The WTO legality of the EU’s GSP+ arrangement

The WTO legality of the EU’s GSP+ arrangement
Dr Lorand Bartels*
Abstract
In EC – Tariff Preferences, the Appellate Body held that the WTO Enabling Clause permitted
developed countries to grant better tariff treatment to some developing countries than to
others, subject to certain conditions. It held further that these conditions were not met by the
EU’s so-called ‘drugs arrangement’, a system of additional preferences (normally duty free
treatment) for certain countries which the EU had determined were in need of special tariff
preferences, thanks to their involvement in combating the production and trafficking of
narcotics. In response to this ruling, when the EU renewed its GSP program in 2005, it
replaced its drugs arrangement and two similar, though less generous, labour and environment
arrangements with a new arrangement popularly known as the ‘GSP+ arrangement’. Under
this arrangement, additional tariff preferences (normally duty free treatment), were made
available to developing countries committing to ratify and implement a list of human rights
and good governance conventions. According to the EU, the GSP+ arrangement complies
with the Appellate Body’s interpretation of the Enabling Clause. This article argues that it
does not. This is firstly because of the substantive criteria chosen by the EU to select GSP+
beneficiaries, which do not meet the Appellate Body’s criteria for differential tariff treatment
of developing countries. Second, it is because the EU’s requirement that would-be
beneficiaries must have applied by a certain date, replicates the problem of the ‘closed list’ of
beneficiaries that was fatal to the earlier incarnation of the EU’s GSP program. The article
concludes with some suggestions for designing a GSP+ arrangement more likely to meet the
Appellate Body’s conditions than the EU’s present arrangement.
* Faculty of Law, University of Cambridge.
2
The WTO legality of the EU’s GSP+ arrangement
Dr Lorand Bartels*
In EC – Tariff Preferences, the Appellate Body held that the WTO Enabling Clause permitted
developed countries to grant better tariff treatment to some developing countries than to
others, subject to certain conditions.1 It held further that these conditions were not met by the
EU’s so-called ‘drugs arrangement’, a system of additional preferences (normally duty free
treatment) for certain countries which the EU had determined were in need of special tariff
preferences, thanks to their involvement in combating the production and trafficking of
narcotics.2 In response to this ruling, when the EU renewed its GSP program in 2005, it
replaced its drugs arrangement and two similar, though less generous, labour and environment
arrangements3 with a new arrangement popularly known as the ‘GSP+ arrangement’.4 Under
this arrangement, additional tariff preferences (normally duty free treatment), were made
available to developing countries committing to ratify and implement a list of human rights
and good governance conventions. One of the conditions was that these countries had to apply
by 31 October 2005, which was four months after the publication of the GSP Regulation. The
new arrangement applies from 1 January 2006 to 31 December 2008.5
According to the EU, the GSP+ arrangement complies with the Appellate Body’s
interpretation of the Enabling Clause. This article argues that it does not. This is firstly
because of the substantive criteria chosen by the EU to select GSP+ beneficiaries, which do
not meet the Appellate Body’s criteria for differential tariff treatment of developing countries.
Second, it is because the EU’s requirement that would-be beneficiaries must have applied by
a certain date, replicates the problem of the ‘closed list’ of beneficiaries that was fatal to the
earlier incarnation of the EU’s GSP program.
* Faculty of Law, University of Cambridge; presently Humboldt Fellow at the Max Planck Institute for
International Law in Heidelberg. This article has benefited greatly from comments and criticisms from
Robert Howse, Joost Pauwelyn, Melaku Geboya Desta, Gustavo Ribiero, Bryan Mercurio, Federico
Ortino, Nathalie Bernasconi, Margaret Young and Holger Hestermeyer and from the participants at the
Conference on WTO and Human Rights, Monash University Prato Centre, 21-22 June 2007.
1 WTO Appellate Body Report, EC – Tariff Preferences, WT/DS246/AB/R, adopted 20 April 2004.
2 See below at n 17.
3 Council Regulation 2501/2001 [2001] OJ L346/1. For discussion see Lorand Bartels, ‘The WTO
Enabling Clause and Positive Conditionality in the European Community’s GSP Program’ 6 JIEL 507
(2003).
4 Arts 8-11 of Council Regulation 980/2005 of 27 June 2005 [2005] OJ L169/1. All further references
to Articles are to this Regulation unless specified otherwise. The legislative history of the regulation is
as follows. A first proposal was set out in Commission Communication, Developing countries,
international trade and sustainable development: the function of the Community’s generalised system
of preferences (GSP) for the ten-year period from 2006 to 2015, COM (2004) 461, 7.7.2004, in which
the GSP+ arrangement was outlined in general terms, and to begin on 1 January 2006. This was
followed by a concrete Proposal for a Council Regulation applying a scheme of generalised tariff
preferences, COM (2004) 699, 20.10.2004, now to commence on 1 July 2005. In its Amended Proposal
for a Council Regulation applying a scheme of generalised tariff preferences, COM (2005) 43,
10.2.2005, the Commission proposed to bring forward the commencement date to 1 April 2005, in
order to assist countries affected by the recent tsunami, with a GSP+ application date of 31 May 2005.
This Proposal was not adopted, and is not referred to in the GSP Regulation.
5 Art 30(2) provides that the GSP Regulation shall apply until 31 December 2008, but also (and a little
ambiguously) that this date ‘shall not apply to the arrangements for least-developed countries’.
3
I. The GSP+ arrangement
As mentioned, the basic administrative condition for receiving GSP+ benefits is that the
applicant country must have applied for the special incentive preferences before 31 October
2005.6 The applicants had to meet a number of substantive conditions:
• An applicant had to ratify and implement a list of sixteen human rights conventions
by 31 October 2005,7 and give an undertaking that it would continue to maintain
ratification and accept regular monitoring and review of implementation.8
Exceptionally, applicants facing constitutional constraints had until 31 December
2006 to ratify and implement a maximum of two conventions.9 This list of
conventions is set out in Part A of Annex III to the GSP Regulation.
