FLACSO Argentina - Lessons from past ExperiencesSerie: Documentos e Informes de Investigación Nº 229 Area: Relaciones Internacionales REGIONAL TRADE ARRANGEMENTS: LESSONS FROM PAST EXPERIENCES Roberto Bouzas February 1999 Trabajo preparado para su publicación en M. Rodríguez Mendoza, P. Low y B. Kotschwar (editores), Trade Rules in the Making: Challenges in Regional and Multilateral Negotiations, Washington DC: The Brookings Institute, 1999. TABLE OF CONTENTS Page WHAT DO MEMBER COUNTRIES LOOK FOR IN PTAS? 2 THE EUROPEAN EXPERIENCE: SETTING THE YARDSTICK 4 The European Community 5 The European Free Trade Association (EFTA) 7 Trade Discrimination in Central and Eastern Europe 8 PREFERENTIAL TRADE ARRANGEMENTS IN THE WESTERN HEMISPHERE 9 PREFERENTIAL TRADE ARRANGEMENTS IN ASIA AND THE PACIFIC 16 PREFERENTIAL TRADE ARRANGEMENTS: AFRICA AND THE MIDDLE EAST 20 LESSONS FROM PAST EXPERIENCES 22 CONCLUSIONS 25 References 27 During the post-war period the major vehicles for liberalization were multilateral trade negotiations (MTNs) through successive rounds at the General Agreement on Tariffs and Trade (GATT) and the unilateral lowering of trade barriers. Preferential trade arrangements (PTAs) played a secondary role, but interest in them peaked in the 1950-60s and again in the late 1980s-1990s. In both periods, developments in Europe set the pace. The two waves of PTAs differ in content, coverage and membership, with the latter displaying a broader coverage of issues and disciplines and a more active role of the United States --- traditionally the champion of the most-favored nation (MFN) principle. When their accomplishments are held up to their stated aims, PTAs have rarely been successful. In most cases commitments have not translated into policies and achievements have fallen short of goals. The European experience (particularly the European Community) has been the most auspicious. With less ambitious objectives and a shorter record, the Canada-US Free Trade Agreement (CUSFTA) and the Australia-New Zealand Closer Economic Relations Trade Agreement (ANZCERTA) have also made significant progress towards their stated aims. The vast majority of PTAs have failed to meet their goals for reasons of both design and implementation. Past experience can suggest some lessons for the future, but care must be taken in drawing general conclusions. First, the success or failure of a PTA can be measured either against the stated objectives or alternative scenarios -- which require speculation about counterfactuals. Second, PTAs differ in structure, as suggested by the existence of areas of preferences, free-trade agreements, customs unions and common markets. Finally, countries engage in PTAs for very different motivations. This chapter examines past experience to draw some policy lessons for the future, but cautions against easy generalizations and focuses on policy outcomes rather than on the theoretical debate. It also side-steps the difficult issue of whether PTAs are helpful or detrimental to the multilateral trading system, taking instead the line that the coexistence of regional and multilateral approaches to trade negotiations suggests that they are complementary rather than mutually exclusive. The chapter provides a brief overview of the various motives for forging PTAs and summarizes the record of PTAs in Europe, the Western Hemisphere, the Asia-Pacific region and Africa, pointing out common threads and peculiarities. Finally, some lessons and conclusions are drawn from the preceding discussion. WHAT DO MEMBER COUNTRIES LOOK FOR IN PTAS? The incentives to enter into PTAs are diverse and may go well beyond economic considerations. In effect, political and security concerns can play a decisive role in binding two or more countries together through a PTA. Yet if such arrangement brings about persistent net losses in economic welfare to member countries, it is unlikely to survive as a voluntary one. Thus, economic considerations will play an important role in any case. Narrow trade considerations are often seen as the objective behind participation in a PTA. However, the strongest case against PTAs comes from standard trade theory, which places PTAs in the realm of second-best outcomes. In effect, Viner’s discussion of static, allocative effects reached the key conclusion that the welfare of countries taking part in PTAs may be reduced if trade diversion (when low-cost sources of supply are replaced by higher-cost suppliers from within the region) out-weighs trade creation (when higher-cost domestic suppliers are replaced by lower-cost suppliers from within the region). Based on this framework, the Cooper-Massell-Johnson proposition concluded that discriminatory liberalization could not raise the welfare of member countries beyond what would otherwise result from unilateral liberalization on an MFN-basis. PTAs may increase members’ welfare at the expense of the rest of the world (worsening the latter’s terms of trade). However, this would be a recipe for conflict and retaliation, and may eventually reduce aggregate welfare. The ambiguous conclusions reached by standard trade theory do not get better when dynamic considerations are taken into account. Even the direction of effects over the rest of the world may change. Another motivation is to gain insurance against future increases in trade protection. This motivation may be relevant for small countries vulnerable to trade policy changes by large and important trade partners. Small countries, especially, may find that effective multilateral rules and enforcement capabilities are the best insurance against undue exercise of market power by larger partners. But policy choice by small economies is generally of little relevance to the performance and health of the multilateral system. Thus, under certain circumstances buying insurance through a PTA may be an attractive option -- depending upon the price to be paid for