Luigi Bocconi University Ph.D. program in International Law and Economics Course of Economics of European integration DOES REGIONAL INTEGRATION FOSTER OPEN TRADE? THE ECONOMICS OF PREFERENTIAL TRADE AGREEMENTS Francesco Argese (francesco.argese@phd.unibocconi.it) Marco Barassi (marco.barassi@phd.unibocconi.it) Abstract: This essay offers an overview of the debate on the merits of preferential trade agreements as instruments of trade liberalization. Conventional welfare analysis suggests that, because of their inherent discriminatory nature, preferential trade agreements are primarily trade diverting and impact negatively on the fragmentation of trade regulation. However, a paradigm shift towards deep integration allows to understand the very nature of the last generation of preferential trade agreements, which encompass positive integration tools and which may thus complement, rather than substitute, the multilateral process of trade liberalization. Keywords: Customs union, Deep integration, Economic regionalism, Multilateralism, Preferential trade agreement, Free trade agreement, Open regionalism, Trade liberalization, World Trade Organization
1. The rise of preferential trade: drivers and design Despite being one of the major exceptions to the most-favored-nation (MFN) treatment in GATT Article I, preferential trade agreements (PTA) in the form of either free trade agreements (FTA) or customs unions (CU) have been accommodated into GATT through Article XXIV. The GATT Contracting Parties and now WTO Members have been allowed to conclude such agreements, provided that they eliminate internal barriers on substantially all the trade (GATT Article XXIV:8) and that external trade barriers shall not on the whole be higher or more restrictive than before the grant of a tariff preference (GATT Article XXIV:5). The addition of Part IV in 1965 and the Enabling clause in 1979 has also provided for the possibility of nonreciprocal preferences between developed and developing countries ( Generalized System of Preferences ) as well as partial preferences among developing countries ( Global System of Trade Preferences ). Lastly, as result of the Uruguay round of multilateral trade negotiations, GATS Article V has introduced provisions for PTAs in services which parallel those in GATT Article XXIV for goods. Historically, economic regionalism has witnessed three major waves. The first was confined to the integration process initiated in Western Europe in the aftermath of WW2 with the establishment of the European Communities (1951, 1957) and the EFTA (1960). Outside Europe trade agreements were undertaken principally in Latin America and Africa, but they were found to be largely ineffective. A second wave started in the mid-1980s when the EC, all along its successive enlargements, first delayed the launching and then the conclusion of the Uruguay round, hence forcing the US to abandon its long-standing commitment to multilateralism grounded in the MFN principle. Recognizing that PTAs were the only means left for keeping the process of trade liberalization afloat, the US went on to conclude FTAs with Israel (1985) and Canada (1989), as well as the NAFTA jointly with Mexico and Canada (1994). A third push toward regionalism initiated in the years 2000s, whose most distinct feature has been the abandon of the original rationale for the US pursuit of PTAs, which turned into a race for securing preferentially the neighbors markets in parallel with the FTAs signed by the EU with its neighbors in Eastern and Central Europe. This race between the two giants has, in turn, led to a renewal of efforts for PTAs in Africa, Latin America, Central and Eastern Europe as well as in South and Central Asia that had previously been undaunted supporter of multilateralism. Today, among the WTO Members only Mongolia is not part to any PTA. 1 The creation of NAFTA in 1994 and the proliferation of PTAs in 2000s instigated a huge policy debate over the compatibility of an increasingly fragmented system of discriminatory preferences with the MFN principle. In this respect, recognizing that PTAs can promote further liberalization and may assist least-developed, developing and transition economies in integrating into the international trading system, the WTO Members renew [their] commitment to ensure that regional trade agreements are complementary to [the multilateral trading system] and consistent with its rules 2. The starting point of our analysis is represented by a paper published by Arvind Panagariya in 1999, which builds on an extensive literature review in the context of the said PTAs emergence. Being limited to trade in goods and taking into account exclusively the conventional static welfare effects related to tariff reduction, Panagariya finds that, because of their intrinsic discriminatory nature, PTAs are trade diverting and impact negatively on multilateral trade liberalization. On the other hand, moving beyond pure welfare analysis and taking account of their recent spread in scope and coverage, PTAs may be qualified as being integrationist devices, which complement, rather than substitute, multilateralism. 1 While in 1990 the PTAs in force (both notified and non-notified) were almost 50, this number raised up to around 300 in 2010. The average number of PTAs that a WTO member is party to is 13. See WTO World Trade Report (2011), p. 47. 2 Singapore WTO Ministerial Declaration (1996), para. 7. ~ 1 ~
