez Robles for her contribution to this chapter, in
particular for drafting Section I and preparing Table 1. In preparing this chapter, the author has
benefited from discussions with Ms. Ramirez Robles, who is currently preparing a doctoral thesis
on dispute settlement in the EU–Mexico FTA. I also wish to thank Lothar Ehring and Lorand
Bartels for their detailed comments. The opinions expressed in this chapter are personal and do not
necessarily represent the views of the European Commission.
384                               Ignacio Garcia Bercero
the vexed question of the relationship between bilateral and WTO dispute
settlement. Section VI draws some conclusions.

         ii. the current network of european union free
                        trade agreements

All the EU agreements which are discussed in this chapter have one feature in
common: they are all free trade agreements (FTAs) within the meaning of
Article XXIV GATT.1 The only exception is the customs union agreement with
Turkey. It should be noted, however, that recent agreements go well beyond
the requirements of Article XXIV GATT, since they often include trade in
services and/or envisage common disciplines in a number of regulatory areas
(standards, procurement, competition policy, investment, etc.)2
   Although different criteria can be used to classify EU FTAs, this chapter uses
a classification based on the degree of ‘proximity’ between the EU and its FTA
partners. On this basis, three types of FTAs can be considered:
(a) FTAs with countries that are part of the ‘European space’ and which are
    normally recognized as potential EU members. In such cases a classic FTA
    is often complemented with provisions by which the partner country
    undertakes to align its legislation with EU norms in a broad range of
    economic areas.
(b) FTAs with neighbouring countries. These often include a component of
    alignment with EU norms, although to a lesser extent than for countries
    that are candidates for EU membership. The EU has established vis-à-vis
    those countries a ‘neighbourhood policy’, under which neighbouring
    countries are expected to participate in the internal market. This often
    implies ‘partial alignment’ with EU norms.
(c) FTAs with non-neighbouring countries. These agreements include no ref-
    erence to alignment with EU norms. However—and paradoxically—they
    are sometimes more comprehensive in their scope and subject to stronger
    enforcement mechanisms.
The agreements within the ‘European space’ are the European Economic
Area (EEA), the Customs Union with Turkey, the Stabilization and Associ-
ation Agreements (SAAs) with the Western Balkan countries, and the

    A free trade area is a group of two or more countries or economies that have eliminated tariffs
and all or most non-tariff measures affecting trade among themselves: W. Goode, Dictionary of
Trade Policy Terms (Cambridge: CUP, 2003), 146.
    Many contemporary FTAs include goods and services, and cover such issues as investment,
government procurement, and competition. FTAs also contain provisions like mutual recognition
of technical standards, intellectual property, etc.: S. Woolcock, ‘A Framework for Assessing
Regional Trade Agreements: WTO-plus’, in G.P. Sampson and S. Woolcock (eds), Regionalism,
Multilateralism and Economic Integration, the Recent Experience (Hong Kong: UN University,
2003), 18–31.
                Dispute Settlement in EU Free Trade Agreements                       385
Europe Association Agreements with Romania and Bulgaria (until their
accession to the EU). The FTAs concluded with neighbouring countries
are the ‘Euro-Mediterranean’ Association Agreements with Egypt, Lebanon,
Morocco, Algeria, Tunisia, Israel, the Palestine Authority, and Jordan. Neigh-
bouring countries in Europe—such as Ukraine and Russia—have so far only
concluded non-preferential Partnership and Cooperation Agreements (PCAs)
but not FTAs. The countries in the last category are South Africa, Mexico,
and Chile. Negotiations are ongoing with Mercosur, ACP countries, and
the Gulf Cooperation Council (GCC), and are currently being considered
with ASEAN, the Andean Community, and the Central American Common

A. Free trade agreements within the European space

The European Economic Area (EEA) is an agreement that has undergone
many changes as regards its Contracting Parties. The EEA was signed in 1992
by the EC and the European Free Trade Association (EFTA).3 At the time,
the members of EFTA were Austria, Finland, Iceland, Liechtenstein, Norway,
Sweden, and Switzerland (the latter, however, did not ratify the EEA Agree-
ment). Today only Iceland, Liechtenstein, and Norway remain parties to the
   Of the agreements concluded with countries within the European space, the
EEA is sui generis in that it provides an alternative to rather than a preparation
for EU membership. In order to participate in the internal EU market, the EEA
requires almost full alignment with the EU acquis, also with respect to legisla-
tion adopted after the entry into force of the EEA. It is the only agreement
which provides for judicial enforcement procedures, modelled on those of the
European Court of Justice. Although the EEA has sometimes been presented
as a possible model for the EU neighbourhood policy, so far no other EU FTA
has adopted similar institutional provisions.
   The Customs Union with Turkey came into force in 1995 within the frame-
work of the Association Agreement between the European Economic Com-
munity (EEC) and Turkey, which was signed in Ankara in 1963.4 The Ankara
Agreement includes an article that regulates the settlement of disputes relating
to the application or interpretation of the agreement. An innovative feature is
that the article provides that, if the parties so agree, they can use other fora to
settle disputes and notably the Court of Justice of the European Communities.
These ‘alternative’ procedures have, however, never been used.
   The Europe Agreements with Romania and Bulgaria are in force until the

    European Economic Area Agreement, signed 2 May 1992, in force 1 January 1994 [1994]
OJ L1/3.
    See the EEC–Turkey Supplementary Protocol to the Association Agreement, signed 28
November 1977, in force 3 January 1978 [1977] OJ L361/1 and Decision 1/95 of the EC–Turkey
Association Council [1995] OJ L96/142.
386                               Ignacio Garcia Bercero
accession of these two countries to the EU. This will take place on 1 January
2007 unless a Council decision postpones it for a further year.5
   The Stabilization and Association Agreements (SAAs) with the Balkans,
have been signed with Bosnia and Herzegovina, Croatia, and the former
Yugoslav Republic of Macedonia. Negotiations have been launched with
Albania and Serbia and Montenegro.6 The aims of these agreements include
the fostering of respect for democratic principles, as well as the reinforcement
of links between this region and the EU single market, and include a member-
ship perspective. These agreements create a free trade area and provide for the
alignment with the EU acquis in several areas such as in competition, state aid
and intellectual property which will allow the economies of the region to
integrate better with the EU.7
   The Europe Agreements and the SAAs are all based on a ‘diplomatic model’
for the settlement of disputes. As further discussed below, the main features of
such a model are consultations within the framework of a Joint Committee,
including the possibility of having recourse to arbitration. However, the
establishment of an arbitration panel requires consensus by the parties.