this insurance. Governments may also find PTA membership useful to gain credibility for their domestic policies. This raises the question of whether the same credibility could not be gained from multilateral commitments. Political economy may offer an answer. Domestic groups (particularly exporters) may mobilize more readily to support a more open trade regime in the context of a PTA (in which gains from market access are more clearly identifiable) than for unilateral or multilateral liberalization. Regional agreements (particularly among asymmetric partners, as in North-South PTAs) may offer more effective constraints on reluctant enforcement or policy reversals than multilateral commitments. International bargaining considerations may also play a role. Countries may find that their bargaining stance strengthens if they act together in trade negotiations. Similarly, frustration over slow progress in multilateral negotiations may stimulate countries that share interests to come together in a PTA. Streamlining differences between countries through PTAs may be an effective way to eventually reach a multilateral deal. THE EUROPEAN EXPERIENCE: SETTING THE YARDSTICK While preferential trade arrangements have proliferated in Europe, not all have been equally successful. The most far-reaching was the European Community (EC), which became the centripetal force of European economic integration. The European experience has been so dynamic and path breaking that the EC is seen as the paradigmatic example of a successful PTA and is frequently used as the yardstick against which success and failure of PTAs is measured. In contrast, the competing European Free Trade Area (EFTA) has found it difficult to maintain its original membership: countries gradually left to join the EC. Moreover, by the early 1990s, those Western European countries that remained in EFTA had undertaken part of the EC deepest integration commitments through the European Economic Area (EEA). At last, the EC turned into a key reference for preferential trade arrangements in Central and Eastern Europe (CEE) as well. The European Community The foundations of European regionalism were laid during the reconstruction that followed World War II. European regionalism benefited from the financial and political support of a security-driven US policy, and went much further than envisaged in US plans. The first truly European (or rather French) initiative was the Schuman Plan of May 1950, which led the "original six" (Belgium, the Netherlands, Luxembourg, France, Italy and Germany) to sign the Treaty of Paris and establish the European Coal and Steel Community (ECSC) one year later. Although the main objective was to eliminate barriers and promote competition in the coal and steel industries, many specific provisions of the agreement were hardly compatible with economic liberalism. The Treaty of Paris also established a supranational organ (the High Authority) empowered to levy taxes, influence investment decisions and even impose minimum prices and production quotas under certain conditions. The signing of the Treaty of Rome in 1957 created two new institutions: the European Atomic Energy Community (EAEC) and the European Economic Community (EEC). The purpose of the EEC was to establish "the foundations for an ever-closer union among the European peoples" through the creation of a common market and, more specifically, a customs union and a limited number of sector policies. The Treaty of Rome established -among other things- the elimination of customs duties and quantitative restrictions; the adoption of a common customs tariff and of a common commercial policy towards third countries; the abolition of obstacles to freedom of movement for persons, services and capital between Member States; the adoption of common policies in the sphere of agriculture and transport; and the promotion of competition in the common market. Following the precedent set by the ECSC, the Treaty of Rome established an elaborate governance structure composed of a Commission (the administrative and technical body), the Council of Ministers (the decision-making authority), the Court of Justice and the Parliament. Since Member states accepted the supremacy of Community law over national law in the Community’s areas of competence, the traditional boundaries of national sovereignty were broken. In 1968, one and a half year in advance of schedule, the EEC concluded the implementation of a common external tariff and the elimination of customs duties and quantitative restrictions on intra-EEC trade. But freedom of movement for goods, services, labor and capital on the basis of the principle of national treatment required a painful process of harmonization. In a precedent-setting decision, the European Court of Justice in 1979 set in motion the application of the principle of mutual recognition by ruling that Germany could not prevent the French alcoholic beverage Cassis de Dijon to be sold in Germany because it did not meet the German standard for liquor. The application of mutual recognition left behind the exhausting phase of detailed harmonization and was probably the single most important step (after the elimination of tariffs) towards the free circulation of goods. In 1986 the Single European Act (SEA) amended the Treaty of Rome and instituted the principles of qualified majority voting and of mutual recognition in areas where there were no community-wide standards. The SEA also launched the EC 92 program to complete the single market. In March 1992 a new Treaty on Closer Economic Union was signed in Maastricht. The Treaty rested on three uneven pillars: the three old Communities (now renamed the European Community), a Common Foreign and Security Policy (CFSP) and cooperation in the fields of Justice and Home Affairs. The provisions for the creation of the European Monetary Union (EMU) were by far the most specific and far reaching. Deepening