How WTO loopholes for the formation of regional trade areas led to the fragmentation of the multilateral trade order

How GATT and WTO loopholes for the formation of regional trade areas led to the fragmentation of the multilateral trade order: Lessons for a future world trade system

International trade has grown rapidly since the establishment of a multilateral trade facilitation regime after World War II, first in the form of the General Agreement on Tariffs and Trade (GATT)[1] and, since 1995, through the World Trade Organisation (WTO).[2] Now in its 26th year, the WTO remains the only global framework for trade regulation and negotiation. It currently comprises 164 members.[3]

While WTO membership has grown over the years, so have the WTO’s problems. The recent proliferation of separate trade agreements, outside the realm of the WTO, has arguably been the main transformation of the multilateral trading system since the establishment of the WTO.[4] Neither the WTO, nor the GATT before it, prevent members from entering into separate agreements with other countries. Such agreements appear in various guises, including customs unions (CUs), free trade agreements (FTAs) and preferential trade agreements (PTAs).

Agreements outside of the WTO regulatory framework – which we can collectively refer to as regional trade agreements (RTAs)[5] – have a commonality in that they all deviate from the GATT’s and WTO’s foundational principle of Most Favoured Nation (MFN).[6] According to the MFN principle, WTO members must extend all advantages, favours, privileges and immunities which they accord to any one country, immediately and unconditionally to all other WTO members. It is immaterial whether the country enjoying the advantage or benefit is a WTO member or not. This ensures that WTO members can not extend privileges to non-members without also extending them to other members. It also ensures that, where a member enters into a separate trade agreement with another member, the benefits accruing under the separate agreement extend to all other WTO members.

By design, MFN negates any benefits that might arise from entering into separate agreements with selected countries only. One of the main rationales behind the MFN principle in the multilateral sphere is that it allows all countries to participate in international trade pursuant to the same rules, which also prevents stronger trading partners from dictating trading terms to weaker trading partners.[7] Such practices were wide-spread in the pre-World War II era and led to unequal trade dependencies.[8] MFN also pre-empts stronger trading partners from making trade benefits conditional on non-economic ends, such as military or political collaboration.[9]

The drafters of the GATT were aware of the inherent conflict between the MFN principle and a country’s ability to enter into separate trade agreements. At first, MFN took precedence but the GATT negotiations, which took place between 1946 and 1948, saw a weakening of the MFN in favour of separate trade agreements. It was ultimately agreed to allow deviations from the MFN principle in cases where countries wanted to enter into smaller agreements between individual member countries. The specific provision that regulates such deviations is Article XXIV of the GATT.[10] A similar provision was included in the General Agreement on Trade in Services (GATS)[11] when it came into force at the inception of the WTO in 1995. There is also the so-called Enabling Clause, which was adopted in 1979 to facilitate preferential trade agreements between developed and developing countries.[12]

The legal framework established in Article XXIV (GATT), Article V (GATS) and in the Enabling Clause, not only allows for deviations from MFN but endorses the notion that such deviations are beneficial when used to further economic integration between some member states. The GATT explicitly “recognizes the desirability of increasing freedom of trade by the development, through voluntary agreements, of closer integration between the economies of the countries parties to such agreements.”[13]

For the best part of the GATT era, the preferential trade agreements between members were only rarely entered into. This trend is neatly captured in Table 1 below, whereby the red line represents all RTAs currently in force, totalling 483.[14] At the time of the WTO’s inception in 1995, there were 13.

