Preferential trade agreements and the WTO

Nadia Rocha, Robert Teh 21 July 2011



Participation in preferential trade agreements (PTAs) has grown rapidly in recent years. In 1990, there were only about 70 PTAs in force. Thereafter, PTA activity accelerated noticeably; by 2010 the number of PTAs in force was close to 300 (see Figure 1). The average WTO member is party to 13 PTAs. PTA activity has transcended regional boundaries and levels of economic development. One half of the PTAs currently in force are not strictly "regional" with the advent of cross-regional PTAs being particularly pronounced in the last decade. Two thirds of all PTAs in force are between developing countries, about a quarter are between developed and developing countries, and the remainder between developed countries only.

Figure 1. Cumulative number of PTAs in force, 1950-2010, notified and non-notified PTAs, by country group

Source: WTO Secretariat

How preferential is trade?

This year's World Trade Report (2011) shows that this explosion of PTAs is not being matched by an expansion of preferential trade flows. Based on data from the 20 largest importers, who account for 90% of world merchandise trade in 2008, only 16% of trade qualified for preferences (see Carpenter and Lendle 2010). In other words, despite the proliferation of PTAs, 84% of world merchandise trade still takes place on a non-discriminatory most-favoured nation (MFN) basis.

This result should not be surprising:

  • First, there has been a huge reduction in MFN tariffs since the end of the Second World War. Half of world trade is already subject to zero MFN tariff rates.
  • Second, PTAs tend to exempt high MFN-tariff items from preferential treatment. In a recent study involving four major trading countries and their partners, Damuri (2009) shows that about 7% of tariff lines, mainly agricultural or food items and labour-intensive manufactured products, are excluded.
  • Third, rules of origin have contributed to these low figures by making the costs of compliance higher than the perceived worth of the underlying preference margins.
  • Fourth, preferential margins are small. The report shows that only 2% of global imports are eligible for preferential tariffs where the preference margins are above 10%. For most large exporters, preferential tariffs matter little for the bulk of their exports. This is not always true for individual sectors, especially in certain smaller economies exporting a narrow set of commodities (mainly sugar, rice, bananas, fish and garments) where preference margins may be more substantial (Keck and Lendle 2011). There is a possibility though that these preferences will be eroded over time as the countries to which they export enter into more PTAs.

Related to this point, the proliferation of PTAs means that the difference between the MFN rate and the PTA rate overstates the competitive advantage of a PTA member, since increasingly its competitors will also enjoy preferential access to the market. Once the preferential access enjoyed by other exporters is taken into account, less than 13% of preferential trade benefits from a competitive advantage exceeding 2 percentage points. The implication of these results is that one has to look beyond tariffs to explain why countries enter into PTAs.

Deep integration

The coverage of policy areas in PTAs, particularly those of a regulatory nature, has been widening and deepening over time. Using a sample of almost 100 PTAs, the report classifies the provisions in PTAs into WTO+ areas and WTO-X areas.1 WTO+ areas refer to policies covered by the WTO, and WTO-X refers to policy areas not covered in WTO agreements. The report also distinguishes whether these measures were legally enforceable or not under the dispute-settlement mechanism of the PTA. As expected, WTO+ provisions universally include industrial and agricultural tariffs (Figure 2). An increasingly large number of PTAs now also include provisions on technical barriers to trade, services, intellectual property, and trade-related investment measures. WTO-X provisions commonly include competition policy, investment, and the movement of capital (Figure 3). About one third of the PTAs in the sample also include environmental laws, labour-market regulations, and measures on visa and asylum.

Figure 2. Number of agreements with WTO+ provisions

Source: WTO Secretariat

Figure 3. Number of agreements with WTO-X provisions

Source: WTO Secretariat

Why is deep integration gathering momentum? First, trade openness increases policy inter-dependency (spillovers) that makes unilateral decision-making inefficient compared with decisions taken collectively. A second reason is that deep integration agreements may be necessary to promote trade in certain sectors and economic integration more broadly. This second explanation applies to international production networks which require a governance structure beyond low tariffs.