in the European Community went hand in hand with widening. In effect, in 1973 the Community broadened its membership to include Britain, Denmark and Ireland, followed by Greece in 1979 and by Spain and Portugal in 1986. In 1995, the inclusion of Austria, Finland and Sweden raised the total number of member states to fifteen. The Community is projected to include at least five additional Central and Eastern European members by year 2010. The European Free Trade Association (EFTA) Shortly after the Treaty of Rome a group of European countries outside of the EC (Austria, Denmark, Norway, Portugal, Sweden, Switzerland and the United Kingdom) created the European Free Trade Association (EFTA). In contrast to the EC, the objectives of EFTA were purely economic and more limited in terms of product coverage (restricted to industrial goods), potential further development and political engagement. In effect, the EFTA process was wholly intergovernmental with no common court or system of law. The centripetal forces unveiled by the EC could not be resisted by EFTA. In the early 1970s the members of EFTA negotiated bilateral industrial free-trade agreements with the EC, to be followed shortly by the accession of Denmark and the United Kingdom to the EC. Portugal followed suit in 1986 and Austria, Finland and Sweden adhered to the EC in 1995. As a result, by the late 1990s only Iceland, Liechtenstein, Norway (where a plebiscite to join the EC was defeated in 1995) and Switzerland remained EFTA members. In 1992, building upon the bilateral free trade agreements concluded in the early 1970s, the EC and EFTA member states created a European Economic Area (EEA). The agreement incorporated many facets of the EC (such as the free circulation of persons, goods, services and capital; competition rules and EC law as a common base of reference), but did not provide for EFTA country participation in EC decision-making, the Common Agricultural Policy (CAP) or the common external tariff. That the EEA agreement did not satisfy all EFTA member states was demonstrated by the fact that Austria, Finland, Norway and Sweden negotiated accession agreements immediately after the EEA was established. Trade Discrimination in Central and Eastern Europe Before the collapse of centrally planned economies in the late 1980s, most Central and Eastern European (CEE) economies were members of the Council for Mutual Economic Assistance (CMEA) founded in 1949 under the coordination of the Soviet Union. The disintegration of CMEA was followed by the signature of regional and bilateral PTAs throughout the region (including some ex-Soviet republics). The two major agreements were the Europe Agreements between the EU and six CEE countries (Bulgaria, Czech Republic, Hungary, Poland, Romania and Slovakia) signed in 1992-1993 and the Central European Free Trade Agreement (CEFTA) signed in March 1993 between the Czech Republic, Hungary, Poland and Slovakia (joined by Slovenia in 1996). The trade component of the Europe Agreements provides for the establishment of a free trade area on industrial products among the EU and the CEE economies over a period of ten years. The agreements cover not only merchandise trade (except agriculture) but also trade in services, foreign investment, payments systems, capital markets regulations and broader economic cooperation. The Europe Agreements are generally regarded as having played a positive role in the transition of CEE countries to a market economy by improving market access into the EU, stimulating mutual trade and foreign investment and reducing uncertainty in the context of far-reaching structural change. The original CEFTA called for the complete elimination of trade barriers in industrial products by year 2001, with exceptions in agriculture and so-called "sensitive" products, with each member country negotiating separate bilateral protocols on the speed and extent of concessions. The rationale for CEFTA was based on "natural" partnership and cultural and historical commonality. Impediments have been sizeable: recurrent problems of implementation (aggravated by a highly unstable macroeconomic environment) and modest political incentives. The scope of CEFTA has been more limited than that of the European Agreements between the CEE economies and the EU. PREFERENTIAL TRADE ARRANGEMENTS IN THE WESTERN HEMISPHERE The history of PTAs in the Western Hemisphere is one of contrasts. Discrimination was largely absent North America for most of the post-war period until the mid-1980s, when the United States embraced PTAs as a complement to its traditional multilateral stance, unexpectedly spurring the first PTA between developed and developing countries. In contrast, PTAs flourished in Latin America and the Caribbean. While the early PTAs in Latin America and the Caribbean were largely paper commitments by the 1980s, a new more outward-looking trade policy environment brought about new incentives to PTAs in the late 1990s. Despite the evident predominance of a multilateral approach, during the post-war period discrimination was not completely absent from US trade policies. Prominent examples of discrimination are the exclusion of communist countries from the benefits of the MFN principle, the US-Canada Auto Pact, the Generalized System of Preferences (GSP), the Caribbean Basin Initiative and the US-Israel Free Trade Agreement. But throughout the period discrimination was subordinate to MFN liberalization under the auspices of the GATT. Moreover, discriminatory arrangements made up a relatively minor share of total US foreign trade and they were basically driven by security and/or foreign policy considerations. This stance has changed since the mid-1980s. Neither CUSFTA nor NAFTA can be adequately understood on purely foreign policy or strategic grounds. International bargaining considerations provided the basic thrust for the US drift towards PTAs with Canada and