B. Association Agreements with neighbouring countries

The Euro-Mediterranean Partnership was launched in 1995 with the Barcelona
Declaration, which establishes a framework of political, economic and social
relations between the EU and the Southern Mediterranean Partners.8 Those
partners currently are Algeria, Egypt, Israel, Jordan, Lebanon, Morocco, the
Palestine Authority, Syria, Tunisia, and Turkey.
   One of the main objectives of the Barcelona Declaration is the establishment
of a Euro-Mediterranean free trade area by 2010. The means of achieving this
free trade area will be through association agreements concluded between the
EU and the Mediterranean Partners as well as FTAs among the Mediterranean
Partners themselves. The FTAs between EU and the Mediterranean Partners
replaced the cooperation agreements signed in the 1970s.9 So far the agreements

     Notice concerning the entry into force of the Treaty of Accession of Romania and Bulgaria
[2005] OJ L157/10.
     Those agreements are part of the of the Stabilization and Association Process (SAP), which is
the framework to support domestic reforms and is based on aid, trade preferences, dialogue,
technical advice, and contractual relations
     See europa.eu.int/comm/external_relations/see/index.htm (visited 10 April 2006).
     On the Barcelona process, see A. Galal and B. Hoekman, Regional Partners in Global Markets:
Limits and Possibilities of the Euro-Med Agreements (London: Centre for Economic Policy
Research, 1997).
     See    europa.eu.int/comm/external_relations/euromed/med_ass_agreements.htm           (visited
10 April 2006).
                  Dispute Settlement in EU Free Trade Agreements                              387
                 10        11              12           13              14
with Tunisia, Israel, Morocco, Jordan, and Egypt have been ratified
and are in force. For Lebanon, the Palestinian Authority, and Algeria the trade
provisions are in force through interim agreements. The Agreement with Syria
has been initialled.
   Euro-Mediterranean Agreements in force include similar dispute settlement
provisions, which correspond essentially to a ‘diplomatic model’. In general,
the dispositions are included only in one article, i.e. Euro-Mediterranean
Agreements with Tunisia (Articles 86 and 34 of the Protocol concerning the
definition of originating products and methods of administrative cooperation),
Israel (Articles 75 and 33 of the Protocol), Morocco (Articles 86 and 34 of the
Protocol), Jordan (Articles 97 and 32 of the Protocol), and Egypt (Articles 82
and 33 of the Protocol).

C. FTAs with non-neighbouring countries

The FTAs with South Africa, Mexico, and Chile are the first signed with
countries outside the European neighbourhood.
   South Africa is a Contracting Party of the EU–African Caribbean Pacific
(ACP) Cotonou Agreement, but is excluded from its trade and financial
co-operation provisions. The Trade, Development and Cooperation Agree-
ment (TDCA) between South Africa and the EU was signed in 1999 and the
ratification process by the Member States is still ongoing.15 However its
trade provisions have been applied under an interim agreement since 2000.
The settlement of disputes under this TDCA goes, in certain aspects, beyond
the provisions of the Euro-Mediterranean Agreements or the agreements
concluded with countries that are candidates to become EU members.
   A provision (Article 104) encompasses a set of rules for arbitration. Con-
sensus is required to establish the arbitration panel and compliance procedures
are not specified. Nevertheless, for trade-related disputes there are certain
innovations, i.e. time limits are established, as well as working procedures for
arbitration, and a reasonable period of time to comply. The arbitrators are
expected to report their findings within 6 months and implementation should
take place within a ‘reasonable period of time’ not exceeding 15 months.

      EC–Tunisia Euro-Mediterranean Association Agreement, signed 17 July 1995, in force
1 March 1998 [1998] OJ L97/2.
      EC–Israel Euro-Mediterranean Association Agreement, signed 20 November 1995, in force
1 June 2000 [2000] OJ L147/3.
      EC–Morocco Euro-Mediterranean Association Agreement, signed 26 February 1996, in
force 1 March 2000 [2000] OJ L70/2.
      EC–Jordan Euro-Mediterranean Association Agreement, signed 24 November 1997, in
force 1 May 2002 [2002] OJ L129/3
      Agreement in the form of an exchange of letters concerning the provisional application of the
trade and trade-related provisions of the EC–Egypt Euro-Mediterranean Association Agreement,
signed 25 June 2001, in force 1 January 2004 [2003] OJ L345/115.
      Council Decision No 1999/753 (Provisional application of the EC–South Africa Agreement)
[1999] OJ L311/1.
388                              Ignacio Garcia Bercero
Issues relating to each Party’s WTO rights and obligations can be referred to
the FTA arbitration proceeding only if the parties agree.
   The EU–Mexico contractual relationship was established in two agreements
(the Global and the Interim) along with a Final Act signed in 1997.16 The
EU–Mexico FTA was established through two Decisions of the Joint Council,
which entered into force in different years.17 Decision 2/2000 established a
free trade area in goods, and entered in to force on 1 July 2000,18 while
Decision 2/2001 established a free trade area in services and entered into force
on 1 March 2001.19
   The Global Agreement provides that the Joint Council shall decide on the
establishment of a compatible dispute settlement procedure with the WTO.20
The set of rules on dispute settlement include detailed norms and specific time
frames. They encompass the different stages of the procedure (consultation plus
arbitration), the appointment of arbitrators, the content of the panel reports
(interim and final), and the procedures to implement the final report. They also
include rules of procedure21 and a code of conduct22 for the arbitrators.
   Negotiations for the EC–Chile Association Agreement began in 2000, and
its trade provisions entered into force on an interim basis in February 2003
under the EC–Chile FTA.23
   The settlement of disputes is regulated in depth in the consultations, arbitral,
and compliance stage. It contains rules about the appointment of arbitrators,
the arbitration panel ruling, as well as model rules of procedure24 and a code
of conduct.25
   FTAs are currently being negotiated with Mercosur (Argentina, Brazil,
Paraguay, and Uruguay) and the Gulf Cooperation Council (Saudi Arabia,
Kuwait, Bahrain, Qatar, United Arab Emirates, and Oman). The FTAs being
negotiated with Mercosur and the GCC have some common features. In both

