Must FTAs Be Vehicles For Discrimination Presentation at

Must FTAs Be Vehicles For Discrimination? Presentation at University of Sussex, 22 May 2008 Simon J. Evenett www. evenett. com

Contents of this presentation 1. 2. 3. 4. 2 Reflecting on the question: Must FTAs be vehicles for discrimination? Relationship to "multilateralising regionalism. " New information on certain classes of FTA non-tariff provisions. Some hard yet-to-be answered questions.

Must FTAs be vehicles for discrimination? 1. 2. Intellectual significance--given Viner's classic analysis. Practical significance today: – For our understanding of what is a "free trade agreement". – For the welfare and policy options available to FTA parties. – For the welfare and policy options available to third parties and "the system. " 3. The answer to the question: NO. So what? – Given historical precedents. No claims about novelty per se. – Logic and Effects? 3

Relationship to "Multilateralising Regionalism" 1. Initial contribution was Baldwin (2006). Work followed by others including Baldwin, Evenett, and Low (2008). 2. From our joint paper: "The notion of multilateralising preferential agreements can be defined with more or less precision. The fundamental idea is obvious—a process of multilateralisation occurs when existing preferential arrangements are extended in a non-discriminatory manner to additional parties. " Hmmm. 3. Examples given in Baldwin (2006): – Information Technology Agreement (ITA) in 1995. – Rules of origin (PECS in 1997). – Baldwin's explanation: Spaghetti bowls and production unbundling. 4

Relationship to "Multilateralising Regionalism" 1. 2. 3. 4. 5 Baldwin, Evenett, and Low (2008) examined whether comparable examples could be found in trade rules that did not relate to what is traditionally referred to as "trade in goods”; some call these rules "behind the border. " Plus BEL (2008) were interested in the generality of the Baldwin (2006) logic. Examined text of FTA provisions from nearly 100 agreements in government procurement, competition law and policy, investment performance, and trade remedies, and drew on secondary literature on services and TBT. What we found surprised us—and has surprised some in the trade policymaking community.

BEL (2008) found FTA provisions where: 1. 2. 3. 4. 5. • 6 The parties agreed to a commitment on a MFN basis. The parties agreed to a commitment that legally-speaking discriminates against third parties—but in fact was implemented on a MFN basis. More parties join an agreement with preferential treatment. The parties agreed to give each other any more favourable treatment that they grant to a third party in a subsequent FTA (so called Third Party MFN). The parties agree not to invoke WTO rights that discriminate or may lead to discrimination against parties to the FTA. Relationship of this taxonomy to Baldwin's ITA and PECS examples.

BEL (2008) also found: 1. Using FTAs to limit non-discrimination is not confined to any one set of FTA provisions. 2. Using FTAs to limit non-discrimination is not confined to any one group of FTA parties. 3. Using FTAs to limit non-discrimination is spreading over time—it is becoming more prevalent. – See the cool charts on the following slides! 4. The Spaghetti Bowl plus production unbundling argument applies in some cases but not all. Other explanations were identified—not least because the Spaghetti Bowl problem isn't always created by every type of FTA provision. 7

De Jure MFN: Spread of NAFTA performance requirements provisions in FTAs. United States Korea Chile Colombia Mexico Singapore Peru Morocco Japan Venezuela Australia NAFTA agreement identical or almost identical language of provisions same NAFTA provisions with some exeptions or omissions Panama Canada Chinese Taipei

Strong Third Party MFN: FTAs with provisions that automatically trigger better treatment in FTA partner procurement markets if that partner subsequently negotiates better coverage with a third party Bosnia-Herzegovina Palestine Authority Croatia Moldova Morocco Albania Macedonia Turkey Slovenia Montenegro Bulgaria Tunisia Kosovo Romania Israel

Weaker Third Party MFN: FTAs with provisions that trigger consultations or negotiations if a party subsequently negotiates greater procurement-related coverage with a third party Lebanon Israel Chile Tunisia EFTA Mexico EC Korea SACU Japan Singapore Australia

Discouraging use of exceptions to WTO principles: FTAs with Joint Committee provisions to oversee antidumping and safeguard measures Romania 1993 - 2007 7. 1. Serbia and Montenegro 2. 19 98 EFTA Morocco Tunisia Turkey 8. . 01 1. 00 99 10. 1. 1. 2000 Bulgaria 4. 1. 05. 1997 1. 12 . 1 997 . 10. 1 6. 30 Bosnia and Herzegovina 5. 1. 7. 1997 EC . 06 Albania Kosovo 00 agreements 1993 – 2000 agreements concluded after 2000 9 99 1. 1 1. 0 EFTA 9. Croatia Moldova . 2000 09 13. 1. 5 99 1. 93 12. 19 2. 31. . 1 01 Algeria Jordan Lebanon Morocco Tunisia 1. 3. 19 93 EFTA FYR Macedonia 19 . 9. 20 11. 1 CEFTA Israel EFTA

Adoption of common/similar rules: Spread of NAFTA Art. 1305. 1 United States Colombia Chinese Taipei Honduras Venezuela Panama El Salvador Mexico Central America Bolivia Guatemala original NAFTA agreement Chile Uruguay CAFTA identical language of provisions Korea almost identical language, same provisions Nicaragua same content, different language EFTA Canada EU

Adoption of common/similar rules: Spread of NAFTA Art. 1305. 2 United States Colombia Honduras Chinese Taipei Venezuela Panama El Salvador Mexico Central America Bolivia Guatemala Chile original NAFTA agreement Uruguay identical language of provisions almost identical language, same provisions Nicaragua same content, different language Canada Korea Note: This pattern has not been spread in different language and same sense of provisions, which was done by Chile in the case of Art. 1305. 1

What remains to be done 1. 2. 3. 4. 14 On Why? Learn what the trade negotiators understood they were doing, with whom, and what outcomes they sought. On Why? Develop models that rationalise the signing of these provisions—from a welfare-theoretic and political economy perspective. On Effects: Better understanding of the effects of these provisions on parties and third parties—are they as benign as they might seem? On Systemic Effects: Examine theoretically whether the sequential adoption of these provisions by more parties does raise world welfare. (Must be some doubts here. )