Mexico in the late 1980s. From the US standpoint, CUSFTA was stimulated by growing dissatisfaction (particularly in Congress) with the evolution of the multilateral trading regime. To a significant extent NAFTA was an outgrowth of this as well. The modest results –from a US standpoint- achieved in the Tokyo Round and the failure to make substantial progress in the early years of the Uruguay Round stimulated a bilateral emphasis to advance US trade interests. Approval of the Single European Act in 1986 and growing fears of a "Fortress Europe" were also influential. Growing Congressional participation in trade policymaking also supported discrimination and a bilateral approach: since the mid-1970s each new trade legislation emphasized the use of unilateral and bilateral mechanisms (including PTAs) to promote US trade objectives. On balance, both CUFSTA and NAFTA served to advance the so-called "new issues" of trade in services, protection of intellectual property rights and foreign investment. Moreover, NAFTA included two side agreements on labor and the environment, which allow for the possibility of lifting trade concessions when governments fail to meet national standards. The Canadian and Mexican incentives to enter into an FTA with the United States varied. For Canada the dominant objective was to guarantee more stable access to the U.S. market by limiting U.S. discretion in the implementation of trade remedy laws and by protecting Canadian exporters from future increases in U.S. protection. Mexico was also interested in increasing its attractiveness to foreign investors, as well as in buying credibility for and locking in domestic reform policies. Both CUSFTA and NAFTA were less ambitious than the EC. PTAs in North America did not consider the creation of a common market (with full mobility of labor) or even a customs union. However, both agreements (and particularly NAFTA) are outstanding examples of a new generation of GATT-plus PTAs which include "deeper integration" commitments. CUSFTA eliminated tariff and non-tariff barriers on goods, but maintained certain exceptions (such as beer) or covered certain goods only partially (such as agriculture or textiles). Provisions on services were innovative, as no multilateral agreement was yet in place, but many important sectors were excluded. Moreover, prevailing discrimination was grandfathered. Canadian restrictions on foreign investment were relaxed but not eliminated, while contingent protection remained untouched, except for the substitution of internal judicial review procedures for binational panels. NAFTA differed from CUSFTA in a number of ways. First, it was the first PTA to bring together developed and developing countries. Second, in some areas NAFTA went beyond CUSFTA, expanding the GATT-plus features of the latter. NAFTA’s provisions on trade in services, protection of intellectual property rights and rules against distortions to investment are more comprehensive. However, NAFTA´s major shortcoming was a complex and detailed system of rules of origin for products such as automobiles, electronic goods, textiles and apparel. The unfinished business include sector exceptions (Mexican energy and Canadian "cultural" industries), harmonization of contingent protection and subsidies (particularly in agriculture).There is agreement that CUSFTA has contributed to expand US-Canada bilateral trade and investment flows. The assessment is more mixed as regards the effects which CUSFTA may have had on contingent protection: although harassment on Canadian exports may have diminished, discretion in the implementation of US "trade remedy legislation" is still high. NAFTA is younger than CUSFTA, which makes any assessment tentative. Since the enforcement of the trilateral agreement trade and foreign investment increased (probably contributing to Mexico’s recovery from the sharp 1995 recession), while inter and intra-industry specialization proceeded along expected lines. NAFTA also contributed to lock in policy reforms in Mexico, although countries such as Chile have been equally capable of sustaining reforms without bilateral commitments as those of NAFTA. Similarly, as may be expected, most of the burden of adjusting to NAFTA was borne by the smaller and most heavily protected economy. This coincided with a worsening of the regional and functional distribution of income in Mexico. However, it is uncertain what the role of NAFTA in this process has been. In contrast to the North American experience, during most of the post-war period Latin American and Caribbean countries engaged actively in PTAs. In 1960, under the inspiration of the United Nations Economic Commission for Latin America (ECLAC), Mexico and six South American countries established the Latin American Free Trade Association (LAFTA), with the objective of eliminating all barriers to intraregional trade in twelve years. By the mid-1960s LAFTA had increased its membership to eleven countries, but each annual round of tariff-cutting negotiations rendered decreasing returns. Special financial arrangements (eg: the Agreement on Multilateral Settlements and Reciprocal Credits) and new mechanisms (eg: complementation agreements enabling two or more countries to liberalize trade on a sectoral basis) contributed to foster intraregional trade. However, the initial objective of establishing a free-trade area in twelve-years was not met. In 1969 a subset of LAFTA member countries (Bolivia, Colombia, Chile, Ecuador and Perú) signed the Cartagena Agreement and established a more ambitious Andean Common Market. Member countries expected that the new institutional arrangements embodied in the Agreement would overcome many of the shortcomings of LAFTA. An executive body with "supranational" powers was established and a schedule for trade liberalization and gradual implementation of a common external tariff to be completed in 1980 was agreed. The pact also included