      The first move towards the signature of these Agreements was in 1975 with a Cooperation
Agreement between Mexico and the EEC which was substituted in 1991 with another Cooper-
ation Agreement but of the third generation type. Later in 1995, in a Joint Declaration, both
parties stated their interest in deepening the relationship. See E. Ramírez Robles, Solución de
Controversias en los Acuerdos Celebrados entre México y la Comunidad Europea (Guadalajara:
Universidad de Guadalajara, 2003), 85–104.
      On the EU–Mexico FTA, see J. Reiter, ‘The EU–Mexico Free Trade Agreement: Assessing the
EU approach to Regulatory Issues in Regionalism, Multilateralism and Economic Integration, the
Recent Experience’, in Sampson and Woolcock, above at n 2, at 66.
      Decision No 2/2000 of the EC-Mexico Joint Council of 23 March 2000 [2000] OJ L157/10
and annexes [2000] OJ L245/1. For the chronology see www.economia-bruselas.gob.mx/
ls23al.php?s=501&p=4&l=2 (visited 10 April 2006).
      EU–Mexico Joint Council Decision No 2/2001 [2001] OJ L70/7.
      Art. 50 of the Global Agreement.
      Annex III with reference of Art. 43 of Decision No 2/2001.
      Appendix I of Decision No 2/2001.
      EC–Chile Association Agreement, in force 1 Mar 2005 [2002] OJ L352/3.
      Model Rules of Procedure for the Conduct of Arbitration Panels. Annex XV referred to in
Art. 189(2).
      Code of Conduct for Members of Arbitration Panels, Annex XVI.
                 Dispute Settlement in EU Free Trade Agreements                         389
cases, the counterpart of the EU is a customs union concluded amongst devel-
oping countries. The objective is to negotiate ‘comprehensive’ FTAs which—as
is already the case in the EU–Chile FTA—cover substantially all trade in goods
and services, as well as common disciplines on a number of regulatory areas,
such as investment, public procurement, intellectual property, competition
policy, etc. Negotiations are well advanced on a dispute settlement mechanism
which is, in many respects, similar to those included in the EU–Mexico and
EU–Chile agreements.
   Negotiations are also ongoing on regional Economic Partnership Agree-
ments (EPAs). These agreements are being negotiated between the EU and
regional groups established among African Caribbean Pacific (ACP) countries
(6 regions with 79 countries).26 The current ACP–EC Agreement was con-
cluded for a twenty-year period in Cotonou. The mechanism for the settle-
ment of disputes in the Cotonou Agreement is innovative, i.e., the option for a
multi-party dispute is incorporated, arbitration is possible after consultations;
where the second and/or third arbitrator is not designated, either party can ask
the Secretary General of the Permanent Court of Arbitration to appoint one.27
Negotiations of EPAs with the six regional groups have not yet started to
consider future dispute settlement mechanisms.

       iii. the traditional ‘diplomatic’ dispute settlement
                        mechanism in eu ftas

With the exception of the EEA, all pre-2000 EU FTAs are based on a tradi-
tional diplomatic approach to dispute settlement. Such an approach relies
essentially on consultations and diplomatic negotiations as the key mechanism
to solve disputes relating to the interpretation or implementation of the
agreement. The option of having recourse to legal adjudication (arbitration)
is normally available but requires agreement of the parties. The limited
effectiveness of legal adjudication is balanced with the preservation of uni-
lateral measures of self-help, through the inclusion of a non-execution clause,
although these often have a high diplomatic cost and are only used as measures
of last resort.
   The Europe Agreements which were concluded with countries in Central
and Eastern Europe after the regime-changes of 1989 are all based on such a
diplomatic model, which had its origins in prior association agreements.28 The
absence in such cases of binding arbitration procedures for trade disputes is
hardly surprising, since the Europe Agreements were negotiated for essentially

     See europa.eu.int/comm/development/body/cotonou/index_en.htm. (visited 10 April 2006).
     ACP–EU Cotonou Partnership Agreement, signed 23 June 2000, in force 1 April 2003 [2000]
OJ L317/3, Art. 98.
     See J. Gomula, ‘Dispute Settlement under Association Agreements with Central and Eastern
European States’ (1995) 22 Polish YIL, 107.
390                             Ignacio Garcia Bercero
political reasons and their main objective was to prepare countries for eventual
EU membership. It is also likely that recourse to legal adjudication was con-
sidered as potentially disruptive of important bilateral relationships.29 Once
countries in Central and Eastern Europe entered into enlargement negoti-
ations, most differences related to the extent to which the acquis was being
properly implemented. Although not explicitly stated, it may be assumed that
the leverage provided by the process of enlargement negotiations was con-
sidered a more effective instrument to achieve compliance than legal adjudica-
tion. Moreover, the frequent references in the Europe Agreements to the EU
acquis raise complex legal issues as regards the extent to which a bilateral
arbitration panel may apply EU law, thereby increasing the reluctance to have
recourse to arbitration.
   The Europe Agreements provided the model for the Stabilization and
Association Agreements with countries in the Western Balkans, which respond
to a similar political rationale. It is, perhaps, less clear why the Euro-
Mediterranean Agreements include dispute settlement provisions which are
almost identical to those in the agreements negotiated with countries that are
candidates for EU accession. Since the political sensitivity of the bilateral
relationship or the preference for asymmetrical leverage are less relevant in
the absence of a membership perspective, it may be assumed that until the
EU–Mexico FTA was concluded in the year 2000, there was a consistent EU
preference for a non-judicial approach to bilateral dispute settlement. While
some of the Euro-Mediterranean Agreements and the Stabilization and
Association Agreements were concluded after 2000, the maintenance of the
traditional diplomatic model was mainly due to institutional conservatism and
the tendency to follow the same approach for all agreements concluded within
a region. Indeed, as previously noted, the basic features of the diplomatic
model of dispute resolution predate the Europe Agreements and, until recently,
had been applied to all EU association agreements, including both their trade
and non-trade related provisions.
   This preference for non-judicial dispute settlement can therefore be best
explained in historical terms. Prior to the entry into force of the WTO Agree-
ments, there was very limited experience with legal adjudication as an instru-
ment to solve commercial disputes. GATT panels were only established and
adopted by consensus and disputes were often solved through complex nego-
tiations, including the threat (or actual application) of retaliation. The first
free trade agreement to include binding arbitration procedures was NAFTA,
which was concluded in 1992, i.e. shortly before the conclusion of negoti-
ations of the new WTO Dispute Settlement Understanding. It took some time

     The Trade Barriers Regulation (Council Regulation (EC) No 3286/94 [1994] OJ L349/71)
excluded from the procedure for industry complaints the enforcement of bilateral agreements.
The main reason was concern by some Member States that the TBR could result in difficult to
solve disputes under the Europe Agreements. For a discussion of the TBR, see M. Bronckers and
N. McNelis, ‘The EU Trade Barriers Regulation Comes of Age’ (2001) 35 JWT, 427, at 434.
                  Dispute Settlement in EU Free Trade Agreements                              391
before countries became used to the new procedures under the WTO and
sought to include similar provisions in bilateral agreements.
   This historical explanation is further corroborated by the fact that the first
EU agreement to include binding arbitration was the FTA with Mexico. The
EU had taken some timid steps towards more effective arbitration procedures
in the FTA with South Africa and in the Cotonou Agreement. It was, however,
Mexico that took the initiative of requesting the establishment of detailed and
binding arbitration procedures, which were largely based on NAFTA. Whereas
the EU had a number of reservations with the NAFTA model, it accepted the
basic rationale for moving towards a quasi-judicial dispute resolution in the
trade field.
   This shift of EU preferences can be best explained through a combination of
three factors. First, as already noted, the EU had by this time become a major
user of WTO dispute settlement and could therefore see the advantages of
quasi-judicial procedures as a means of seeking neutral and less political reso-
lution of disputes. Second, the EU–Mexico Agreement was the first FTA in
which leadership within the Commission was allocated to DG Trade, which is
institutionally more inclined towards procedures that ensure a rapid solution
of commercial disputes. Finally, and most importantly, there was growing
dissatisfaction within the EU with the effectiveness of the traditional diplomatic
approach to dispute resolution.
   In order to better understand the weaknesses of the traditional diplomatic
model of dispute resolution, it is useful to refer briefly to the key elements of
the dispute settlement clauses included in EU association agreements. These
are as follows:
(a) Any of the parties may refer to the Association Council any dispute con-
    cerning the application or interpretation of the Association Agreement.
    The Council may, by consensus, adopt a binding decision to resolve the
    dispute. In practice this procedure is used to encourage consultations and
    diplomatic negotiations to reach a mutually acceptable solution. This dip-
    lomatic approach has often proved effective to solve a number of ‘low
    profile’ trade irritants. Indeed, consultations remain the preferred means
    to solve disputes under the quasi-judicial procedures followed in more
    recent agreements. However, in the absence of a credible arbitration option,
    there have been numerous examples of issues that figure year after year in
    the agendas of the Association Council without any effective resolution.30
    This is particularly so if influential producer interests are affected. In those
    circumstances, diplomatic negotiations may prove ineffective, whereas