2. PTAs as trade diverting devices Relying on the Vinerian static welfare analysis of economic integration, the fundamental assumption of Panagariya is that when [ ] countries import from and export to union members as well as outside countries, trade diversion is inevitable 3. Analyzing the economic effects of PTAs on member countries, Panagariya rejects the argument that when members are natural trading partners, i.e. a disproportionate share of world trade takes place among countries which are geographically proximate, the gains from trade creation within blocks are likely to outweigh any possible losses from external trade diversion. Especially when PTAs include both high-tariff and low-tariff countries, the more the former import from the latter, the more a PTA is trade diverting. The PTA-led redistributive effects make high-tariff countries worse off and low-tariff countries better off. In the end, if a PTA s members are small in relation to the outside world, the PTA turns out to be wholly trade diverting. Empirical evidence supports trade diversion, as it is in the case of MERCOSUR and NAFTA. A related issue concerns the rules of origin, which allow in an FTA to avoid trade deflection and determine whether a good is granted duty-free movement within the area. Panagariya argues that the rules of origin maximize trade diversion in high-tariff countries by requiring that goods destined to them be imported through their own borders rather than through low-tariff countries. Moreover, if traded intermediate inputs are used in production, in order to take advantage of the tariff preference in the final good FTAs members may be forced to shift their purchases towards internal sources of inputs even though these sources are more expensive than if purchased from more efficient external sources. These distortions due to the existence of complex sets of rules of origin are even multiplied because of overlapping FTAs. Aside from static welfare effects, entering into PTAs with developed countries would not grant any additional alleged benefits to developing countries. In particular, PTAs do not provide any further guarantee of market access than WTO agreements do. Moreover, at least in principle PTAs do not contain any provisions that can shield developing countries from administered protection in developed countries by, for instance, escaping anti-dumping and safeguards actions. 4 Even more, PTAs are not likely to impart credibility to domestic reforms, as proven by the peso crisis in Mexico in mid-1990s. An alleged side element of this lock-in effect would be the more effective dispute settlement of PTAs which, contrary to the State-to-State WTO dispute settlement system, is available to private parties. Yet, the access of private parties may result in governments being induced not to sign agreements that are otherwise beneficial. Turning to the implications for the multilateral trading system, some authors identify a domino effect whereby idiosyncratic events (such as the completion of the European single market in the mid-1980s) create the economic incentives for outside countries to seek entry into an existing PTA. Then, as long as the entry into the PTA is free, this process can lead to global free trade. Actually, in Panagariya s view, this process stops short of encompassing all countries and yielding global free trade. After a PTA reaches a certain size, insiders have an incentive to block further entry, whereas the profits of outsiders decline monotonically as the PTA expands. Even assuming that countries behave as Nash players and choose tariffs to maximize welfare, their welfare peaks before a trading block expands to include all countries. 3 Panagariya (1999), p. 483. 4 In this respect, for instance, the special agreements on sugar and orange juice in place in the NAFTA explicitly allow for the play of administered protection in the event of import surges from Mexico. Also, side agreements on labor standards and environment, which give the US huge powers to subject Mexico to dispute settlement procedures (with fines up to USD20 million), bear testimony to the fact that NAFTA generally resists deep penetration of the US markets by Mexico. ~ 2 ~