mechanisms to promote an equitable distribution of benefits. PTAs were not restricted to South America. In 1960, superseding a 1958 free-trade agreement, Costa Rica, El Salvador, Guatemala, Honduras and Nicaragua set the more ambitious objective of creating a Central American Common Market (CACM). The aim was to reach a free trade area (in industrial goods) in a period of five years and to implement a common external tariff. Although not all objectives were fully met, the CACM was regarded as one of the few successful agreements among developing countries. Intraregional liberalization proceeded swiftly and trade expanded rapidly, especially in industrial products. To complete the regional chessboard a Caribbean Free Trade Agreement (CARIFTA) was signed in 1967, to be superseded six years later by the Caribbean Community, the objective of which was to create a customs union and a common market. In the initial years of CARIFTA intraregional trade increased, but stalled as of the mid-1970s. This first generation of Latin American and Caribbean PTAs shared a number of common features. First, all agreements were targeted at strengthening weak trade relations rather than at providing a more positive framework for already existing economic ties. Second, PTAs were regarded as a complement to import substitution industrialisation: broader regional markets would remove the constraints posed by small domestic markets. Third, pragmatism was dominant and many original commitments (such as regional planning and co-ordination of direct foreign investment) were over-ambitious and were never met. Despite these common threads, performance was not completely similar: in particular, and in contrast to LAFTA and CARIFTA, the CACM made significant progress during its first decade. Reasons for CACM´s relative success were both political and economic. One factor was a tradition of colonial administrative unity and repeated attempts at political unification. Economic factors include the favorable contribution of emerging industrial interests and the limited opposition posed by established producers (industrialization in the CACM proceeded pari passu with the implementation of trade preferences). By the 1970s it was apparent that economic integration had failed to fulfil its promise. Contributing factors were protectionist trade regimes, conflicting interests, weak economic ties, chronic macroeconomic instability, and over-ambitious targets. Lack of progress in the 1970s was followed by collapse in the 1980s. The Montevideo Treaty of 1980 created the Latin American Integration Association (LAIA) in substitution for LAFTA. LAIA included more flexible mechanisms, such as a wider scope for bilateral agreements, the benefits of which may not be extended to the rest of LAIA on a MFN basis. Despite Venezuela’s accession in 1973 the Andean Pact waned in the 1970s, with Chile withdrawing in 1976 after a new government departed radically from prevailing economic policies. The CACM and CARICOM also stalled, the former seriously challenged by growing political tensions. The turning point for PTAs was in the 1980s, when the debt crisis led to an increase in import barriers and a sharp contraction of intraregional trade. By the end of the decade most Latin American and Caribbean PTAs were largely paper commitments, with a remarkable gap between reality and objectives. After the NAFTA negotiations, and launching of the Enterprise for the Americas Initiative in 1990, PTAs gained new life throughout the region. Since the late-1980s, trade policy reform and more outward-oriented trade regimes also spurred PTAs. The result of this revival was a "spaghetti bowl" of PTAs linking pairs or groups of countries. In contrast to the previous goal of expanding protected domestic markets, the new focus of trade liberalization in Latin America and the Caribbean has been to improve and secure market access, facilitate industrial restructuring and enhance the economies attractiveness for foreign investment. International bargaining considerations also played a role, particularly after the swift change in the US trade policy stance. The revival of trade discrimination in the region materialized in two stylized approaches. One was multiple FTA negotiations. The second was the creation of customs unions. The first approach was characteristic of Mexico and Chile, which have signed bilateral or trilateral FTAs with Latin American and Caribbean partners. The second is best exemplified by the Southern Common Market (MERCOSUR) comprised by Argentina, Brazil, Paraguay and Uruguay. MERCOSUR made significant progress towards a free trade area in a brief period of time and intra-regional trade flows experienced a significant rise. Although non-tariff barriers and non-border measures still pose market access problems, preferences have reached 100% over MFN tariff rates for nearly all products, with the remainder subject to an automatic schedule for liberalization to be concluded in 1999. The major question mark regarding PTAs in the Western Hemisphere is whether negotiations to create a Free Trade Area of the Americas (FTAA) will succeed in bringing together these multiple strands of trade discrimination. Negotiations were launched in April 1998 but differences remain as to the shape of the agreement and its mode of implementation. PREFERENTIAL TRADE ARRANGEMENTS IN ASIA AND THE PACIFIC PTAs have been neither popular nor successful in the Asia-Pacific. The major initiatives are the Association of Southeast Asian Nations (ASEAN) in 1977 (transformed into the ASEAN Free Trade Area in the early 1990s), the Australian-New Zealand Closer Economic Relations Trade Agreement (ANZCERTA) of 1983, and the South Asian Preferential Trade Area (SAPTA) created in 1993 by the South Asian Association for Regional Co-operation (SAARC). The Asia Pacific Economic Co-operation (APEC) forum launched in 1989 has a Pacific rather than an