Description: Macintosh HD:Users:hans:Desktop:RTAs in world trade.png

Table 1: Evolution of RTAs in the World, 1948 to 2020.[15]

One of the earliest examples of Article XXIV of the GATT being invoked was the creation of the European Economic Community, which took the form of a CU.[16] GATT drafters recognized the dangers of MFN deviations and therefore only foresaw an exception for CUs and not other forms of RTAs, such as FTAs.[17] The reason for this was simple. CUs are the most advanced form of economic integration, comprising both free trade within the CU’s borders, as well as common rules and tariffs for trade from outside the CU. From a trade perspective, CUs can be viewed as single entities, meaning that countries outside of a CU deal with the CU as a whole, and not with individual constituent countries. In other words, when individual countries enter into a CU, together they become a akin to a larger country, at least for the purposes of international trade. This is, for instance, reflected in the fact that EU members states are collectively represented at the WTO by the EU. FTAs only affect internal trade between FTA members, who each retain the right to regulate external trade. FTAs therefore represent a far less advanced form of economic integration.

When the GATT was negotiated, even the CU deviation was thought to be exceptional in nature, as captured by Jackson, who later described its inclusion in the Geneva Draft[18] as something that “even the United States recognized”.[19] Thus, while the CU exception found its way into the final draft, it was, for almost the entire negotiation period, regarded as exceptional. There was no discussion of broadening it.

At the Havana conference that led to the adoption of the GATT, a number of countries, including the United States, were against softening up the rules, and preferred a system whereby any deviations – other than for CUs – had to be approved by members based on a two-thirds voting majority.[20] However, other countries insisted on loosening the rules to allow for regional co-operation without having to satisfy the high CU threshold.[21] In the end, by the time GATT negotiations were completed in March 1948, the scope of the RTA exception was expanded beyond CUs to include FTAs, as well as interim agreements. This represented a significant softening of the previously agreed disciplines. Table 2 illustrates the progression of GATT negotiations in relation to MFN deviations. Notably, the disciplines were only weakened at the very final stage.

Draft Section Date MFN deviations
US draft proposal 33 September 1946 CUs only
London draft 38 November 1946 CUs only
New York draft 38 February 1947 CUs only
Geneva draft 42 October 1947 CUs only
Havana Charter 44 March 1948 CUs, FTAs, and interim agreements

Table 2: Timeline of MFN deviations in GATT negotiations.[22]

Notwithstanding the lag between adoption of the rules and the explosion in their use four decades later, the softening of the RTA disciplines have undoubtedly facilitated more recent developments.[23] If RTAs were limited to CUs, they would be far more difficult to set up, and there would be fewer of them. Easier access to loopholes means that they will be used more often. In turn, this creates an institutional threat, as is evident from the increasing redundancy of the WTO. There is also academic commentary that points the blame at poor application of the existing rules.[24] It is likely that a combination of both factors, weak disciplines, as well as weak application, has led to the current situation.

The result of these developments is that the multilateral trade system has been seriously undermined.[25] Smaller and developing countries are particularly affected as RTAs are dominated by the major trading countries.[26] Dominance is reflected both in terms of participation, as well as the substance of the PTAs.[27] This reintroduces the asymmetry which had been present before World War II and which the GATT and WTO were meant to put a stop to. The proliferation of PTAs has also meant that there have been little or no multilateral negotiations for many years. Some might argue that other factors are at play, such as the WTO’s lack of a modern investment regime.[28] While there are undoubtedly many different reasons why countries chose to enter into separate agreements, it is also certain that if countries did not have the option of going separate ways, there would be a much stronger incentives to word under the WTO umbrella, for instance in developing a multilateral investment regime.

Overall, it appears as if the WTO is on a path towards redundancy. While, in 2015, WTO members reaffirmed “the need to ensure that Regional Trade Agreements (RTAs) remain complementary to, not a substitute for, the multilateral trading system[29], there is little evidence that the trend has been halted, let alone reversed, with new and bigger RTAs in the pipeline.[30] The impending economic crisis in the aftermath of the COVID-19 pandemic is another factor that will likely lead to countries opting to shape trade rules to their own benefit, which may lead to additional pressure on weaker trading partners. In a similar vein, calls for a new wave of protectionism are also likely to adversely affect weaker trading countries, who will be left with the choice of trading according to the stronger parties’ rules or being left out altogether. These were the kinds of trends the GATT and WTO were meant to suppress and they need to be countered if the multilateral trading system is to survive.