In light of the foregoing, Panagariya concludes that PTAs serve mainly as stumbling blocks to global free trade. 5 Much of the evidence suggests a rise in trade barriers against non-members following the formation of a PTA. While a PTA can lead to a reduction in tariffs on intermediate inputs, yet such a reduction increases effective protection to the final good and can turn out to be welfare-reducing. In addition, to the extent that a country s applied tariff is below its bound rate, as usually for developing countries, producers lobbying can very well result in increased tariffs. Especially in the case of FTAs, if a country is dependent on tariffs for revenue purposes to maintain fiscal balance or to face fiscal crises, as is true again of many developing countries, the removal of tariffs on the partner countries may force to raise the external tariffs. 6 Even in the absence of any increase in external trade barriers, the proliferation of crisscrossing FTAs leads to a replacement of the non-discriminatory MFN tariffs by a spaghetti bowl phenomenon (i.e. each FTA has its own rules of origin varying across goods and different tariff rates apply to the same good depending on what origin is assigned to it) or hub-andspoke systems (where the hubs could have zero tariffs vis-à-vis each country with whom it has an FTA, while the spokes could have free trade only with few of those countries). Finally, in a political-economy setting in which producers lobbying play the central role in determining trade policies, PTAs can energize and unify protectionist lobbies, turning them into effective obstacles against multilateral liberalization. This is especially true in case of PTAs between developed and developing countries, which in public mind are associated with large inflows of labor-intensive goods and lower wages for unskilled labor. 3. Beyond pure welfare analysis: the deep integration agenda The position of the GATT/WTO and of the advocates of the PTAs consistence with the multilateral trading system relies on the concept of open regionalism, which tends to identify approaches of regional integration different from the closed, import-substituting regionalism of the 1950s and 1960s. According to the Council of Economic Advisors to the US President, open regionalism refers to plurilateral agreements that are non-exclusive and open to new members to join. It requires first that plurilateral initiatives be fully consistent with Article XXIV of the GATT [ ]. Beyond that, it requires that plurilateral agreements not constrain members from pursuing additional liberalization either with non-members on a reciprocal basis or unilaterally. [ ] Finally, open regionalism implies that plurilateral agreements both allow and encourage non-members to join 7. Yet, even in presence of an open membership, discrimination against non-members at any point in time remains in place by definition and this does not prohibit the world trading system to become fragmented. It is just this fear of fragmentation that prompted Mr. Renato Ruggiero to go a step beyond in defining open regionalism. Apart from the mere compliance with GATT Article XXIV, to be consistent with the multilateral trading system a gradual elimination of internal barriers to trade within a regional grouping will be implemented at more or less the same rate and on the same timetable as the lowering of barriers towards non-members. This would mean that regional liberalization would in practice as well as in law be generally consistent with the [MFN] principle 8. If one was looking for an example of 5 By using the median voter model, Levy (1997) shows that it is especially in a Krugman (rather than Heckscher-Ohlin) set up with product differentiation that PTAs can turn into a stumbling block to global free trade. With the gains from trade deriving from differences in factor endowments as well as increased variety, it is possible for median voters in member countries to achieve a higher level of utility under this alternative than under the multilateral option. 6 For instance, in the aftermath of the peso crisis in Mexico (1994-1995), the external tariffs on more than 500 items went up from 20 to 35 per cent. The same phenomenon was observed in the wake of the fiscal crisis in Brazil (1998-1999), where MERCOSUR raised its common external tariff by three per cent. 7 CEA (1995), p. 220. 8 WTO Secretariat (1996), p. 11. ~ 3 ~