• An applicant had to ratify and implement at least seven of eleven listed ‘good
governance’ conventions by 31 October 2005,10 and the remainder by 31 December
2008,11 and give an undertaking that it would continue to maintain ratification and
accept regular monitoring and review of implementation.12 This list of conventions is
set out in Part B of Annex III to the GSP Regulation.
• An applicant had to be a ‘vulnerable’ country.13 This is defined in terms of three
cumulative conditions: (a) poverty (ie not classified by the World Bank as a highincome
country); (b) non-diversification of exports (ie the five largest sections
represent more than 75 per cent of its GSP-covered exports to the EU) and (c) share
of EU GSP-covered imports (ie no more than 1 per cent of these imports).14
The preferences may be withdrawn for reasons common to the other GSP arrangements,15 and
in addition if the relevant conventions are no longer incorporated into domestic legislation or
being effectively implemented.16
The original beneficiaries of the drugs arrangement were five Andean and six Central
American countries and Pakistan.17 Except for Pakistan, all of these are on the new list. These
6 Art 10. Art 11 specifies the procedure to be followed by the Commission in examining compliance
with these conditions.
7 Art 9(1)(a).
8 Art 9(1)(d).
9 Art 9(2). This derogation derives from a recommendation made by the European Parliament when it
was consulted on a draft of the regulation: see European Parliament legislative resolution on the
proposal for a Council regulation applying a scheme of generalised tariff preferences,
P6_TA(2005)0066, adopted 3.9.02 [2005] OJ C320E/145, Amendment 23. It was expressly intended to
El Salvador, which was done in Council Decision 978/2006 of 19 December 2006 [2006] OJ 365/86.
10 Art 9(1)(b).
11 Art 9(1)(c).
12 Art 9(1)(d).
13 Art 9(1)(e).
14 Art 9(3).
15 Arts 16 and 17.
16 Art 16(2). The Commission is also required to keep the ratification and effective implementation
status under review and report to the Council in time for the next Regulation: Art 9(4).
17 Andean countries (Bolivia, Colombia, Ecuador, Peru, Venezuela); Central American countries (Costa
Rica, El Salvador, Guatemala, Honduras, Nicaragua and Panama). The ‘drugs arrangement’ dates from
1 January 1991, when Bolivia, Colombia, Ecuador and Peru received exemptions from quotas as well
as duty free access on certain industrial and agricultural products: Council Regulation 3835/90 [1990]
OJ L370/126. On 1 January 1992, Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and
Panama were granted equivalent access on agricultural (but not industrial) products: Council
Regulation 3900/91 [1991] OJ L368/11. On 1 January 1995, Venezuela was granted equivalent access
on both agricultural and industrial products: Council Regulation 3281/94 [1994] OJ L348/1 and
4
are also now joined by Sri Lanka, Moldova, Georgia and Mongolia,18 making fifteen GSP+
beneficiaries in total. Of these four countries, the first two were existing beneficiaries of the
EU’s labour arrangements, while the last two were in the process of applying for these
preferences.19 This means that, with the exception of Pakistan and with the addition of two
applicants, the beneficiaries of the EU’s new GSP+ arrangement are the same as the
beneficiaries of its former drugs and labour arrangements.20
II. WTO rules
The WTO rules applicable to GSP programs are set out in the 1979 Enabling Clause21 (now
part of the GATT 1994).22 The Enabling Clause states that the most-favoured-nation
obligation in Article I GATT is not applicable to the grant of preferences by developed WTO
Members to developing countries in accordance with the Generalized System of Preferences
as described in the 1971 GSP Decision of the GATT Contracting Parties.23 As a result,
preferences under the Enabling Clause must be ‘generalized, non-reciprocal and nondiscriminatory’.
24
In EC – Tariff Preferences, India argued that the reference to ‘non-discriminatory’
preferences in the Enabling Clause meant that no difference in treatment between developing
countries was permitted.25 The Appellate Body rejected this argument, stating that ‘non-
Regulation 3282/94 [1994] OJ L348/57. From 1 July 1999, all of these countries received duty free
access on most industrial and agricultural products: Council Regulation 2820/98 [1998] OJ L357/1. On
14 November 2001, the European Commission proposed adding Pakistan to the list of ‘drugs
arrangement’, and this was effected in the new GSP Program as of 1 January 2002: Council Regulation
2501/2001, above at n 3. The EU’s rationale was that as a result of the events of 11 September 2001,
including increased refugee flows, Pakistan’s campaign against illicit drugs needed to be supported,
and that ‘[t]he GSP drug regime is therefore likely to stabilise its economic and social structures and
thus consolidate the institutions that uphold the rule of law’. See the Explanatory Memorandum to the
Amended Proposal for a Council Regulation applying a scheme of generalised tariff preferences for the
period 1 January 2002 to 31 December 2004, COM (2001) 688, 14.11.2001. It was the addition of
Pakistan to the list of ‘drugs arrangement’ beneficiaries that sparked India’s complaint in EC – Tariff
Preferences.
18 Commission Decision 924/2005 of 21 December 2005 [2005] OJ L337/50. According to Art 11(3),
this Decision was supposed to have been published by 15 December 2005.
19 Commission Regulation 1649/2000 (Moldova) [2000] OJ L189/13, Commission Regulation
2342/2003 (Sri Lanka) [2003] OJ L346/34; Notice regarding request (Mongolia) [2004] OJ C243/2;
Notice regarding request (Georgia) [2001] OJ 2001 C127/13.