[1] General Agreement on Tariffs and Trade, Final Act Adopted at the Conclusion of the Second Session of the United Nations Conference of Trade and Employment, 30 October 1947, UN Sales No.1947.II/10, 55 UNTS 187, entered into force 1 January 1948. (This Agreement has since been incorporated into the WTO agreements.)

[2] Agreement Establishing the World Trade Organisation, 1994, Final Act Embodying the Results of the Uruguay Round of Multilateral Trade Negotiations (MTN), 15 December 1993, entered into force 1 January 1995, The Results of the Uruguay Round of Trade Negotiations – The Legal Texts, GATT Secretariat, (1994).

[3] Source: (viewed on July 21, 2020)

[4] See, e.g., F. Bonciu and M. Moldoveanu, The Proliferation of Free Trade Agreements in the Post-Doha Round Period: The Position of the European Union, Procedia Economics and Finance, Vol. 8 (2014), pp. 100-105; M. Matsushita and Y.S. Lee, Proliferation of Free Trade Agreements and Some Systemic Issues – In Relation to the WTO Disciplines and Development Perspectives, The Law and Development Review, Vol. 1, Iss. 1 (December 2008), pp. 23-50; L. Trakman, The Proliferation of Free Trade Agreements: Bane or Beauty, Journal of World Trade, Vol. 42, Iss. 2 (2008), pp. 367-388. 

[5] While not all such agreements are regional and the term can be seen to be misleading, it is adopted here because the term RTA represents the terminology used at the WTO.

[6] Supra note 1, Article I.

[7] R. Senti and H. Mahncke, A Guide to the World Trade Organisation, LexisNexis, (2008), p. 26.

[8] Ibid.

[9] The Generalized System of Preferences (GSP), for which the GATT and WTO’s Enabling Clause is the legal basis, is said to operate is such a manner. See G. Grossman and A. Sykes, A Preference for Development: The Law and Economics of GSP, World Trade Review, Vol. 4, Iss. 1 (2005), pp. 41-67, at p. 55; see infra footnote 12.

[10] Supra note 1, Article XXIV.

[11] General Agreement on Trade in Services, 15 April 1994,1869 UNTS 183,

ILM 1167 (1994), entered into force 1 January 1995.

[12] Decision on Differential and More Favourable Treatment, Reciprocity and Fuller Participation of Developing Countries, Decision of 28 November 1979 (GATT Document L/4903, BISD 26S/203).

[13] Supra note 1, Article XXIV(4).

[14] Source: (accessed on July 21, 2020).

[15] Ibid.

[16] Treaty Establishing the European Economic Community, 25 March 1957, 298 UNTS 3.

[17] K. Dam, The GATT: Law and International Economic Organization, The University of Chicago Press (1970), p. 274.

[18] The Havana Charter, incorporating the GATT, went through a number of drafts. The first draft was put forward by the United States in September 1946 and was called “Suggested Charter for an International Trade Organization of the United Nations”. The revised, second draft, came out of the London meeting of the preparatory committee of the International Conference on Trade and Employment that took place between October and November 1946. It is known as the London draft, 1946. A third draft came out of the New York meeting of the preparatory committee in February 1947. The fourth draft is the Geneva draft and was the result of revisions made to earlier drafts by the same preparatory committee meeting in Geneva, between April and August 1947. The final version is called the Havana Charter, which was the result of the Havana negotiations that took place between November 1947 and March 1948. The Havana Charter version is identical to what became GATT Article XXIV.

[19] Jackson, John H., World Trade Law and the Law of the GATT, Mitchie Co., Virginia (1969), p. 577. (emphasis added)

[20] United Nations Conference on Trade and Development, E/CONF.2/C.3/SR.7, 19 December 1947, pp. 3-4; see, e.g. comment by the United States representative: “economic regional preference arrangements were not a promising device for economic development. Special circumstances justifying such an arrangement should be submitted to thee Organization for its decision as to the net gain to world trade, otherwise the whole object of eliminating preferences would be undermined.”, at p. 4.