     An example is provided by the dispute concerning the non-adoption by Turkey of the EU-
acquis as regards the protection of data exclusivity for pharmaceuticals. This issue had been first
raised in the Association Council in April 2002. However, a partial solution has only been found in
2005 following a complaint under the Trade Barriers Regulation. The TBR report recommended a
negotiated solution, joint reference to the ECJ, or recourse to WTO dispute settlement.
392                                Ignacio Garcia Bercero
    the EU, in the absence of a major economic impact, may be reluctant to
    exercise political leverage or apply retaliatory measures.
(b) In the event that a dispute is not settled through a decision of the Associ-
    ation Council, any of the parties may have recourse to arbitration. Each of
    the parties appoints an arbitrator and the third arbitrator is appointed by
    the Association Council. The parties are required to take steps to imple-
    ment the decision of the arbitrators. This procedure has three obvious
    (i) arbitration procedures can be blocked if one of the parties fails to
          appoint an arbitrator or opposes a consensus decision on the third
    (ii) there are no deadlines or established procedures for the conduct of
          the arbitration, and
    (iii) there are no procedures for compliance with the arbitration ruling.
    In the light of this, it is not surprising that the arbitration procedure has
    hardly ever been used.
(c) Most agreements also include a non-execution clause, which allows a
    party to take ‘appropriate measures’ if it considers that the other party
    has failed to fulfil an obligation under the agreement. Normally the parties
    should first seek a solution within the framework of the Association
    Council, although in cases of urgency, measures may be taken immedi-
    ately. In comparison with the retaliation options provided under WTO-
    type dispute settlement procedures, ‘appropriate measures’ have two
    distinguishing features:
    (i) recourse to dispute settlement procedures is not a precondition for the
          application of remedial action, and
    (ii) there are no clearly defined limits on the measures that may be
          adopted as a response to a failure by the other party to fulfil its
          obligations under the agreement.32
    It might have been thought that the high degree of self-help involved in
    the procedure leaves the EU—as the strongest party in the bilateral
    relationship—with substantial leverage to ensure asymmetrical compli-
    ance. In practice, however, this procedure has been used very sparingly

      A party that blocks arbitration proceedings can be said to be in breach of its obligations under
the agreement, thereby triggering the right of the other party to resort to other remedies under
international law, including invocation of the non-execution clause. In practice, however, the
parties may be reluctant to have recourse to such ‘exceptional’ remedies and prefer to continue
efforts to seek a diplomatic solution to a dispute. The absence of a pre-established procedure with
clear time limits acts as a disincentive for having recourse to arbitration. For instance, in a case
concerning the Ukraine, arbitration proceedings were not pursued because of a disagreement on
how to share the costs of the arbitration.
      Non-execution clauses normally provide that the measures adopted must be those that ‘least
disturb the functioning of the agreement’. Some also include reference to the principle of ‘pro-
portionality’ or to the need to ensure consistency with international law. For a discussion on the
non-execution clause, see L. Bartels, Human Rights Conditionality in the EU’s International
Agreements (Oxford: OUP, 2005), 29, 131.
                Dispute Settlement in EU Free Trade Agreements                         393
    and most of the cases involved Russia, a partner with not insignificant
    bargaining power. A possible explanation is that decision making in the
    EU Council on the adoption of retaliatory measures is facilitated if such
    action is as the outcome of a quasi-judicial adjudication procedure.33

    iv. towards a quasi-judicial model of adjudication:
        the eu-mexico (2000) and eu-chile (2002) ftas

As noted in the prior section, the introduction of quasi-judicial procedure for
dispute settlement was the result of dissatisfaction with the effectiveness of the
diplomatic model as an instrument to solve trade disputes, experience with
WTO dispute settlement and the demands of EU’s trading partners. Some of
the more obvious weaknesses of the standard arbitration clauses in EU agree-
ments were already corrected in the 1999 EU–South Africa Trade and Devel-
opment Cooperation Agreement (TDCA) and in the 2000 Cotonou Agreement.
However, a fully developed arbitration procedure has only been introduced, so
far, in the FTAs concluded with Mexico and Chile. The EU–Mexico Agreement
includes a Title on Dispute Settlement comprising seven articles, as well as a
separate set of rules of procedure and a Code of Conduct for arbitrators. The
EU–Chile Agreement has a Title on Dispute Settlement comprising nine articles,
rules of procedure and a Code of Conduct. These innovations only relate,
however, to the trade and trade-related parts of the agreements. No dispute
settlement has been envisaged as regards the political or co-operation provi-
sions of the agreements. It should be noted that the EU–South Africa TDCA
was different in that the traditional dispute settlement model was retained for
economic, development, financial or other areas of cooperation, while stricter
time-limits were limited to disputes relating to trade or trade-related areas.
   The main source of the arbitration procedures in the two agreements is WTO
dispute settlement, which provides the basic structure of a procedure based on a
consultation phase, arbitration panels, an obligation to comply within a ‘rea-
sonable period of time’ and commensurate retaliation in case of lack of compli-
ance (or agreement on temporary compensation). At the same time, the parties
tried to improve on WTO dispute settlement, in particular, in order to have
faster procedures. The parties—and, particularly, the EU and Chile—also
experimented with some of the ideas being discussed in the WTO DSU negoti-
ations. NAFTA was also an important source for the EU–Mexico Agreement,
particularly as regards the Rules of Procedure and the Code of Conduct.
   The arbitration procedures in the EU–Mexico and the EU–Chile agree-
ments have many elements in common, but also important differences. For
ease of reference, Table 1 summarizes the main differences between the two
     The Council adopted Regulations on retaliatory measures following WTO disputes with the
US concerning steel safeguards, the Foreign Sales Cooperation, and the so-called Byrd
Table 1.(*) Main Differences in the Dispute Settlement Mechanism
FTAs EC–Mexico and EU–Chile

LEGAL PROVISIONS                          EC/MEXICO                                  EC/CHILE FTA
                                          OJ L157/26 of 30.06.2000                   OJ L353 of 30.12.2002