20 This is not however to say that these countries were automatically granted the benefits. In terms of
the mandatory List A conventions, of the fifteen beneficiaries, only five (Costa Rica, Ecuador, Panama,
Peru and Sri Lanka) had ratified all of the required human rights conventions before the arrangement
was announced. Eight beneficiaries (Bolivia, Colombia, Georgia, Guatemala, Honduras, Moldova,
Nicaragua and Venezuela) needed to ratify one outstanding convention, and two (Mongolia and El
Salvador) needed to ratify two. I am extremely grateful to Gustavo Ribiero for sharing his data on this.
21 Differential and More Favourable Treatment, Reciprocity and Fuller Participation of Developing
Countries (‘Enabling Clause’), Decision of 28 November 1979, GATT Doc L/4903.
22 WTO Appellate Body Report, EC – Tariff Preferences, para 90.
23 Generalized System of Preferences, Decision of 25 June 1971, GATT Doc L/3545.
24 WTO Appellate Body Report, EC – Tariff Preferences, paras 145-147.
25 The case is discussed in Steve Charnovitz et al, ‘The Appellate Body’s GSP Decision (Internet
Roundtable)’ 3 World Trade Review 239 (2004), Lorand Bartels, ‘The WTO Ruling on EC - Tariff
Preferences to Developing Countries and its Implications for Conditionality in GSP Programmes’,
Gregory Shaffer and Yvonne Apea, ‘GSP Programmes and Their Historical-Political-Institutional
Context’ and Jane Bradley, ‘The Enabling Clause and Applied Rules of Interpretation’, all published in
Thomas Cottier et al (eds), Human Rights and International Trade (Oxford: OUP, 2005), and Gene
Grossman and Alan Sykes, ‘A Preference for Development: the Law and Economics of GSP’ 4 World
Trade Review 41 (2005).
5
discriminatory’ does not require identical treatment of all developing countries; rather,
additional preferences may be made available to developing countries that share the same
‘development, financial or trade need’.26 Drawing on other provisions of the Enabling Clause,
the Appellate Body also set out further conditions for any such differential treatment: first, the
identified ‘development, financial and trade need’ must meet an ‘objective standard’. The
Appellate Body added that ‘[b]road-based recognition [as] set out in the WTO Agreement or
in multilateral instruments adopted by international organizations, could serve as such a
standard’.27 Second, the identified need ‘must by its nature, be such that it can be effectively
addressed through tariff preferences’. Third, on the facts of the case, ‘a sufficient nexus
should exist between, on the one hand, the preferential treatment provided … and, on the
other, hand, the likelihood of alleviating the relevant “development, financial [or] trade
need”.’28
As it happened, it was unnecessary for the Appellate Body to determine whether the drugs
arrangement at issue met these conditions, because this arrangement was operated through a
‘closed list’. On this, the Appellate Body said as follows:
We find that the measure at issue fails to meet this requirement for the following
reasons. First, as the European Communities itself acknowledges, according benefits
under the Drug Arrangements to countries other than the 12 identified beneficiaries
would require an amendment to the Regulation … Secondly, the Regulation contains
no criteria or standards to provide a basis for distinguishing beneficiaries under the
Drug Arrangements from other GSP beneficiaries.29
Whether or not, aside from this administrative aspect, the drugs regime would have met the
other conditions set by the Appellate Body is debatable. Even if the tariff preferences under
the drugs arrangement could, in principle, have been able to respond to the relevant
‘development needs’ of the beneficiary countries by encouraging substitute crops, it seems
that in fact they did not. As the EU’s European Economic and Social Committee observed,
‘there is no evidence that the special incentive arrangements for combating the production and
trafficking of drugs … has had any impact whatsoever on the drug trade.’30
III. Application of WTO rules to GSP+ arrangement
The argument in this article is that the GSP+ arrangement violates the conditions in the
Enabling Clause, as interpreted by the Appellate Body. This is for the following reasons.
26 WTO Appellate Body Report, EC – Tariff Preferences, para 165.
27 Ibid, para 163.
28 Ibid, para 164.
29 Ibid, paras 187-8.
30 Opinion of the European Economic and Social Committee on the ‘generalised system of preferences
(GSP)’, adopted 25 February 2004 [2004] OJ C110/34, para 6.6.2. But overall benefits do apparently
accrue to GSP+ beneficiaries: see Maria Persson, ‘Everything but Drugs: Export Effects of the
European Union Trade Preferences for Drug Producing and Least Developed Countries’, Working
Paper, Swedish Network for European Studies in Economics and Business, undated, available at
www.snee.org/filer/papers/251.pdf (visited 25 July 2007). By contrast, the benefits of general GSP
arrangements are far less certain: Persson, ibid, Grossman and Sykes, above at n 25; and Çaglar Özden
and Eric Reinhardt, ‘The Perversity of Preferences: GSP and Developing Country Trade Policies,
1976–2000’ 78 Journal of Development Economics 1 (2005).
6
A. The conventions
1. ‘Development, financial and trade needs’
In Recital 7 to the GSP Regulation, the EU gives the following rationale for its GSP+
preferences:
The special incentive arrangement for sustainable development and good governance
is based on an integral concept of sustainable development as recognized by
international conventions and instruments such as the UN Declaration on the Right to
Development of 1986, the Rio Declaration on Environment and Development of
1992, the ILO Declaration on Fundamental Principles and Rights at Work of 1998,
the UN Millennium Declaration of 2000 and the Johannesburg Declaration on
Sustainable Development of 2002.