Asian scope (it includes the United States, Canada, Mexico and Chile) but so far it has set only loose commitments and has failed to come up with a blueprint for trade liberalization. ANCZERTA brought together two countries with similar political institutions and cultural traditions at a time when both were undergoing a far-reaching process of economic reform. In practice, it is the only experience that can be called a success in the region. ASEAN´s record with APTA and AFTA is poor: preferential liberalization advanced slowly and intraregional trade flows remain modest. The record in South Asia is gloomier. Various reasons have been offered to explain the comparatively short experience with PTAs in the Asia-Pacific and their limited success. One is the region’s reliance on world rather than regional or domestic markets. Moreover, in the past the GATT regime allowed ample room to domestic trade policy autonomy, enabling the survival of local institutions and trade policy regimes. Despite these features intraregional trade has expanded fast, mainly as a result of the region’s high rates of economic growth than PTAs. AFTA was the outgrowth of the ASEAN Preferential Trade Area (APTA) launched in 1977 by the Association of Southeast Asian Nations (ASEAN). ASEAN was established in 1967 by Brunei, Indonesia, Malaysia, Philipines, Singapore and Thailand on security and strategic concerns, but with no economic objectives at the onset. In 1977 member countries launched an ASEAN Preferential Trading Area (APTA), but APTA remained essentially dormant until the early 1990s. In 1992 it was superseded by an agreement to establish an ASEAN Free Trade Area (AFTA) within a period of 15 years. APTA, based on a "positive" list approach, took a piecemeal approach to PTAs. A decade after inception only a handful of products had been covered by preferences with a very modest impact on bilateral trade flows. One reason for slow progress was the prevailing differences in trade protection and the anticipated asymmetric effects of trade liberalization on member countries. Trade within the region was very modest and increased little after the agreement. The AFTA agreement established the removal of all quantitative restrictions and the reduction of tariffs to the 0-5% range by year 2007, later advanced to 2003. The tariff-cutting mechanism or Common Effective Preferential Tariff (CEPT) excluded unprocessed agricultural products, natural resources and services. Preliminary evidence suggests that the effects of AFTA on intraregional trade flows have been very modest, partly due to the slow pace of implementation, multiple exceptions and the limited industrial complementation among member countries. The predecessor of the ANCZERTA is a 1965 bilateral free trade agreement which provided for the establishment of a free trade area by 1977 for forest products and selected manufactures, but which left untouched non-tariff barriers. In contrast, ANCZERTA covered all merchandise trade and included tariff and non-tariff barriers, providing external support to far-reaching domestic economic liberalization programs. The 1988 and 1992 reviews went beyond to include, among other issues, non-border barriers, services, anti-dumping procedures, business law and regulatory practices and government procurement. ANCZERTA has been the only successful PTA in the Asian-Pacific region and it displays several peculiarities that are worth noting. The first is its relatively low degree of institutionalization: ANCZERTA rests on a continuous review process that gradually extended its agenda, imposed requirements for increased transparency and fostered consultations when required by any government. The second feature is the weak economic interdependence among the member countries. Finally, New Zealand and Australia have similar institutions and preferences, and both undertook far-reaching economic liberalization. In 1993 the members of the South Asian Association of Regional Co-operation (SAARC) created a South Asian Preferential Trade Area (SAPTA). The agreement aimed at periodic negotiations to exchange trade concessions on a sector or product-by-product basis. In the first round of negotiations held in 1995 concessions -mainly tariff concessions- were minimal. During the second round in 1996 member countries decided to establish a free-trade area by 2005, a deadline advanced to 2000 in 1997. These commitments seem beyond reach considering the relatively high level of protection that prevails in the region, the low volumes of intraregional trade and the tradition of political and even military conflict among member countries. Finally, the Asian-Pacific Economic Co-operation (APEC) forum was established in 1989. Its first four years were devoted to issues of organization, membership and process. At the 1993 Seattle summit the intention to promote a free trade area in the region was first mentioned. One year later in Bogor (Indonesia) a commitment was made to establish free trade by the year 2010, with an extension to 2020 for developing country members. How APEC will progress towards free trade is still unknown. There are at least two conflicting visions on how to go along. While many Asian members prefer the so-called "concerted unilateral approach" (unilateral liberalization on a MFN basis), the United States and others are inclined towards a negotiated liberalization on a reciprocal basis. Internal differences go beyond the formal means toward trade liberalization. Trade regimes, development levels and cultural and political traditions also differ widely among members. PREFERENTIAL TRADE ARRANGEMENTS: AFRICA AND THE MIDDLE EAST Regional integration in Sub-Saharan Africa has a long history, with the Southern Africa Customs Union (SACU) created in 1910 and the East African Community (EAC) established in 1919. SACU still survives but the EAC was dissolved in 1978. Most PTAs established in the 1970s were an