Ruggiero s preferred form of regionalism, East Asia undoubtedly provides it. Yet, in Panagariya s view, even in this case regionalism can neither escape being discriminatory nor overcome the fundamental contradiction between openness to and discrimination against outsiders. Having said that, this does not necessarily mean that the argument on the compatibility of PTAs with multilateralism should be resolved in the sense of the impossibility to reconcile the two solutions. If the first wave of regionalism may be characterized as falling into the patterns described by Panagariya, newly signed PTAs seem to encompass a different approach to economic integration. Multilateral MFN tariffs have been decreasing for decades and have now reached their lowest level (4 per cent on average in 2009). Moreover, given the considerable number of zero duty MFN rates in many countries and widespread goods exclusions, only 16 per cent of world trade is eligible for preferential tariffs. Therefore, rather than being confined to shallow, negative integration tools aiming at eliminating tariff barriers among members, the recently-signed PTAs promote deep 9, positive integration through the coordination, if not a complete harmonization, of inside-the-border policy areas and regulatory frameworks, which are not (or only partially) dealt with at multilateral level. Data from the WTO show that most of the new PTAs cover areas such as competition, product standards, regulatory regimes, investments, movement of capital, agriculture, environmental and labor standards. This explains why the deep integration agenda cannot be pursued independently of PTAs. 10 The rationale for a deep integration agenda is the development of increasingly fragmented production networks, which reflect the need for companies to gain competitive advantages through the exploitation of factor price differences. The more fragmented the production chain is, the more additional costs are due to the difficulties of doing business in different countryspecific settings. The aim of deep integration agreements is hence to fill the governance gap between countries resulting from the globalization process on a range of regulatory issues that are essential to the success of the networks. Signing deep integration PTAs is estimated to increase trade in production networks by almost 8 per cent on average. In turn, high levels of trade in production networks significantly raise the likelihood of signing deep integration agreements. For the debate on the implications of PTAs for the multilateral trading system, the most relevant consequence from the above is that, since deep integration provisions are embedded in regulatory frameworks that apply to all, they de facto extend their scope to non-members as well. For instance, if transparency provisions are implemented as a consequence of a bilateral agreement, third countries will automatically benefit from them. The same holds for the principle of mutual recognition, which applies regardless of the origin of the products. It is thus clear that, even if difficult to quantify, PTAs encompassing policy areas other than tariff reduction may generate positive spillover effects for third countries, hence complementing (rather than substituting) the multilateral process. 4. Rethinking economic regionalism Consistently with his methodological and scientific approach, Panagariya suggests that trade diversion in world trade may be minimized by: (i) placing a moratorium on the expansion of existing PTAs while, at the same time, moving speedily to bring the unfinished multilateral agenda to its logical conclusion so as to eradicate the need for new PTAs; (ii) modifying GATT Article XXIV so as to require each FTA/CU members to bind their tariffs/common external tariff to the rates applied at the time of negotiations and ensure that tariff protection against non- 9 Lawrence (1997), p. 4. 10 As it has been observed, nowadays a discriminatory tariff may matter less than the selective recognition of product safety standards or selective access to government procurement matters (Mattoo and Subramanian (2011), p. 6). ~ 4 ~
members is not raised in the future; (iii) requiring that within a specified time period (e.g. five to ten years) following the grant of a tariff preference to FTA members, the external tariff should be reduced to the level of the preferential tariff as well; and (iv) eliminating the rules of origin on a product in the member country with the lowest tariff on that product or, alternatively, not permitting any rules of origin whenever the difference between the highest and lowest external tariff rates in a PTA is less than a pre-specified limit (e.g. five per cent). A focus exclusively on trade diversion effects of tariff concessions may nevertheless lead to misleading results when evaluating the contribution of recently-signed PTAs vis-à-vis the multilateral trading system. As observed earlier, deep PTAs can be a complement of the multilateral process. Two important observations follow from this conclusion. On the one hand, the economic theory needs to go beyond the standard welfare analysis of PTAs. On the other hand, further research is