Consequently, developing countries which due to a lack of diversification and
insufficient integration into the international trading system are vulnerable while
assuming special burdens and responsibilities due to the ratification and effective
implementation of core international conventions on human and labour rights,
environmental protection and good governance should benefit from additional tariff
preferences. These preferences are designed to promote further economic growth and
thereby to respond positively to the need for sustainable development. … (emphasis
added)
In summary, the EU’s GSP+ preferences are designed to compensate countries for ‘assuming
special burdens and responsibilities due to the ratification and effective implementation of
core international conventions’ by ‘promot[ing] further economic growth and thereby …
respond[ing] positively to the need for sustainable development’. This is more easily
understood when turned around: additional preferences will lead to ‘economic growth’ for
certain countries, this being compensation for the expense incurred by those countries when
(voluntarily) they seek to improve their ‘need for sustainable development’ by ratifying and
implementing the listed conventions. It is not entirely easy to identify the ‘development,
financial or trade need’ in this rationale. On the one hand, it seems to be considered
equivalent to the ‘need for sustainable development’, while on the other it seems more limited
to the expense involved in seeking to improve this ‘need’. Nonetheless, it is not necessary to
attempt to parse these possibilities, as the EU’s rationale is in any case of secondary
importance in determining the legality of the EU’s GSP+ arrangement.
The proper approach to this question is to begin with first principles: to interpret the meaning
of the treaty term ‘development need’31 in its legally relevant context.32 This, as US – Shrimp
made clear, includes the acknowledgement, in the Preamble to the WTO Agreement, of the
objective of ‘sustainable development’,33 and, as the Appellate Body did in that case, it seems
appropriate to look to other legal instruments to gain an understanding of the current meaning
of the concept of ‘development need’. A good starting point would seem to be the 1986 UN
Declaration on the Right to Development, which stated that ‘development is a comprehensive
economic, social, cultural and political process, which aims at the constant improvement of
the well-being of the entire population and of all individuals on the basis of their active, free
and meaningful participation in development and in the fair distribution of benefits resulting
31 It is not necessary to discuss ‘trade’ or ‘financial’ needs.
32 Art 31(3)(c) of the Vienna Convention on the Law of Treaties. The context includes the term
‘sustainable development’, found in the preamble of the WTO Agreement, and somewhat oddly
omitted from the list of instruments in Recital 7.
33 WTO Appellate Body Report, US – Shrimp, WT/DS58/AB/R, adopted 6 November 1998, paras 129-
30. It is a matter of some curiosity that Recital 7 of the GSP Regulation omitted to mention this, while
listing a variety of other instruments relevant to the concept of sustainable development.
7
therefrom’.34 To this an intergenerational dimension was added by the 1987 Brundtland
Report, which coined the term sustainable development as ‘development that meets the needs
of the present without compromising the ability of future generations to meet their own
needs’.35 In recent years, the concepts of development, and sustainable development, have
come to be seen ever more holistically. The 2002 UN Johannesburg Declaration on
Sustainable Development referred to the ‘interdependent and mutually reinforcing pillars of
sustainable development – economic development, social development and environmental
protection’.36 And, strongly influenced by the work of Amartya Sen,37 the Human
Development Index (and the annual Human Development Reports) prepared by the UN
Development Programme (UNDP), defines human development as ‘expanding the choices
people have to lead lives that they value’.38
Against this background, it seems inappropriate to understand the term ‘development needs’
solely in economic terms.39 Accordingly, one may suggest that perhaps the Appellate Body
misspoke when it said that a ‘development need’, for purposes of the Enabling Clause, ‘must
by its nature, be such that it can be effectively addressed through tariff preferences’.40 There is
in fact no reason why a ‘development need’ for these purposes needs to be any narrower than
usually understood. On the other hand, this is very far from saying that tariff preferences will
necessarily be able to meet ‘development needs’ without any resource implications. In this
sense, there is certainly an economic dimension to the preferences that are protected under the
Enabling Clause. The point, however, is that this is ensured by focusing on the link between
the preferences and the stated ‘development need’. Thus, it would have been better to say – as
indeed the Appellate Body said in the immediately preceding sentence in EC – Tariff
Preferences – that it is only those tariff preferences that can effectively address legitimate
development needs that have the protection of the Enabling Clause.41
How, then, does this apply to the conventions listed in the EU’s GSP+ program? In general, it
can be said that the objectives of the listed conventions correspond to ‘development needs’,
interpreted broadly, in the sense that achieving these objectives usually corresponds to a
greater standard of development in those countries.42 But to require ratification and
implementation of these conventions as a basis for granting additional tariff preferences risks
not granting preferences to countries which do have a relevant need, and granting preferences
to countries that do not have a relevant need.
34 UN Declaration on the Right to Development, Adopted by General Assembly resolution 41/128 of 4
December 1986.
35 World Commission on Environment and Development, Our Common Future (Oxford: Oxford
University Press, 1987).
36 Johannesburg Declaration on Sustainable Development, 4 September 2002, para 5, available at
www.un.org/esa/sustdev/documents/WSSD_POI_PD/English/POI_PD.htm (visited 25 July 2007). See
also the very broad definition in set out in the ILA New Delhi Declaration of Principles of International
Law Relating to Sustainable Development, available as UN Doc. A/57/329 (2002).
37 See especially Amartya Sen, Development as Freedom (Anchor, 2000).
38 ‘What is HD’, http://hdr.undp.org/hd/ (visited 25 July 2007).
39 Sen notes that the average life expectancy of black US residents is lower than residents of certain
developing countries with a lower per capita GDP: Sen, above at n 37, at 21-3. In this sense, the United
States has a ‘development need’ similar to, if not greater than, some developing countries, though as a
non-developing country it remains excluded from GSP benefits under the Enabling Clause.