Forum Exclusion                           WTO and the procedures of the FTA,         WTO and the procedures of the FTA (Art. 189(4)(c)) are
                                          are not mutually exclusive but the         mutually exclusive
                                          proceedings can not be concurrent
                                          (Art. 47(4))
Transparency                              Not regulated                              Hearings: Opportunity to the Parties to have open
                                                                                     hearings (Rule 23)
                                                                                     Amicus curiae: submissions are allowed but the
                                                                                     arbitration panel shall not be obliged to address them, in
                                                                                     its ruling (Rule 35 to 37)
Establishment of the panel (timeframes)   Within 45 days of the request for the      Within 3 days of the request for the establishment of the
                                          establishment, the panel shall be          arbitration the panel shall by constituted (Art. 185).
                                          constituted (Art. 44).
Composition of the panel                  Panel composed of 3 arbitrators.           Panel composed of 3 arbitrators selected by lot, from a
                                          Each Party appoint one arbitrator.         roster of 15 persons (5 EU, 5 Chile, 5 non-nationals).
                                          Each party proposes up to 3                (Art. 185). Roster to be established by Association
                                          arbitrators to serve as chair. If one of   Committee no later than 6 months after entry into force of
                                          the parties fail to propose an             agreement.
                                          arbitrator or if there is no agreement
                                          on the chair, they will be chosen by
                                          lot (Art. 44)
Initial report                          Initial Report will be issued within 3   No initial report.
                                        or 5 months from the constitution of
                                        the panel. (Art. 45(1))
Timeframe for panel examination         Final Report shall be issued within 30   Final report will be issued in 3 or 5 months from the
                                        days form the presentation of the        constitution of the panel. (Art. 187(1))
                                        initial report (Art. 45(2)) (Initial +
                                        Final = 4 or 6 months)
Implementation procedures               The Party concerned shall notify the     The complained against party will notify the measures
                                        measures adopted in order to             required to comply, a reasonable period of time (RPT) for
                                        implement the final report before the     doing so, and a concrete proposal for a temporary
                                        expiry of the reasonable period of       compensation until full implementation (Art. 188(3)).
                                        time (RPT) previously determined         In case of disagreement of any of the 3 previous issues,
                                        (either by agreement of the parties or   there is the possibility to ask arbitration.
                                        by arbitration, the ruling should be     If the party does not agree with the notification, could ask
                                        given within 15 days). (Art. 46)         the original panel to make a ruling within 45 days (Art.

(*) Source: Edna Ramirez Robles, 2006
396                               Ignacio Garcia Bercero
agreements. What follows is a presentation of the key features of the new
‘quasi-judicial’ approach:

A. Shorter timeframes than under the WTO DSU

A potential advantage of bilateral dispute settlement is to secure speedier reso-
lution of commercial disputes than is possible under the WTO DSU. This was
an important consideration when negotiating the arbitration procedures
under both agreements. The main source of procedural time-savings is the
absence of an appellate review, which in any event would have been difficult and
unusual to establish within the institutional framework of a bilateral agree-
ment. There are, in addition, other time-saving provisions. The mandatory
consultation period has been reduced from 60 days under the DSU to 45 days.
The arbitration panel is to be established within no more than 45 days (Mexico)
and within three days (Chile). The shorter timeframe in the EU–Chile Agree-
ment is due to the differences in panel-establishment procedures, which are
discussed below. As regards the timeframe for the issuance of a final report
(since the date of establishment of a panel), this is four to six months (Mexico)
and three to five months (Chile). The reason for this difference is the elimi-
nation in the EU–Chile Agreement of the DSU requirement to prepare an
interim panel report. As a result of all these time-savings, under the EU–Chile
Agreement, it should in principle be possible to obtain a definitive ruling on a
dispute within less than seven months of a complaint being launched.34

B. Automatic procedure for the establishment of an arbitration panel

Under both agreements the need for consensus in the Association Council
disappears, although the institutional solutions found to avoid blockages are
different. Under the EU–Mexico Agreement, each of the parties is to appoint
one arbitrator and propose up to three candidates as a Chair. If no agreement is
reached on the Chair, or if a party fails to appoint its arbitrator, the remaining
arbitrators are selected by lot. In negotiating these procedures, an important
consideration for both the EU and Mexico was to avoid the procedural com-
plexities of the panel composition procedure under NAFTA, which had often
resulted in a failure by the parties to agree on a panel.35
   The EU–Chile panel composition procedure is more innovative in that
it departs from the DSU model, which relies on efforts by the parties to agree
on ad-hoc panelists. Instead, the parties are to agree on a roster of fifteen
      A word of caution is, however, necessary since, under the WTO, Panels have often taken
longer to finalize their work than the time periods stipulated in the DSU. This is despite the fact
that panelists can rely on the support of the WTO secretariat. It can not be excluded, therefore,
that bilateral panelists, which have very limited institutional support, may require more time to
issue a ruling, particularly in complex disputes.
      On NAFTA dispute settlement, see A.L.C. de Mestral ‘NAFTA Dispute Settlement: Creative
Experiment or Confusion?’, in this volume, at xxx.
                 Dispute Settlement in EU Free Trade Agreements                             397
panelists—five from Chile, five from the EU and five non-nationals. All three
arbitrators are then selected by lot within three days of the request for the
establishment of the panel. Unlike in the case of NAFTA, the establishment of
a roster is conceived as an alternative to the ad-hoc selection procedure. Despite
this, it has proved easier than in NAFTA to agree on the roster, which has
already been established.36

C. Detailed procedural rules for the arbitration

Whereas prior EU agreements only had one article to regulate all aspects of
dispute settlement procedures, the FTAs with Mexico and Chile include
detailed rules on arbitration procedures both in the main body of the agreement
and in the annexes on rules of procedure and Code of Conduct. As already
noted, the procedural rules are a combination of DSU and NAFTA rules. In
one important aspect, however, the EU–Chile Agreement is more innovative
than either the DSU or NAFTA (and, indeed, than the EU–Mexico Agreement).
The EU–Chile Agreement explicitly provides that rulings should be made pub-
lic and that, provided both parties agree, hearings may also be opened to
the public. It also provides for the admissibility, under certain conditions, of
amicus curiae. These transparency provisions are essentially identical to those
proposed by the EU in the DSU negotiations. The agreement with Chile was
the first bilateral FTA to provide for open hearings and admissibility of amici,
although similar provisions have been included in agreements subsequently
negotiated by the US.