integral component of African post-colonial development strategy. This fact explains the large number of arrangements and their different objectives. However, PTAs in Africa and the Middle East have shown little progress either in implementation or in fostering closer economic ties. Indeed, the vast majority of the agreements have failed to materialize and display a high degree of non-compliance. Several reasons can explain this, among others the prevailing development pattern (largely inward-oriented and led by the public sector), complementary production, slow economic growth, divergent regulatory frameworks and poor infrastructure. Diverse political and governance system also limited the effectiveness of regional economic integration. Most African preferential trade arrangements have taken place among Sub-Saharan countries. The Economic Commission for Africa promoted three sub-regional arrangements, two of which, the Economic Commission for West African States (ECOWAS) and the Common Market for Eastern and Southern Africa (COMESA), have been in place for a number of years. Each of these includes, in turn, several sub-regional arrangements. ECOWAS was established in 1975, with sixteen member countries, seven out of which were already part of the French-speaking West African Economic Community (CEAO). Three other countries were members of the Mano River Union. ECOWAS failed to establish free trade among its members. Some of the smaller groupings, such as the West African Economic and Monetary Union (UEMOA)-- which arose out of the demise of CEAO – survived. UEMOA has a parallel Central African organization (the Central African Economic and Monetary Union -UDEAC-), with six members. Overall, intraregional trade is modest and largely concentrated in livestock, agricultural and petroleum products. The Common Market for Eastern and Southern Africa (COMESA) embraces twenty-two countries and began in 1981 as the Preferential Trade Area for Eastern and Southern Africa (PTA). Its original aim was to establish a common market by year 2000. The PTA was best known for its institutional development. Smaller sub-regional arrangements in Southern Africa include the Southern African Development Community (SADC) and the Southern Africa Customs Union (SACU). SADC replaced the Southern African Development Co-ordination Conference (SADCC), that was made up of a set of loose economic cooperative and development arrangements aimed at ending economic dependence on South Africa. In contrast to its predecessor, SADC has adopted measures that involve tariff reductions and the removal of non-tariff barriers. The longevity of the Southern Africa Customs Union (SACU) has been explained by side-payments made by its dominant partner, South Africa. Preferential trade arrangements in Northern Africa and the Middle East date back to the 1957 Agreement for Economic Unity among Arab League States, which comprised twenty-one Arab member countries and provided for the creation of an Arab Common Market. Yet achievements fell short of stated commitments. The most serious attempt to establish a preferential trade arrangement by a North African group of countries was the Arab Maghreb Union (AMU) formed in September 1989, but its success has been limited. Intraregional trade was limited by like resource endowments, limited potential for internal specialization and proximity to European markets. The leading regional agreement in the Middle East has been the Gulf-Cooperation Council (GCC) formed in 1981 by the Arabic gulf states. Its success was slightly higher than that of the AUM, with some preferences exchanged and some progress made in coordinating monetary policy and enabling the free movement of people between member countries. LESSONS FROM PAST EXPERIENCES Drawing clear-cut lessons from past experience is not straight-foward: preferential trade arrangements differ widely in motivations, objectives and scope. Successful PTAs usually share some of the following attributes: they increase trade interdependence, they display a high correlation between commitments and achievements and they include a broadening coverage of issues. The European Community is the classic example, but CUSFTA and ANCZERTA may also qualify as success stories. Failure, in contrast, can be postulated when there is no perceived effect on economic interdependence or there is a large gap between commitments and achievements. In such cases the agreement tends to remain largely symbolic. As in the case of other international institutions, they rarely die. The EC is generally taken as the yardstick to measure success or failure in regional integration. But for many reasons (including historical and political factors) the European experience is highly specific and, probably, non-replicable. Yet a number of conclusions can be drawn from it. Their relevance and applicability need to be assessed in the specific circumstances concerned. 1) Political Commitment. Political motivations have been frequently cited as motoring factors for PTA --- but they are neither necessary nor sufficient conditions for success. A large number of PTAs (particularly among developing countries) had strong political motivations at their origin, but fell rapidly into lethargy. Political motivations played a relatively unimportant role in successful agreements such as CUSFTA. Past experience suggests that rather than political motivation, a key condition for a success is political commitment. Although the basis and motivations for such a commitment may differ, a PTA will most likely fail in its absence. 