40 WTO Appellate Body Report, EC – Tariff Preferences, para 164.
41 Ibid.
42 One might be more sceptical of certain extraterritorial aspects of some of the conventions, such as
the requirement in the Genocide and Apartheid Conventions to prosecute these crimes even if they are
committed outside of the jurisdiction of the ratifying country. It is difficult, if not impossible, to see
how this requirement corresponds to a ‘development need’ of the beneficiary country, even if it
corresponds to a ‘development need’ of the country in which the crime took place.
8
(a) Objective standard
According to the Appellate Body, legitimate ‘development, financial or trade needs’ must also
conform to an objective standard. As mentioned, the Appellate Body considered that ‘broadbased
recognition [as] set out in the WTO Agreement or in multilateral instruments adopted
by international organizations, could serve as such a standard’.43 In saying this, the Appellate
Body is likely to have been referring to such instruments as evidence of a standard. This is
wholly different from giving any importance to the formal act of ratifying such conventions
(this being an independent condition for receiving the preferences).44 In fact, a country that
has not ratified a convention may have precisely the same development needs as one that
has.45 Furthermore, at least one developing country WTO Member (Macao), which receives
general EU GSP preferences,46 lacks the international legal personality required to ratify the
conventions, and is therefore ipso facto ineligible for GSP+ preferences For these reasons, the
EU’s GSP+ arrangement must be considered to be in clear violation of the condition set out
by the Appellate Body that all would-be beneficiaries in a similar situation be granted the
same preferences.47
(b) Selective ‘needs’
It is perhaps also appropriate at this point to make some remarks on the EU’s choice of
conventions, and in particular on the ‘core’ human rights conventions in Part A. Somewhat
surprisingly, this list includes the Apartheid Convention, which two-thirds of EU Member
States have not ratified,48 and the Genocide Convention, which one EU Member State (Malta)
has not ratified. On the other hand, it does not include the UN Migration Convention, which
the UN considers a ‘core’ human rights convention,49 perhaps because it has not been ratified
by any EU Member State.
This selectivity in the choice of conventions designed to serve as evidence of ‘development
needs’ is not in itself legally problematic. The Appellate Body never required that all possible
‘development, financial or trade needs’ be identified in a differential tariff. It also
acknowledged that ‘certain development needs may be common to only a certain number of
43 See above at n 27.
44 As the GSP Regulation requires ratification and implementation it is not necessary to consider the
actual results of ratification alone. On this question see Oona Hathaway, ‘Do Human Rights Treaties
Make a Difference?’ 111 Yale Law Journal 1935 (2002).
45 Indeed, logically speaking, it is tempting to suggest that a country that has ratified and properly
implemented a convention is less likely to have the relevant development need than one that has not.
46 Macao also receives GSP preferences from Canada, New Zealand, Norway, Turkey and Switzerland.
Hong Kong, another non-State WTO Member, does not receive GSP benefits from the EU, but does
from Canada and Norway (and the non-WTO Member Russia): see UNCTAD, Generalized System of
Preferences – List of Beneficiaries, UNCTAD/ITCD/TSB/Misc.62/Rev.1, 2005, at 5.
47 Ironically, this very same point has been made by the EC Trade Commissioner, Peter Mandelson,
when criticizing a proposal of an eco-tax on products produced by countries not party to the Kyoto
Protocol. He pointed out that ‘China has ratified Kyoto but has no Kyoto targets because of its
developing country status. The US has not, but states like California have ambitious climate change
policies’. See Larry Elliott, ‘Mandelson Calls for Open Markets in Green Products’, The Guardian, 18
December 2006.
48 Among EU Member States the Apartheid Convention has only been ratified by Bulgaria, Czech
Republic, Estonia, Hungary, Latvia, Poland, Romania, Slovakia and Slovenia: see www.unhchr.ch/
html/menu3/b/treaty8_asp.htm (visited 25 July 2007).
49 See www.ohchr.org/english/law/ (visited 25 July 2007). Some of the previously captured
beneficiaries (Costa Rica, Georgia, Moldova, Mongolia, and Venezuela) have also not ratified this
convention.
9
developing countries’.50 However, this example also underlines how ineffective the Appellate
Body’s test is in preventing ‘development needs’ from being selected illegitimately. The
Appellate Body sought to protect against such a result in two ways. First, it stated that a
legitimate ‘need’ could not simply be ‘based merely on an assertion to that effect by, for
instance, a preference-granting country or a beneficiary country’ but ‘must be assessed
according to an objective standard’.51 But ‘objectivity’ means little in a case of de facto
discrimination. After all, the classic example of such discrimination, the 1904 German tariff
preferences in favour of Switzerland to cows grazing at 800m above sea level, was equally
‘objective’.52 The Appellate Body’s second condition was that a legitimate ‘need’ must be
‘widely-recognized’.53 The problem with this condition is that, as the EU’s GSP+
arrangement shows, it can be manipulated by a requirement that beneficiaries have more than
one ‘widely-recognized’ needs. In the EU’s GSP+ arrangement, beneficiaries must have no
fewer than twenty-seven ‘widely-recognized’ needs – and these all simultaneously. It is
tempting to speculate whether this is in any way connected with the almost perfect identity of
beneficiaries under the EU’s old and new arrangements.
The failure of the Appellate Body to prevent a GSP+ arrangement being targeted at selected
beneficiaries is also somewhat ironic from a historical perspective, given its similarity to the
ill-fated ‘Brasseur Plan’ of the early 1960s, according to which developed countries would
grant selective GSP preferences on the basis of individual negotiations with each beneficiary
country. The author of the proposal offered this explanation: ‘[i]n this way, true nondiscrimination
would be brought about, for each developing country would be free to ask for
the negotiations it thought useful’.54 Aside from doubts as to this understanding of the
principle of non-discrimination, it was also obvious that such a system would enable the EEC
to protect existing preferences to European ex-colonies under the Yaoundé Convention.55 In
the event, the Brasseur Plan was soundly rejected on the grounds that it would be open to
political manipulation, would fragment the bargaining power of developing countries, would
be impossible to negotiate, and would complicate the world trading system, and the present
system of non-discriminatory preferences was established in its place.56 The resurrection of
selective preferences, even if limited to GSP+ preferences, is therefore not just dubious from a
normative perspective, but seems to contradict the very purposes of the original system.