D. Compliance proceedings

Both agreements have established detailed compliance proceedings inspired
by Articles 21 and 22 of the DSU, although they go beyond the current rules
of DSU in two respects: (a) the establishment of rules on sequencing—i.e.
retaliation is only allowed if the original panel rules that the measures taken
to comply with the initial ruling are still in breach of the FTA, and (b) a
procedure to allow for the lifting of sanctions if compliance is achieved after
concessions have been suspended. These provisions are very similar to rules
supported by most WTO Members in ongoing DSU negotiations.37
   The EU–Chile Agreement is again more innovative. In particular, it provides
that, within thirty days of the ruling, the defending party must notify: (a)
the specific measures it intends to take in order to comply, (b) a proposed

      The NAFTA roster is larger than in EU–Chile—30 individuals—and arbitrators are still
selected by the parties rather than by lot. At the time of writing, it appears that NAFTA parties
have not yet agreed on the roster.
      For a discussion of the DSU Negotiations, see I. Garcia Bercero and P. Garzotti, ‘DSU
Reform—Why have Negotiations to Improve WTO Dispute Settlement failed so far and what are
the Underlying Issues’ (2005) J World Invest & Trade, 847.
398                               Ignacio Garcia Bercero
timeframe for compliance, and (c) a concrete proposal on temporary compen-
sation until full implementation. If there is disagreement on any of these three
issues, the matter may be referred to arbitration by the same panel that ruled
in the original dispute. This procedure is more far-reaching than what has
been proposed so far in the DSU negotiations. First, it requires the defending
party to indicate, within a very short deadline, the specific measures it intends
to take to comply with the ruling. Moreover, the arbitration panel is given
jurisdiction to rule on a measure before it has been adopted. This represents a
significant departure from DSU rules and practice, under which compliance
proceedings normally only start at the end of the reasonable period of time
and where there is no obligation to formally notify in advance the measures a
party intends to take in order to comply. Second, an obligation is introduced
to present an offer for temporary compensation during the time required to
achieve compliance with bilateral obligations. The arbitration panel has juris-
diction to rule on whether the compensation offer presented is ‘manifestly
disproportionate’. The obligation to present a compensation offer is, however,
non-enforceable since retaliation may only be applied in case there is lack of
compliance at the end of the reasonable period of time.
   Both agreements include rules on retaliation which are similar to those in
the DSU. The original arbitration panel may be asked to rule on whether a
proposed suspension of concessions is equivalent to the level of nullification and
impairment. The arbitrator may also rule on whether the principles governing
the selection of countermeasures have been followed. These principles require
specific justification in case the suspension relates to a different title to that in
which the arbitrators have found a breach and to the need to select measures
that ‘least disturbs the functioning of this agreement’.38 It should be noted that
parties are only authorized to ‘suspend the application of benefits granted
under this Part of the Agreement . . .’. This appears to exclude the suspension
of benefits under the non-trade related part of the agreement.39

      Under the EU–Chile Agreement, the trade-related part includes separate titles on Free Move-
ment of Goods, Trade in Services and Establishment, Government Procurement, Current payments
and Capital movements, Intellectual Property Rights, and Competition (this last title is excluded
from the application of Dispute Settlement).
      In principle, it would appear that retaliation could include measures which are in breach of
obligations under the WTO Agreement. However, the application of such retaliatory measures
would give rise to complex issues as regards the relationship between bilateral and WTO dispute
settlement. An important question is whether a WTO Panel has jurisdiction to apply WTO law in
relation to countermeasures that have been properly adopted under a bilateral dispute settlement
procedure. This issue underlies the WTO dispute between US and Mexico on discriminatory taxes
applied by Mexico on US sugar products as a retaliation against a US breach of NAFTA provi-
sions, although in this case Mexico’s retaliation was not based on the provisions of NAFTA but
was rather a response to the obstruction by the US of NAFTA procedures. The Appellate Body,
however, has ruled that it has no jurisdiction to adjudicate non-WTO disputes. (See WTO Appellate
Body Report, Mexico—Soft Drinks, WT/DS308/AB/R, adopted 6 March 2006, para. 56). It should
be noted that these findings were not strictly necessary to solve the dispute and could be regarded
as obiter.
                Dispute Settlement in EU Free Trade Agreements                         399
   As a conclusion, it should be noted that both the EU–Mexico and the EU–
Chile Agreements represent a radical departure from prior EU practice. In
both cases the main purpose was to introduce an arbitration proceeding
for trade disputes based on the WTO model, but with reduced timeframes.
The EU–Chile FTA represents a more significant departure from the estab-
lished WTO model, because of the willingness of the parties to experiment
with a number of ideas being discussed in the DSU negotiations. Some of
these innovations are close to the current consensus of the broader WTO
membership—i.e. sequencing. Others may be more difficult to emulate at the
multilateral level-procedures for amicus curiae, establishment of a roster of
panelists, obligation to present a compensation offer. On one issue, however,
the two agreements reached highly contrasting approaches. This is the vexing
question of the relationship between bilateral and WTO dispute settlement.
This issue is discussed in the next section.

  v. an unresolved issue: articulation between bilateral
          dispute settlement and wto proceedings

The relationship between the dispute settlement proceedings of the WTO and
of FTAs is a subject of great legal complexity. This article will not consider
the extent to which there may be overlaps and conflicts of jurisdiction between
the two legal orders or whether WTO dispute settlement proceedings may
take into account provisions of regional trade agreements or the fact that the
same measure may be (or have been) subject of bilateral dispute settlement
proceedings under a regional trade agreement.40 The focus of the present
inquiry is limited to explaining how this relationship has been addressed in
recent EU FTAs and to suggest that further reflection on the relationship
between bilateral dispute settlement and the WTO is needed, since neither of
the approaches followed in the agreements with Mexico and Chile are fully
  The complexity of finding an ‘ideal’ solution to the relationship between
WTO and bilateral dispute settlement procedures is due to the fact that the
substantive law of free trade agreements is very largely influenced—and,
indeed, dependent upon—WTO law. In order to illustrate this point, it is
useful to consider some examples drawn from the substantive provisions of
the EU–Chile FTA. On the basis of these examples, it appears possible to
distinguish at least four possible ‘types’ of relationship between WTO law and
the law of the FTA.

     See the contribution by K. Kwak and G. Marceau, ‘Overlaps and Conflicts of Jurisdiction
between the World Trade Organization and Regional Trade Agreements’, in this volume, at xxx.
400                                Ignacio Garcia Bercero
A. Bilateral provisions which are largely independent from WTO obligations

Examples are the preferential schedules of market access commitments for
goods and services, the provisions on the right of establishment or capital
movements (areas not regulated by WTO rules), or the title on government
procurement (since Chile is not a party to the WTO plurilateral Government
Procurement Agreement). However, even in these areas, WTO law may exercise
an influence since many of the underlying legal concepts are common in WTO
and FTA law. It can not be excluded, therefore, that in interpreting the mean-
ing of a market access concession or of a term in the government procurement
title, a bilateral arbitration panel may wish to consider how similar issues have
been addressed in WTO jurisprudence. Indeed, it may well be argued that, in
order to enhance the predictability and coherence of international trade law,
it would be highly desirable if bilateral panels were to consider relevant
WTO rulings.