needed to focus increasingly on the economic, legal and political reasons for establishing PTAs that go beyond the mere reduction in tariffs and evaluate the possibility for positive integration in order to outweigh the negative consequences of trade diversion which is still generated by tariff preferences. We believe that this paradigm shift is perfectly embedded in the European integration process. The beginnings of such a process relied upon a negative approach to integration in view of the establishment of a common market and a customs union (the latter by July 1st, 1968). It is worth observing that the abolition of tariff barriers as well as quantitative restrictions to trade and any other measures having equivalent effect was on the whole trade creating, with a significant expansion of both intra- and extra- EC trade in industrial goods. Such a significant development was led particularly by the very rapid post-ww2 economic growth experienced by the EC countries and the reduction of average tariffs on imports of industrial goods as a result of the GATT rounds of trade liberalization. Only trade in agricultural products was (and actually continue to be) characterized by trade diversion. Despite the establishment of a customs union, the EC was still a fragmented market especially because of the proliferation of non-tariff barriers throughout the 1970s. The resulting industrial structure was thus marked by too many, inefficient, small firms providing poor and/or lowquality goods at relatively high prices. To overcome this situation, which was considered one of the major causes of the Euro-sclerosis underlying the productivity gap with the US and Japan at that time, in the mid-1980s the Single Market Program aimed at abolishing the remaining non-tariff, technical barriers to trade between Member States and moving towards a paradigm of positive, deep integration. As a result of the completion of the internal market by December 31 st, 1992, between 50 and 60 per cent of the world trade now occurs in the EC/EU. 11 Notwithstanding these achievements, there is still room for further liberalization as remarked by the 2010 Monti s report on the future of the EU 12, which outlines a comprehensive strategy focusing on missing links and new frontiers to make a re-launch of the internal market politically successful and economically and socially viable. The current integration fatigue eroding the appetite for more Europe and a market fatigue with a reduced confidence in the role of the market, however, leave open a huge question: Is the EU still the model for regional integration efforts around the globe and a valuable asset for the future of international trade regulation? 11 If the static gains from the establishment of a customs union in 1968 were quantified as being 1 to 2 per cent of EEC GDP and the dynamic gains to be around 5 per cent of EEC GDP, the 1988 Cecchini Report Europe 1992: the overall challenge estimated ex ante that, by making the allocation of resources within the EC more efficient, the completion of the internal market would have been produced a 2.5-6.5 per cent increase in the level of EC GDP. 12 Monti (2010), p. 1. ~ 5 ~
References Altomonte C., Nava M. (2005), Economics and Policies of an Enlarged Europe, Edward Elgar Council of Economic Advisors (CEA) (1995), Economic Report of the President, Washington DC Krueger A. (1995), Free trade agreements versus Custom unions, NBER Working Paper n. 5084 Lawrence, R. (1997), Regionalism, multilateralism and deeper integration, Brookings Institution Papers on Economic Activities Mattoo A., Subramanian A. (2011), A China Round of Multilateral Trade Negotiations, Peterson Institute for International Economics, Working Paper Series WP 11-22 Monti M. (2010), A new strategy for the single market: At the service of Europe s economy and society, available at: http://ec.europa.eu/bepa/pdf/monti_report_final_10_05_2010_en.pdf Panagariya A. (1999), The regionalism debate: An overview, The World Economy, 22(4), p. 477-512 Rodrik D. (2000), How far will international economic integration go?, Journal of Economic Perspectives, 14(1), p. 177-186 Sapir A. (2011), European Integration at the Crossroads: a Review Essay on the 50th Anniversary of Bela Balassa s Theory of Economic Integration, Journal of Economic Literature, 2011, 49 (4), p. 1200-1229 Singapore WTO Ministerial Declaration (1996), WT/MIN(96)/DEC, adopted 13 December 1996, http://www.wto.org/english/thewto_e/minist_e/min96_e/singapore_declaration96_e.pdf WTO Secretariat (1996), The road ahead: International trade policy in the era of the WTO, Fourth Annual Sylvia Ostry Lecture, Ottawa, WTO Press/49, Geneva, 29 May WTO Secretariat (2011), The WTO and preferential trade agreements: From co-existence to coherence, World Trade Report 2011, Geneva ~ 6 ~