(c) Positive response
A further problem with the EU’s GSP+ arrangement is presented by the Appellate Body’s
requirement that there must be a causal link between the additional tariff preferences granted
in a GSP+ type program and the improvement of a legitimate ‘development, financial or trade
need’. The Appellate Body said that:
50 WTO Appellate Body Report, EC – Tariff Preferences, para 160.
51 Ibid, para 163.
52 Report of the International Law Commission on the Work of its Thirtieth Session, 8 May-28 July
1978, Doc A/33/10, II(2) YBILC 1 (1978), at 31-2. This document contains an interesting gloss on the
story from an (unreferenced) Food and Agriculture Organization document stating that ‘the specialized
tariff may have been technically justified because of the genetic improvement programme which was
carried out in Southern Germany at that time’.
53 WTO Appellate Body Report, EC – Tariff Preferences, paras 164 and 179.
54 Statement by Mr Maurice Brasseur, in Proceedings of the United Nations Conference on Trade and
Development, Policy Statements, Vol II (New York: United Nations, 1964), 108 at 111.
55 For more on this, see Lorand Bartels, ‘The Trade and Development Policy of the European Union’
18 European Journal of International Law (2007) forthcoming.
56 John Huhs, ‘Trade Preferences for Developing Countries: Options for Ordering International
Economic and Political Relations’ 20 Stanford Law Review 1150 (1968), 1175. See also Gardner
Patterson, Discrimination in International Trade: The Policy Issues 1945-1965 (Princeton: Princeton
University Press, 1966), 360-64.
10
… the response of a preference-granting country must be taken with a view to
improving the development, financial or trade situation of a beneficiary country,
based on the particular need at issue. As such, in our view, the expectation that
developed countries will ‘respond positively’ to the ‘needs of developing countries’
suggests that a sufficient nexus should exist between, on the one hand, the
preferential treatment … and, on the other hand, the likelihood of alleviating the
relevant ‘development, financial [or] trade need’.57
Thus, even if the ‘development needs’ are broad and (according to the suggestion made
already) non-economic in nature, but this does not mean that all forms of preferential
treatment will necessarily be protected by the Enabling Clause.
One situation in which preferential treatment would clearly not be permitted is when a
beneficiary country does not, in fact, have the need at issue. There can be no ‘positive
response’ to a non-existent need. This has some clear implications for the EU’s existing GSP+
arrangement, which grants preferences at least partly to compensate beneficiaries for needs
that they do not necessarily have. This applies in particular to the ratification and
implementation of the Genocide and Apartheid Conventions. In a general sense, all
developing countries have an interest in preventing genocide and apartheid; indeed, they have
an immediate and unconditional erga omnes obligation to do so. But this does not mean that
they also have a ‘need’ to prevent genocide or apartheid, in the sense of the Enabling Clause.
The Appellate Body’s references to ‘improving … the situation’ and to ‘alleviating the …
need’ are clearly linked to material situations, most likely with resource implications.
Arguably, there would be some flexibility in the degree of risk that a situation will be present
in any given country. Some countries clearly have a ‘need’ to prevent damage from
foreseeable natural disasters. But this does not mean that every possible risk, no matter how
remote, can be considered a ‘development need’ of any given country. Bolivia is at no risk of
sustaining direct damage from a tsunami; nor, fortunately, does it appear to be at any risk of
suffering apartheid or genocide. This is not to say, of course, that no countries have a relevant
‘need’ to prevent genocide or apartheid, nor that tariff preferences would necessarily be an
inappropriate means of meeting such a need.58 But in the concrete case, it does appear that the
EU is granting GSP+ preferences to compensate countries for ‘development needs’ that they
do not have. Consequently, these preferences are, at least in this respect, not authorized by the
Enabling Clause.
The situation is different where a country does have a legitimate development need. In this
case, the question is one of causation: is the preferential treatment likely to alleviate this
‘need’? Conceivably, there are two different ways in which this might occur. First, one might
imagine a specific tariff preference targeted at products related to the ‘need’ at issue. A good
example of such a preference (ignoring its actual effects) would be the EU’s former drugs
arrangement, which was designed to create incentives for non-drug substitute crops. A second
possibility is that identified in Recital 7 of the GSP Regulation, which is to increase revenue
and possibly general welfare, which can then be used to contribute to the alleviation of the
relevant ‘need’.59 In both cases, evidence that the preferences are not in fact alleviating the
identified ‘need’, and are not likely to do so, would break the necessary causal link. This
could be either because the preferences are not generating any benefits, or because the ‘needs’
are not being addressed. In the first case, it is difficult to imagine why there would be a
complaint, so in this sense the argument resolves itself. Most likely, argument would be
directed to whether the identified ‘needs’ are being improved. Such an argument would be
57 WTO Appellate Body Report, EC – Tariff Preferences, para 164.
58 Bosnia-Herzegovina is an obvious example. I am grateful to Nathalie Bernasconi for this comment.
59 This assumes that the additional preferences do generate increased economic growth. On this, see
Persson, above at n 30.
11
very difficult to mount in the abstract, but it is conceivable that after some time, a failed
GSP+ preferential arrangement might be challengeable on this basis.