2) An Enabling Environment. The strength of political commitment will be dependent upon the broader governance and economic policy environment. All success stories involve developed countries with mature governance institutions, stable macroeconomic environments and a relatively high degree of commonality in values and objectives. These environmental factors provide a shared and stable background to implement commitments. When governance institutions are subject to abrupt change, commitments rarely translate into implementation. Similarly, in a highly unstable macroeconomic environment the sustainability of trade policy reforms (unilateral or preferential) is highly uncertain. Lastly, shared values and objectives provide a common framework from which to build. The trade policy environment is a particularly relevant factor, as suggested by the experience of the EC, CUSFTA and ANCZERTA. Relatively low protection in member countries will contribute to the success of a preferential trade arrangement because it will mean less resistance to lowering tariffs. A shift towards trade liberalization is also likely to dissolve opposition towards a PTA. The latter may also (but not necessarily will) turn into a supporting factor for overall trade liberalization. 3) Institutions and PTAs. The kind of institutionalization adopted by the EC is often referred to as a key to success. But conventional stories about supranational authorities and the role of the European Court of Justice (ECJ) are exaggerated. In reality, EC institutions are the result of an uneasy, evolving mix of inter-governmental and supranational competence and decision-making. Compliance with EC regulations or ECJ decisions has been high, but lower than implied by conventional stories about the EC. Moreover, after almost two decades of painful harmonization, the EC abandoned centralized regulations and adopted the principle of mutual recognition. The existence of a bureaucratic locus representing the "common interest" has also been highlighted as a lesson from the EC success. But neither CUSFTA nor ANCZERTA has such a "common interest" locus. Moreover, both have rather slim institutions. Although not all share the view that the EC bureaucracy was a key to progress in economic integration, there is a widespread agreement that it played an important role at times of stagnation. 4. Distribution of Costs and Benefits. The issue of the distribution of costs and benefits occupied a central role in the EC experience. The existence of institutions and mechanisms addressing distributive considerations is generally regarded as having prevented the latter from becoming a barrier to more integration. Occasionally, distributive considerations have been addressed at the expense of more liberal trade policies (such as the Common Agricultural Policy). In the case of the EC, "variable geometry" arrangements have also been used as a means to circumvent the dilemmas posed by distributive issues or disparate preferences or policy priorities. Other PTAs, such as CUSFTA and ANCZERTA, have placed more trust in market mechanisms. Whether this is a viable option for PTAs involving countries with large differences in per-capita incomes and/or levels of development may be clarified by NAFTA’s record. In PTAs among developed countries, the issue of the distribution of costs and benefits has been facilitated by intraindustry specialization. Intraindustry trade involves less income redistribution and less dislocation of factors of production, thus limiting the extent of domestic opposition to trade liberalization. 5. Deepening and Leveling the Playing Field. Functional interpretations of the EC experience underline that successful economic integration calls for continuous deepening. As market integration develops, the demand for a more stable and predictable environment increases, thus stimulating a deepening of the scope of regional agreements. Similarly, as market integration grows demands for a level playing field mount. Yet for the issue of policy harmonization to be effectively addressed from the start, a dominant partner is necessary. If this is not the case, harmonization is unlikely to be fully dealt with from the beginning. Eventually, as demands mount policy harmonization and deepening initiatives may be negotiated and enforced. During the initial phases of an agreement the issue of market access must bear absolute priority. If market access cannot be guaranteed, harmonization and deepening initiatives will remain paper commitments. CONCLUSIONS PTAs are the result of a variety of motives, which include political, strategic, bargaining and trade policy considerations. PTAs have rarely resulted from a careful objective examination of economic costs and benefits. But the existence of net economic benefits and the issue of their distribution among member countries have turned into important factors as agreements were implemented. Political motivations may not be necessary to launch an agreement, but political commitment is a necessary condition for success. This makes mature institutions, a stable macroeconomic environment and a relatively high degree of commonality of values and objectives important preconditions. Liberal trade regimes (or a shift towards liberalization) are also necessary to succeed in preferential liberalization. The institutional design of successful PTAs differs. At one extreme is the EC, with highly structured institutions and supranational decision-making. Other successful initiatives have adopted alternative frameworks. Considering the multitude of failed regional integration experiences, institutional design does not appear to be a major explanatory factor. The revival of PTAs in the last decade displays two major novelties. The first is the broad coverage of issues, concomitant to the extension of the international trade agenda. The other is the adoption by the United States of PTAs as a complement to its traditional multilateral stance. This latter fact deserves close attention, as its evolution will influence whether the present regionalist drive will complement or conflict with the multilateral system. At the end, the question of whether PTAs will be building or stumbling blocks to an effective multilateral system will remain an empirical issue.