INTRODUCTION

By June 1, 1995, the World Trade Organization (WTO) – the successor institution to General Agreement on Tariffs and Trade (GATT) – had been notified of more than 130 regional arrangements (Bhagwati and Panagariya, 1996b). Today, with the exception of Hong Kong and Japan, every WTO member boasts of a membership in at least one regional arrangement. Neither the successful completion of the Uruguay Round nor the recent peso crisis in Mexico has been able to reverse the rising tide of regionalism. Having already concluded Free Trade Area (FTA) agreements with its neighbors in Central and Eastern Europe, the European Union (EU) is now beginning to bring the countries in North Africa into its fold. The United States, on the other hand, has set a target date of 2005 for the creation of the Free Trade Area of the Americas (FTAA), consisting of the entire Western Hemisphere. Germany has even made overtures for a Trans-Atlantic Free Trade Area (TAFTA) between the United States and the EU.

Historically, East Asia has not been enthusiastic toward preferential trading. Until recently, the only functioning preferential trading scheme in the region was the Association of Southeast Asian Nations (ASEAN) Preferential Trading Area (APTA), which was very limited in scope. Starting in 1989, the rise of regionalism in North America and Europe led to some rethinking in the region. In 1989, the forum for Asia Pacific Economic Cooperation (APEC) was created to bring the Uruguay Round to a speedy conclusion and to promote liberalization of trade and investment policies at the regional level. APEC, which has eighteen members currently, includes countries from not only East Asia but also the Americas.

Though the theory of preferential trading had its birth in Jacob Viner's (1950) celebrated work for the Carnegie Endowment, The Customs Union Issue, the first complete general-equilibrium model of preferential trading was provided by James Meade (1955) in the de Vries Lectures, delivered at the Netherlands School of Economics while the Benelux union was in progress and published as The Theory of Customs Unions. Remarkably, at a time when two-good models dominated the thinking of international trade theorists, Meade constructed a complete three-good, three-country model and even went on to extend it to a multicountry, multicommodity context. The model has proved as durable as Viner's concepts of trade creation and trade diversion with Lipsey (1958, chaps. 5–6, 1960), Mundell (1964), Vanek (1965, Appendix), Corden (1976), Collier (1979), McMillan and McCann (1981), and Lloyd (1982) making significant contributions to its further development. Insights emerging out of the model have also shaped the policy debate on regional integration (see Bhagwati and Panagariya, 1996a).

Peter Kenen has much in common with Meade. Like Meade, he has the unusual distinction of having advanced both branches of international economics: pure trade and finance. Those of us who had the opportunity to sit through his lectures on international trade can also recall his clever use of geometry, as was done by Meade (1952) in A Geometry of International Trade.

Introduction

Three decades ago, under the impetus of European arrangements, the developing world launched the first wave of regional integration (RI). Free-trade areas (FTAs) and customs unions (CUs) mushroomed in Latin America and Africa. Unfortunately, expectations of economic development through RI were not realised, and two decades later virtually all regional arrangements among developing countries were judged as failures.

By the early 1980s, multilateral tariff cutting by developed countries and unilateral trade liberalisation (UTL) by developing countries had substantially weakened the case for regional arrangements. Yet, paradoxically, it is then that a second round of regionalism got under way. More arrangements (eight) were signed during the 1980s than during the 1960s, and still more (half a dozen or so) are under consideration. The GATT process is running out of steam and many countries are turning back to the bilateral alternative.

From the viewpoint of developing countries, the current regionalism differs from the regionalism of the 1960s in two important respects. First, the regionalism of the 1960s represented an extension of the importsubstitution- industrialisation strategy from the national to the regional level and was therefore inward-looking. The current regionalism is by contrast taking place in an environment of outward-oriented policies. Second, in the 1960s, developing countries pursued RI exclusively with other developing countries. Today these countries, especially those in Latin America, have their eyes on integration with large, developed countries.

In this chapter, we review and extend the theory of RI and evaluate empirically its contribution to growth. Our objective is to assess the benefits of RI from the viewpoint of participating countries rather than the world as a whole.

The cornerstone of the post-World War II world trading system is the most favoured nation (MFN) clause that underlies all GATT-negotiated reductions in tariffs. Through the MFN clause, non-discrimination creates global order out of an essentially mercantilist system. Most importantly perhaps from the standpoint of lesser nations with little bargaining power, through the application of the MFN clause, purely bilateral bargains negotiated under the auspices of the GATT become available to all. Enormous progress towards global free trade has been made under this non-discriminatory approach to tariff reductions. For example, between the opening of the first round of multilateral talks in Geneva in 1947 and the close of the Tokyo round in 1979, the average US tariff had declined by nearly 92 percent. On the eve of the Uruguay round in 1987, the average tariff for the United States was 4.9 percent, for the European Community it was 6.0 percent, and for Japan it was 5.4 percent.

This long period of progressive trade liberalisation also witnessed two waves – waves because they came suddenly, and the first one at least, foundered – of regional trading arrangements. The first came during the early 1960s. Under the impetus of the European Common Market, regionalism spread throughout Africa, Latin America and other parts of the developing world. The United States was then a hegemon and a staunch supporter of multilateralism. Regionalism then came to a halt during much of the 1970s.

The second wave started in the middle 1980s. This time around, the United States was a major player.

This volume is the outcome of a World Bank–CEPR conference held at the World Bank, Washington DC, on 2–3 April 1992. Partial funding for the conference was provided by the Research Administration of the World Bank under RPO 677-12. The support is gratefully acknowledged. We owe special thanks to Ghislaine Bayard, Karla Cabana and Rebecca Sugui for a flawless administration of the conference. The large number of congratulatory messages we received after the conference were testimony to the superb organisation they carried out and to their untiring contribution to activities related to the production of the volume on the Western side of the Atlantic.

The editing, coordination and production of the volume have been carried out on the Eastern side of the Atlantic at CEPR. David Guthrie and Kate Millward were most effective at ensuring that authors (including the usually delinquent editors) delivered their manuscripts and revisions on time. Barbara Docherty, as Production Editor, was swift and thorough in editing the manuscripts.

Finally, we wish to thank Francis Ng of the World Bank for his superb statistical assistance on a number of chapters in the volume.

Introduction

This paper is part of an ongoing research effort aimed at analysing the interactions among commodity exports, real incomes and trade policies. We are particularly interested in evaluating the concern that efficiency or policy-induced changes in the supply of exports of primary commodities may lead to such a large decline in the prices of the latter that export revenues and incomes of the exporting countries actually decline. The commodities in question include cocoa, coffee, and tea.

The possibility of income and revenue loss arises principally because the world demand for primary commodities is relatively inelastic. The natural instrument the exporting countries can employ to counteract this problem is trade policy. In doing so, one must, however, recognize that since the exporting countries are neither small nor monopolies in the world markets, their policies are interdependent. Because the conventional trade models rely on one of these two extremes, they fail to capture the interdependence of policies central to the problem under consideration.

In our recent work, we have analysed in detail the implications of interdependence among countries when the policy instrument is an export tax. In Panagariya and Schiff (1991a), we derive Nash optimum taxes in a 10-country model of the world cocoa market. In our simulations, we find that compared to the initial equilibrium, tax competition implicit in Nash behaviour leads to a loss in real income for 8 out of 9 exporting countries. In Panagariya and Schiff (1991b), we compare income and revenue-maximizing Nash taxes. A key result in this paper is that under plausible circumstances, revenue-maximizing Nash taxes can yield higher levels of income than income-maximizing Nash taxes.