Twenty-first century trade, as defined by Baldwin (2010), is a much more complex phenomenon than trade prior to the early 1980s. The rise in international production networks illustrates the complementarity between trade and governance which is at the core of successful deep agreements. For cross-border production networks to operate smoothly, certain national policies need to be harmonised or rendered mutually compatible to facilitate business activities in several countries (see Lawrence 1996). Empirical evidence2 and case studies presented in the report support this link between production networks and deep PTAs.

Deep integration may involve a trade-off between the benefits of common policies and the costs of harmonisation when policy preferences differ among member countries. Deep integration lowers trade costs and provides shared benefits, such as common rules and a stable monetary system that the market or national governments fail to offer. Deep integration with advanced economies may also create advantages for developing countries from importing best-practice institutions. However, costs may be involved if the common rules are distant from national preferences and the needs of developing countries.

Coherence between PTAs and the WTO

Deep PTAs present a different challenge to the multilateral trading system, allowing for the deeper integration sought by members of a PTA while maintaining compatibility with the non-discrimination principle of the WTO. The traditional building-bloc stumbling-bloc paradigm does not provide the appropriate analytical framework for deep PTAs since the underlying question behind this approach is whether preferential tariff opening would eventually lead to multilateral opening.

The existing literature suggests that deep integration is often non-discriminatory. By their very nature, some deep integration provisions are de facto extended to non-members because they are embedded in broader regulatory frameworks that apply to all trading partners. Preferential trade agreements may also directly refer to WTO rules on deep integration measures, automatically supporting the multilateral trading system. However, some deep provisions in preferential trade agreements can contain discriminatory aspects, creating a tension with the multilateral trading system.

There are a number of options for increasing coherence between PTAs and the multilateral trading system.

  • One option would be for members of PTAs to extend their existing preferential arrangements in a non-discriminatory manner to additional parties.
  • A second option would be to fix deficiencies in the WTO's legal framework on PTAs by clarifying and improving existing disciplines.
  • A third option would be to adopt a more nuanced and non-litigious approach to PTAs in the context of the existing transparency mechanism. This would allow WTO members to better understand one another's interests and to make progress in the legal interpretation of WTO provisions on PTAs.
  • Lastly, WTO members can accelerate their efforts at multilateral trade opening. An immediate step in this direction will be for WTO members to bring to a successful conclusion the long-standing negotiations under the Doha Round.

These are not mutually exclusive options and it is likely that all these options would have to be pursued.

The opinions expressed here should be attributed to the authors only. They are not meant to represent the positions or opinions of the WTO or its members and are without prejudice to members' rights and obligations under the WTO.


Baldwin, R (2010), “21st Century Regionalism: Filling the gap between 21st century trade and 20th century trade rules”, CEPR Policy Insight No. 56.

Carpenter, T and A Lendle (2010), “How Preferential is World Trade?”, CTEI Working Paper 2010-32, Geneva: The Graduate Institute Centre for Trade and Economic Integration.

Damuri, YR (2009), “How preferential are preferential trade agreements? Analysis of product exclusions in PTAs”, Swiss National Centre of Competence in Research, Working Paper 2009/30.

Horn, H, P Mavroidis, and A Sapir (2010), "Beyond the WTO? An anatomy of EU and US preferential trade agreements”, The World Economy, 33(11):1565-1588.

Keck, A and A Lendle (2011), “Determinants of preference utilization in the EU and US”, mimeo.

Lawrence, RZ (1996), “Regionalism, Multilateralism and Deeper integration”, Brookings Institution.

Orefice, G and N Rocha (2011), “Deep integration and production networks: an empirical analysis”, WTO Working Paper 2011-11.

World Trade Organization (2011) World Trade Report 2011: From Co-existence to Coherence (Geneva: WTO).

1 This follows the method outlined in Horn et al (2010).

2 For a complete analysis on the relationship between deep integration and production networks see Orefice and Rocha (2011).



Topics:  Global governance International trade

Tags:  WTO, Doha Round, preferential trade agreements, regionalism

Senior Economist, Trade and Regional Integration Unit, World Bank

Counsellor in the Economic Research and Statistics Division, WTO


CEPR Policy Research