[21] United Nations Conference on Trade and Development, E/CONF.2/C.3/SR.4 – SR.8, 19 December 1947.

[22] Supra note 18.

[23] P. Rosendorff and H. Milner, The Optimal Design of International Trade Institutions: Uncertainty and Escape, International Organization, Vol. 55, Iss. 4 (2001), pp. 829–857; K. Chase, Multilateralism compromised: the mysterious origins of GATT Article XXIV, World Trade Review, Vol. 5, Iss. 1 (2006), pp. 1–30.

[24] See, e.g., C. Picker, Regional Trade Agreements v. the WTO: A Proposal for Reform of Article XXIV to Counter this Institutional Threat, U. Pa. J. Int’l Econ. L., Vol. 26. Iss. 2 (2005), pp. 267-319. GATT and WTO cases involving RTA measures include: EEC – Tariff Treatment of Imports of Citrus Products from Certain Countries in the Mediterranean Region, L/5776, 7 February 1986; EEC – Member States’ Import Regimes for Bananas, DS32/R, 3 June 1993; EEC – Member States’ Import Regimes for Bananas, DS38/R, 11 February 1994, Turkey- Restrictions on Imports of Textile and Clothing Products, Report of the Panel, WT/DS34/R, 31 May 1999, Report of the Appellate Body, WT/DS34/AB/R, 19 November 1999; European Communities – Conditions for the Granting of Tariff Preferences to Developing Countries, Report of the Panel, WT/DS246/R, 1 December 2003, Report of the Appellate Body, WT/DS246/AB/R, 20 April 2004.

[25] See, e.g., J. Bhagwati, Termites in the Trading System: How Preferential Agreements Undermine Free Trade, Oxford: Oxford University Press (2008); S. Baier and J. Bergstrand, Economic Determinants of Free Trade Agreements, Journal of International Economics, Vol. 64 (2004), pp. 29-63; N. Limao, Preferential Trade Agreements as Stumbling Blocks for Multilateral Trade Liberalization: Evidence for the US, American Economic Review, Vol. 96 (2006), pp. 896-914.

[26] See, e.g., S. Ellis Barnekow and K. Kulkarni, Why Regionalism? A Look at the Costs and Benefits of Regional Trade Agreements in Africa, Global Business Review, Vol. 18, Iss. 1 (2017),  pp. 99-117; A. Palit, Mega-regional trade agreements and non-participating developing countries: Differential impacts, challenges and policy options, Competition & Change, Vol. 21, Iss. 5 (2017), pp. 417-434.

[27] Supra note 25.

[28] The WTO agreements include the Agreement on Trade-Related Investment Measures (TRIMS), supra note 2, but this is a fairly rudimentary agreement compared to those included in many recent PTAs.

[29] Nairobi Ministerial Declaration (adopted on 19 December 2015) WTO Doc WT/MIN(15)/DEC, para. 28.

[30] The largest of these is the Regional Comprehensive Economic Partnership (RCEP), which is a proposed FTA in Asia which would comprise about half the world’s population.

[31] J. Anderson and Y. Yotov, Terms of trade and global efficiency effects of free trade agreements, 1990–2002, Journal of International Economics, Vol. 99 (2016), pp. 279-298. C. Carrere, Revisiting the effects of regional trade agreements on trade flows with proper specification of the gravity model, European Economic Review, Vol. 50, Iss. 2 (2006), pp. 223–247. T. Kohl, Do we really know that trade agreements increase trade?, Review of World Economics, Vol. 150, Iss. 3 (2014), pp. 443-469.

[32] J. Bhagwati, United States Trade Policy: The Infatuation with FTAs, in J. Bhagwati and A. Krueger, The Dangerous Drift to Preferential Trade Agreements, AEI Press, Washington, DC (1995).

[33] There are surprisingly few cases involving the interpretation of the rules pertaining to MFN deviation. A list of cases dealing with the issue directly is provided in supra 24 above.