Contagious FTAs: New evidence on the domino theory of regionalism

Dany Jaimovich, Richard Baldwin

02 September 2010



Regionalism is sweeping the world trade system like wildfire. Multilateral WTO talks, meanwhile, proceed at a glacial pace. This correlation has led many observers to fear that regionalism’s boon is multilateralism’s bane. Jadish Bhagwati’s last book, for example, is titled Termites in the Trading System: How Preferential Agreements Undermine Free Trade (Bhagwati 2008).

This fear has pushed regionalism far up the global economic agenda and prompted a new research. But a focus on regionalism is nothing new. The profession’s best and brightest were focused on regionalism in the 1940s and 1950s – Jacob Viner, James Meade, Richard Lipsey, and Harry Johnson inter alia. The main issue then was what could be called “small-think regionalism”, i.e. are free trade agreements beneficial to the signers?

A-list economists turned back to the issue in the early 1990s, which was the last time multilateral agreement proved difficult but regional agreements proved easy. In 1991, Paul Krugman, Larry Summers, and Jagdish Bhagwati laid out what could be called “big-think regionalism”, i.e. the systemic implications. Two principle lines of inquiry were launched.

  • Does spreading regionalism harm world welfare?

Krugman (1991a) sketched a model that crystallised the profession’s thinking around this “Is bilateralism bad?” question.

  • Does regionalism help or hinder multilateralism?

Krugman (1991b, 1993) sketched a bargaining model where regionalism can help or hinder multilateralism, but it was Jagdish Bhagwati’s bon mot that organised the profession’s thinking. Do trade blocs “more readily serve as building blocks of, rather than stumbling blocks to, GATT-wide free trade”? (Bhagwati 1991 p. 77). Specifically, the second line of inquiry crystallised around the question of whether an exogenous variation in regional trade blocs made multilateral tariff cooperation more or less likely.

Larry Summers’ contribution was to come down firmly on one extreme of the debate. He declared that all the ‘isms are good: unilateralism, bilateralism, plurilateralism and multilateralism (Summers 1991).

More recently, a third line of inquiry has emerged with a focus on the causes of regionalism (Baldwin 1993, 1995, 1997, Grossman and Helpman 1995, Yi 1996, Freund 2000a, b, Aghion et al. 2004).

Theories of spreading regionalism

Since only two multilateral GATT Rounds have been concluded in the last 40 years, testing the first two lines of inquiry proved difficult. Instead, the empirical literature has focused on the third line of thinking. The most common theoretical explanations are:

  • “Slow multilateralism”, i.e. the assertion that regionalism is spreading because multilateral talks are progressing so slowly (Krugman 1991, Bhagwati 1993, 2008).
  • Idiosyncratic shocks, such as the US’s opening of the US-Canada Free Trade Agreement (FTA) talks in 1986 (Bhagwati 1991 p.71), the breakup of the USSR in 1991 (Lester and Mercurio 2009 p.3), and the Asian Crisis of 1997 (Harvie et al. 2006 p.3).
  • The global spread of democracy (Mansfield et al. 2002, Wu 2004), and the quest for geopolitical stability (Mansfield and Pevehouse 2000, Martin et al. 2008, 2010, and Vicard 2008).
  • The bandwagon effect, which argues that nations sign FTA since they see other nations signing them (Bhagwati 1991 p.73, and Solis et al. 2009).
  • The domino theory of regionalism produces what looks like a bandwagon effect but identifies the precise political economy links producing it (Baldwin 1993).

The domino theory posits that signing one FTA induces outside nations to sign new FTAs that they previously shunned since the trade diversion effect of the first FTA creates new political economy forces in excluded nations. Specifically, excluded nations seek to sign FTAs as a means of redressing the new discrimination. The second-round FTAs in turn create their own trade diversion that may in turn lead to more FTAs. Using an obvious metaphor, the domino theory explains why FTAs seem to be “contagious”.

New evidence on the domino effect

In a recent CEPR Discussion Paper (Baldwin and Jaimovich 2010), we present empirical evidence that domino effects are important. The value added is threefold.

  • First, we extend the domino theory model to allow for FTAs (the original contributions concerned customs unions).
  • Second, we use the model to develop a theory-based measure of contagion. Importantly, our measure of contagion captures asymmetries in the country pairs, while Egger and Larch (2008) implicitly assume contagion is bilaterally symmetric (since their measure depends on distance).
  • Third, we test the domino effect against alternative hypotheses using a broader sample of FTAs than previous studies.

The key empirical lever that allows us to distinguish the contagion is the extent to which trade ties connect the new FTA signers. The contagion hypothesis works on trade diversion, so the spread of FTAs should follow a pattern that is clearly related to the new signers’ trade patterns. In particular, a pair of nations should be more likely to sign a new FTA if either of them has recently signed FTAs with third nations that are the pair’s exporting rivals.