(d) Timing
A further difficulty with the EU’s GSP+ arrangement concerns the sequencing of the grant of
preferences and the conditions to be met. Here the EU seems to have created a paradox. If one
accepts the stated rationale for this arrangement, the GSP+ preferences are designed to enable
countries to meet the cost of ratifying and implementing certain conventions. The problem is
that the preferences are also only made available once this has been done. As there is a lead
time before the benefits begin to flow, this means that the GSP+ beneficiaries are required to
bear an immediate cost, which, in theory, will be compensated later.
This has a number of consequences. In the first place, it is difficult to see how the imposition
of an immediate cost on a developing country can be seen as a ‘positive response’ to a
development need, even if that cost is compensated for later by increased economic growth.
More problematically, there is a clear risk of discrimination in favour of would-be
beneficiaries that are able to meet this immediate cost. By implication, this was recognised by
the European Parliament in a 2004 Resolution on the proposed GSP scheme. While generally
welcoming the GSP+ arrangement, the Parliament still ‘[insisted] that the EU should commit
itself to providing development assistance towards capacity building in order to help
developing countries to qualify for “GSP Plus”, failing which the standards set could act as
non-tariff barriers and a large number of countries could miss out on the potential benefits of
the system.’60 This is clearly not equal treatment of developing countries with the same
development needs, as required by the Appellate Body.61 At a minimum, a conditional GSP+
arrangement must allow a sufficient grace period between the grant of the preferences and the
satisfaction of the conditions established.
2. The ‘vulnerability’ criterion
The EU’s GSP+ program is only available to countries meeting a ‘vulnerability’ criterion,
defined in terms of export diversification and proportion of EU imports. Again, this criterion
needs to be identified in terms of an objective ‘development, financial or trade need’ to which
tariff preferences are a ‘positive response’. Here the problem concerns the element of the
definition of ‘vulnerable’ which specifies that a country’s GSP exports may amount to no
more than 1 per cent of total EU imports. This criterion is not defined in terms of the country
at issue, but in terms of EU imports, an entirely independent factor. As such, by definition it
cannot be a relevant criterion for discriminating between developing countries.
3. The ‘closed list’
Perhaps the most surprising aspect of the GSP+ program is the 31 October 2005 application
deadline for GSP+ benefits, meaning that there is no possibility of adding new beneficiaries to
the GSP+ arrangement until 2009, when a revised GSP program comes into effect. Not only
60 European Parliament Resolution on the Commission Communication, Developing countries,
international trade and sustainable development: the function of the Community’s generalised system
of preferences (GSP) for the ten-year period from 2006 to 2015 (COM (2004) 461) [2005] OJ
C166E/65, para 12. Whether development assistance could alleviate discriminatory trade measure is an
interesting question. In this context, however, it is moot, as the Regulation was adopted without any
incorporation of this suggestion.
61 This author has argued elsewhere that where the conditions reflect customary international law,
developing countries might be deemed to be in the same situation as countries meeting these
conditions: see Bartels, above at n 3, 525. Even if this is true, the obligations in the EU’s current GSP+
arrangement do not all reflect customary international law (as noted of the previous regime: ibid, 526).
12
does this contrast with the open procedure applied in the labour and environment
arrangements in the EU’s previous GSP program,62 and completely ignore a recommendation
by the European Parliament to leave this as an open procedure.63 It also directly contradicts
the Appellate Body’s primary objection to the earlier program, when it said that ‘according
benefits under the Drug Arrangements to countries other than the 12 identified beneficiaries
would require an amendment to the Regulation’.64 Given these factors, it is something of a
mystery why the EU adopted this criterion, but whatever the reason, the fact that it did means
that the EU’s GSP+ regime cannot be permitted under the WTO Enabling Clause.
IV. An alternative
The EU’s GSP+ arrangement is open to challenge on the basis that it discriminates between
developing countries on unjustifiable grounds and in an arbitrary manner, in violation of the
principles set out in the Enabling Clause as interpreted by the WTO Appellate Body in EC –
Tariff Preferences. A legitimate GSP+ program may only differentiate between developing
countries on the basis of ‘development, financial or trade needs’ that are objective and able to
be improved by the grant of tariff preferences. These conditions are not met by the EU’s
GSP+ arrangement, which (at least in part) is based on non-economic needs that are not
objectively identified, and cannot be ameliorated by the grant of tariff preferences. In
addition, the fact that the GSP+ arrangement is based on a ‘closed list’ condemns it as a
simple repetition of the problem in that earlier case.
None of this means that the EU is unable to devise a GSP+ arrangement that does, in fact,
comply with the Appellate Body ruling. Most obviously, such an arrangement would have to
be permanently open to new applicants. This is a question of present relevance, as a number
of non-least developed African, Caribbean and Pacific (ACP) countries decide whether it is in
their interests to enter into reciprocal Economic Partnership Agreements (EPAs) with the EU,
following the expiry of the non-reciprocal regime under the Cotonou Agreement at the end of
2007, or whether they should instead seek to apply to the GSP+ arrangements, which provides
broadly equivalent benefits.65 Should the EU deny such applications simply on the formal
basis that the time for applications was October 2005, the threat of a WTO dispute on this
point alone (should there be standing) might prove a useful tactic. Furthermore, in order to
avoid the ‘timing paradox’ discussed above, there should be a significant grace period
between the grant of preferences and the requirement to ratify and implement conventions, in
order to allow the benefits of the preferences to help meet the costs thereby incurred.
The additional difficulties with the EU’s present arrangement could only be remedied by
more radical amendments. One option would be to base a GSP+ program on a broader and
more objective category of economic vulnerability than is currently the case. Alternatively –
though this would mean leaving out many current beneficiaries and would therefore be
politically difficult – would be to focus on recognised UN categories, such as small island
62 See Art 15 (labour) and Art 22 (environment) of Council Regulation 2501/2001, above at n 3.
63 See the European Parliament legislative resolution, above at n 9, Amendment 22. The justification
was that ‘[t]he special incentive arrangement is an essential part of the GSP as an incentive for
sustainable development and good governance. It should therefore not be limited to the countries that
are sufficiently developed when this Regulation enters into force, but should remain an incentive in the
years to come.’