B. Areas where the FTA essentially confirms WTO obligations

The clearest example is Article 78 on Antidumping and Countervailing Meas-
ures. This article states that the parties retain the right to apply such measures
in accordance with the relevant WTO provisions. Presumably, the reason for
including such a provision in a bilateral agreement is simply to confirm that
the right to apply such measures is not affected by the FTA. Regardless of
whether the parties explicitly exclude the application of bilateral dispute
settlement to such provisions or not, it could hardly be their intention that a
bilateral panel adjudicates on the interpretation and application of the WTO
Antidumping Agreement.41 A more complex case is Article 92 on the Safe-
guard Clause. Article 92(1) essentially confirms the application of Article XIX
GATT and the Safeguards Agreement. However, the article also provides for a
number of obligations that are additional to those provided for in the WTO.
While most of these obligations are procedural—and, thus, not easily subject
to adjudication—at least one of the provisions is substantive and largely
independent from WTO law. It can not be excluded therefore that a bilateral
panel may be asked to adjudicate on the interpretation and application of
Article 92(6) of the EU–Chile Agreement.42

      The EU–Mexico Agreement explicitly excludes the application of bilateral dispute settlement
to antidumping and countervailing measures, whereas EU–Chile is silent on the matter. However,
in both cases, the purpose of referring to the Antidumping Agreement is to confirm an exception
from the obligation to eliminate duties in bilateral trade.
      Art. 92(6) obliges the parties to select safeguard measures which least disturb the achievement
of the objectives of the agreement, which do not exceed what is necessary to remedy serious injury,
and which preserve the level/margin of preference under the agreement.
                Dispute Settlement in EU Free Trade Agreements                        401
C. Areas of overlap between WTO and bilateral obligations

In certain areas, parties to an FTA seek to establish rules that go beyond WTO
obligations (so-called WTO-plus provisions). However, in doing so the nego-
tiators are not drawing on a blank piece of paper since often the point of
departure is a confirmation of existing WTO obligations. For instance, the
EU–Chile Agreement seeks to develop disciplines on technical standards and
regulations and on sanitary and phytosanitary measures which go beyond the
provisions of the WTO, TBT, and SPS Agreements. (Arguably, however,
many of the provisions included in the agreement are non-binding in nature).
At the same time, the FTA confirms WTO obligations under the TBT and SPS
Agreements and clarifies that a number of terms are used with the same mean-
ing than they have under the WTO Agreements. In such cases, it is difficult
to see how a bilateral panel could properly apply the provisions of the FTA
without also referring to WTO law. To a certain extent, similar issues can arise
as regards the EU–Chile provisions on trade facilitation, intellectual property
and, possibly, customs valuation.

D. Areas where the FTA reproduces (without referring to) WTO provisions

A clear example is Article 77 of the EU–Chile Agreement on national treatment
of internal taxes and regulations which simply reproduces Article III GATT.
Similarly, the regulatory provisions on services are largely modelled on the
equivalent provisions of the GATS. Even if no explicit reference is made to the
WTO, it is difficult to envisage how a bilateral panel could interpret these
provisions without making reference to relevant rulings of the WTO Dispute
Settlement Body. NAFTA panels, for instance, have often referred to GATT
and WTO law in their rulings.43
  This complex relationship between WTO law and FTA law gives rise to two
questions: (a) May bilateral arbitration panels apply WTO law which is
reproduced or referred to in the provisions of the FTA or, at least, take into
account WTO provisions for the interpretation of FTA law? and (b) Do the
parties retain the freedom to use both bilateral and WTO dispute settlement or
do they need to operate a choice of forum? Both issues were highly contro-
versial in the negotiations with Mexico and Chile and could only be solved at
the final stage of the negotiations. As will be seen the solutions found are not
only different, but also not fully satisfactory.

1. Application of WTO Law by bilateral panels
As was already the case in the EU–South Africa FTA, the agreement with
Mexico provides that ‘Arbitration proceedings established under this title will
     For a theoretical framework on the extent to which WTO law may be applied by bilateral
panels, including a discussion of NAFTA practice, see L. Hsu, ‘Applicability of WTO law in
Regional Trade Agreements: Identifying the Links’, in this volume, at xxx.
402                              Ignacio Garcia Bercero
not consider issues relating to each party’s rights and obligations under the
Agreement establishing the World Trade Organization’ (Article 47(3)). More-
over, the articles that have a closer connection to WTO rules—antidumping
and countervailing measures, balance of payments restrictions, trade agree-
ments with third parties, and intellectual property—are explicitly excluded
from the application of bilateral dispute settlement.
   The EU–Chile FTA only explicitly excludes from bilateral dispute settlement
a non-WTO related area—i.e. competition. It includes, however, a provision
that, if broadly interpreted, could result in bilateral panels having to decline
jurisdiction whenever WTO law appears to be relevant for the resolution of a
dispute. Article 189(4)(c) reads as follows: ‘Unless the Parties otherwise agree,
when a Party seeks redress of a violation of an obligation under this part of the
agreement which is equivalent in substance to an obligation under the WTO, it
shall have recourse to the relevant rules and procedures of the WTO . . .’.
   What are the consequences of these two approaches for the jurisdiction of a
bilateral panel to apply WTO law or to use WTO law as a tool for the inter-
pretation of provisions in the bilateral agreement that are closely related to
WTO law? First, it would appear that under both agreements a bilateral panel
may not rule on bilateral provisions that go no further than a confirmation of
existing WTO rules and obligations. In any event, in order to increase legal
certainty, the EU–Mexico FTA explicitly excludes a number of provisions that
appear to fall under this category. Second, the EU–Chile FTA appears to go
much further in that it excludes—unless the parties agree—bilateral provisions
that are equivalent in substance to WTO obligations, even if the FTA may not
at all refer to WTO law. Thus, it would appear that, under such a clause, the
responding party could object to the jurisdiction of a bilateral panel to rule,
for instance, on a breach of national treatment on the grounds that the
bilateral obligation is substantially equivalent to Article III of GATT. Third, it
is submitted that bilateral panels should always be able to use WTO law as a
tool to interpret FTA provisions. This appears to be, in any event, required
under public international law and both EU–Mexico and EU–Chile indicate
that arbitration panels shall interpret the provisions of the agreement in
accordance with customary rules of public international law.44 Finally, in so
far as interpretation is not a sufficient tool to take WTO law into account,
none of the approaches above provides a satisfactory solution to address areas
in which there is a genuine overlap between WTO and bilateral obligations. In
those cases, it may be impossible for a bilateral panel to apply a bilateral
provision without at the same time considering issues relating to each party’s
rights and obligations under the WTO. Moreover, recourse to WTO dispute

      An unfortunate legal ambiguity has been created with the requirement that arbitration pro-
ceedings should not ‘consider’ issues relating to WTO rights and obligations. A strict reading of
such provision could even imply a prohibition to use WTO law as a tool to interpret FTA provi-
sions. It is submitted that such a ‘clinical isolation’ of FTA and WTO law would be highly
                Dispute Settlement in EU Free Trade Agreements                         403
settlement does not provide an alternative, since WTO panels may not enforce
the WTO plus elements included in the FTA.