Measuring contagion effects

According to our theory, the contagion comes from the threat of trade diversion. This suggests that two elements must be considered: the number of FTAs signed among a nation’s trade partners, and the importance of those markets to the nation’s exporters. To operationalise this, we follow the spatial econometrics literature and represent the structure of the spatial interdependence in the contagion index as an N*N weighting matrix, where each element is a measure of the “distance” between two nations. Previous studies that use spatial econometrics to analyse FTA interdependence, Egger and Larch (2008) and Manger (2006), use a symmetric weight matrix where the elements are the inverse of the average geographical distance. We instead use the “economic distance”, which allows, for example, for Chile to care more about the US’s FTA with Mexico than the US does about Chile’s FTA with Brazil.

Data and results

Our data on FTAs comes from the recently released Preferential Trade Agreements Database developed by Hufbauer and Schott (2009). This source is the most extensive and detailed currently available (e.g. it is more comprehensive than the WTO database). The Hufbauer-Schott database has 570 agreements recorded from 1948 to 2007; of these, 329 agreements were still force in 2007. We limit the period of investigation to 2005 due to lack of comprehensive data on our “right-hand side” variables. In the original database, a cumulative total of 1319 country-pair trade deals are registered up to 2005, but just 1134 are still in force or signed for later implementation (most of the dropped FTAs are due to expansion of customs unions, especially the EU, which force new members to drop old bilateral FTAs).
In our regression results, we find that our contagion index does a good job of explaining which FTAs are signed. The other explanations usually work when they are entered alone, but their significance drops off or disappears when the contagion index is included.


The domino theory is not a “primitive” explanation of why regionalism is spreading, but it explains how a few exogenous shocks could produce a cascade of political economy effects that makes it look like regionalism is spreading like a wildfire. On the whole, this makes recent regionalism look more benign. Nevertheless, regionalism does pose a threat to the world trading system.

Despite what one reads in the newspaper, trade liberalisation is as popular as ever among policymakers. The new century has seen massive liberalisation of trade in goods and services – much of it by nations that disparaged trade liberalisation for decades. But unlike last century, almost none of this has occurred multilaterally. Developing nations have cut their tariffs unilaterally and via South-South FTAs. Rich nations have relied on regional trade with developing nations to achieve their market-opening goals.

To date, these changes seem more like challenges than threats. The key players believe the world trade system will continue to be anchored by the WTO’s shared values, such as reciprocity, transparency, non-discrimination, and the rule of law. WTO-anchorage allows each member to view their own FTAs as minor derogations. Yet, at some point there will be so many FTAs that WTO rules become the exception and the derogations become the new norm – expectations about WTO’s central role will come unmoored. Nations feel justified in ignoring WTO norms since everyone else does. This is something world leaders should perhaps think harder about.


Aghion, P, P Antràs and E Helpman (2004), “Negotiating Free Trade”, NBER Working Paper 10721.

Baldwin, Richard (1993), “A domino theory of regionalism”, CEPR Discussion Paper 857; NBER Working Paper, 4465.

Baldwin, Richard (1995), “A domino theory of regionalism”, Chapter 2 (pp. 25-48) in Baldwin, Haaparanta and Kiander (eds.), Expanding Membership in the European Union, Cambridge University Press.

Baldwin, Richard (1997), “The causes of regionalism”, The World Economy, 20(7): 865-888(24).

Baldwin, Richard and Dany Jaimovich (2010), “Are Free Trade Agreements Contagious?” CEPR Discussion Paper 7904.

Bhagwati, Jagdish (1991), The World Trading System at Risk, Princeton University Press.

Bhagwati, Jagdish (2008), Termites in the Trading System: How Preferential Agreements Undermine Free Trade, Oxford University Press.

Freund, C (2000b), “Multilateralism and the endogenous formation of preferential trade agreements”, Journal of International Economics, 52:359–376.

Grossman, G. and E. Helpman (1995), “The Politics of Free-Trade Agreements”, American Economic Review, 85(1995b):667-690.

Hufbauer, G and J Schott (2009). “Fitting Asia-Pacific agreements into the WTO system,” in R Baldwin and P Low (eds.), Multilateralising Regionalism: Challenges for the global trading system, Cambridge University Press.

Summers, L (1991), "Regionalism and the World Trading System", Economic Review, Federal Reserve Bank of Kansas City, issue Nov: 295–302.

Yi, S (1996), “Endogenous formation of customs unions under imperfect competition: Open regionalism is good”, Journal of International Economics, 41:151-175.




Topics:  International trade

Tags:  WTO, regionalism, multilateralism

Postdoctoral Researcher, Department of Applied Econometrics and International Economic Policy, Goethe University Frankfurt

Professor of International Economics, Graduate Institute, Geneva; President of CEPR; Vox Editor-in-Chief