64 WTO Appellate Body Report, EC – Tariff Preferences, para 187.
65 See, eg, Jonathan Adabre, ‘Ghanaian MPs Adopt GSP-Plus as Alternative to EPAs’, 30 April 2007,
Public Agenda, available at http://allafrica.com/stories/printable/200704301176.html. There are certain
differences in terms of product coverage, safeguards clauses, rules of origin, but in practice for many
such countries these differences are not significant in practice.
13
developing states66 or landlocked developing countries.67 Both of these categories figure in
the UN Millennium Development Goals68 and are taken up in the UN Secretary-General’s
2001 ‘Road Map’ on the implementation of these goals.69 Though the ‘Road Map’ does not
make specific reference to trade preferences, it does consider one means of ‘moving forward’
to be ‘[e]nsuring progress towards a special and differential treatment of small island
developing States in the financial and trade-related spheres’.70 Indeed, even the original
European Commission Communication setting out the guidelines for the new GSP regime
recognised this as an appropriate distinction between developing countries. It stated that ‘the
GSP should focus on the countries most in need, such as the LDCs and the most vulnerable
developing countries (small economies, land-locked countries, small island states, and low
income countries) in order to help them play a greater role in international trade’. Having
acknowledged this, however, the Commission’s solution was, somewhat illogically, to
graduate other countries from the normal GSP program, rather than by adding the specially
affected countries to a more favourable GSP+ arrangement.71
None of this is to say that it is necessarily simple to devise a GSP+ program that would
respond appropriately to the particular needs of such countries (though market access
equivalent to that granted to least developed countries would be a useful starting point). In
principle, though, it would seem that such an approach would be more likely than the EU’s
current arrangement to conform to the Appellate Body’s ruling in EC – Tariff Preferences.
Annex
Conventions referred to in Article 9 of the EU GSP Regulation
Core human and labour rights UN/ILO Conventions
1. International Covenant on Civil and Political Rights
2. International Covenant on Economic, Social and Cultural Rights
3. International Convention on the Elimination of All Forms of Racial Discrimination
4. Convention on the Elimination of All Forms of Discrimination Against Women
5. Convention Against Torture and other Cruel, Inhuman or Degrading Treatment or
Punishment
6. Convention on the Rights of the Child
7. Convention on the Prevention and Punishment of the Crime of Genocide
66 For a comprehensive discussion, though without linking this to GSP programs, see Barbara von
Tigerstrom, ‘Small Island Developing States and International Trade: Special Challenges in the Global
Partnership for Development’ 6 Melbourne Journal of International Law 402 (2005).
67 The United Nations Office of the High Representative for the Least Developed Countries,
Landlocked Developing Countries and the Small Island Developing States (UN-OHRLLS) was
established by the UN General Assembly in 2001: see http://unohrlls.expressiondev.com/en/about/
(visited 25 July 2007).
68 UN Millennium Declaration, UN Doc A/RES/55/2, adopted 18 September 2000, paras 17-18.
69 United Nations, Road Map towards the Implementation of the United Nations Millennium
Declaration, Report of the Secretary-General, UN Doc A/56/326, 6 September 2001, Goal 8 Target 14.
70 Ibid, para 155. The Road Map makes no similar recommendation for landlocked developing
countries (paras 156-7). This is not entirely easy to understand, given that the increased transport costs
of these countries, which is mentioned by the report, affects their development, and would seem to be
appropriately addressed by trade preferences: see eg WTO Cancun Ministerial Conference, Paragraph
18 of the Draft Ministerial Text (JOB(03)150/Rev.2), Small Economies, Joint Proposal by Landlocked
Developing Countries, WT/MIN(03)/W/23, 14 September 2003; and Third Party Written Submission
of Paraguay to the Appellate Body in EC – Tariff Preferences, 30 April 2003, para 26, available at
www.acwl.ch (visited 25 July 2007).
71 European Commission, COM (2004) 461, above at n 4, para 6.2
14
8. Convention concerning Minimum Age for Admission to Employment (No 138)
9. Convention concerning the Prohibition and Immediate Action for the Elimination of
the Worst Forms of Child Labour (No 182)
10. Convention concerning the Abolition of Forced Labour (No 105)
11. Convention concerning Forced or Compulsory Labour (No 29)
12. Convention concerning Equal Remuneration of Men and Women Workers for Work
of Equal Value (No 100)
13. Convention concerning Discrimination in Respect of Employment and Occupation
(No 111)
14. Convention concerning Freedom of Association and Protection of the Right to
Organise (No 87)
15. Convention concerning the Application of the Principles of the Right to Organise and
to Bargain Collectively (No 98)
16. International Convention on the Suppression and Punishment of the Crime of
Apartheid.
Conventions related to the environment and governance principles
17. Montreal Protocol on Substances that Deplete the Ozone Layer
18. Basel Convention on the Control of Transboundary Movements of Hazardous Wastes
and Their Disposal
19. Stockholm Convention on Persistent Organic Pollutants
20. Convention on International Trade in Endangered Species of Wild Fauna and Flora
21. Convention on Biological Diversity
22. Cartagena Protocol on Biosafety
23. Kyoto Protocol to the United Nations Framework Convention on Climate Change
24. United Nations Single Convention on Narcotic Drugs (1961)
25. United Nations Convention on Psychotropic Substances (1971)
26. United Nations Convention against Illicit Traffic in Narcotic Drugs and Psychotropic
Substances (1988)
27. United Nations Convention against Corruption (Mexico)