2. Choice of Forum
Under the NAFTA and subsequent US FTAs, there is a ‘choice of forum’ clause
which essentially provides that: (a) the complaining party may select whether
to refer a dispute to the WTO or bilateral dispute settlement, and (b) once a
panel has been requested under either one of these fora, the forum selected
shall be used to the exclusion of the others.
   In the negotiations of the EU–Mexico FTA, Mexico took the position that
there should be an equivalent ‘choice of forum’ clause to that included in
NAFTA. The EU was strongly opposed to the NAFTA approach. It was
argued that the consequence of a ‘choice of forum’ clause would be that,
whenever both options were available, the complainant would have a clear
preference to have recourse to WTO dispute settlement, since the DSU was
well tested. Moreover, in the light of the evolutionary nature of WTO law and
jurisprudence, there was a reluctance to any procedural solution which
implied giving up future WTO rights. The compromise reached was a rule
against concurrent proceedings. Thus, while there is no forum exclusivity,
once a forum has been chosen, proceedings in a different forum may only be
initiated after the first proceedings have ended. This is intended to limit the
risk of procedural harassment by the complainant.
   The compromise above was not acceptable to Chile. Thus the EU–Chile
Agreement includes a rule on forum exclusivity. However, within 10 days of
the establishment of a bilateral panel, a jurisdictional objection may be raised,
on which a preliminary ruling must be made within 30 days. In practice, this
means that the defendant would normally be able to avoid a bilateral proceed-
ing if the claims by the complainant include a breach of bilateral provisions
that are substantially equivalent to WTO obligations. As indicated above, the
consequence is likely to be a much greater reluctance by the parties to have
recourse to bilateral dispute settlement.
   To sum up, it would appear that a simpler and more effective solution
should be found to the relationship between bilateral and WTO dispute
settlement. In this connection, the author would favour the following
approach: (a) bilateral dispute settlement panels should be in a position to
apply WTO law which is referred to in a bilateral FTA, unless the parties have
explicitly excluded the applicability of bilateral dispute settlement to a particu-
lar provision.45 Article 41(3) of the EU–Mexico FTA is not satisfactory since
it can limit the scope to apply WTO-plus provisions, where such provision

      Some may argue that such an approach contradicts Art. 23.1 DSU, under which WTO
Members may only seek a redress of a violation of WTO law through WTO dispute settlement
proceedings. However, from the moment WTO law is incorporated into a bilateral agreement, it
becomes part and parcel of the bilateral rights and obligations which may be enforced under
bilateral dispute settlement.
404                         Ignacio Garcia Bercero
can only be applied in conjunction with existing WTO obligations. Article
189(4)(c) of EU–Chile is, of course, even more limitative and could seriously
impair the effectiveness of bilateral dispute settlement. b) A rule against con-
current proceedings is sufficient to prevent any risk of legal harassment. A rule
on forum exclusion is, on the other hand, likely to be to the detriment of
bilateral dispute settlement except for issues not covered by WTO disciplines,
since complainants are likely to be reluctant to lose WTO remedies.

                 vi. conclusions: lessons learned?

The evolution towards a quasi-judicial model of adjudication in EU FTAs
appears now irreversible. Ongoing negotiations of FTAs with Mercosur and
the GCC include dispute settlement provisions which are similar to those
included in the FTAs concluded with Mexico and Chile. The initialled EU–Syria
FTA is the first Euro-Mediterranean Agreement which includes quasi-judicial
dispute settlement procedures. Recently, the EU Council authorized the
Commission to open negotiations with all Euro-Mediterranean countries with
a view to introducing more effective dispute settlement procedures. Consider-
ation of such an approach is also being discussed in ongoing negotiations for
the conclusion of an SAA with Serbia and Montenegro. Indeed, quasi-judicial
dispute settlement appears to be the norm in all recently concluded FTAs, not
only of the EU, but of all major trading partners.
   Despite the above, it is perhaps premature to conclude that lessons have
been learned and that the EU has decided to move away from the traditional
diplomatic approach of bilateral dispute resolution. After all, at the time of
writing, there has not been a single case in which the dispute settlement provi-
sions of the FTAs with Mexico and Chile have been invoked. This contrasts
with the relatively frequent recourse by the EU to WTO dispute settlement.
What explains this limited recourse to the new quasi-judicial dispute settlement
   A first possible reason is that the trend towards quasi-judicial dispute settle-
ment is still very recent and limited to a few trading partners, which are not
frequent ‘targets’ of EU dispute settlement. In these circumstances, it is likely
that the traditional diplomatic approach to dispute resolution is still preferred.
The fact that the EU Trade Barriers Regulation does not allow industry com-
plaints to enforce obligations under FTAs may also limit the potential for
recourse to bilateral dispute settlement. It can be expected, however, that both
the EU (and its trading partners) will over time overcome this initial reluctance
to test the new quasi-judicial procedures. One or two successful examples may
be all it takes for recourse to bilateral arbitration to become accepted as a
‘normal’ occurrence in bilateral trade relations. It may also be the case that the
mere existence of quasi-judicial arbitration procedures acts as an incentive to
solve more disputes at the consultation/diplomatic phase.
                   Dispute Settlement in EU Free Trade Agreements                         405
   Another possible reason relates to the substantive normative content of EU
FTAs. As indicated above, bilateral dispute settlement is not applicable to
disputes concerning trade remedies measures (i.e. antidumping, countervailing
duties, and safeguards). These are, however, a large majority of the cases that
the EU (and other WTO Members) bring to WTO dispute settlement. Indeed,
the only WTO case launched by the EU vis-à-vis Mexico concerns the applica-
tion of countervailing duties on olive oil and the only WTO case launched by
Chile against the EU concerned a safeguard measure on salmon. None of
these cases could have been brought under the bilateral dispute settlement
procedures in the FTAs. Bilateral dispute settlement would appear to be par-
ticularly appropriate in the case of WTO plus provisions in the FTAs. How-
ever, most of these relate to the schedules of market access concessions, which
normally are faithfully implemented and are not a source of trade disputes. In
case future FTAs were to include more ambitious WTO plus provisions in
regulatory areas, it could be expected that more frequent recourse would be
made to bilateral dispute settlement.
   Finally, there is possibly a more structural reason. As Bill Davey has noted,
WTO dispute settlement may be viewed as more legitimate and effective
both in terms of the adjudication process and the pressure to comply with a
ruling.46 This is due to a combination of factors, including the existence of a
well established WTO jurisprudence, the role of the Appellate Body in ensur-
ing high legal standards and coherence of rulings, the support provided to
panels by the WTO Secretariat, and the existence of multilateral surveillance
for the implementation of rulings. As a result of these factors, it is likely that if
a certain trade dispute can be addressed under either bilateral or WTO dispute
settlement, there will still be a clear preference to have recourse to WTO
procedures. So, in all likelihood while bilateral dispute settlement will become
a more important part of EU toolbox, it will not eclipse the priority given to
WTO dispute settlement.

       See W.J. Davey, ‘Dispute Settlement in WTO and RTAs: A Comment’, in this volume, at xxx.